No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Real-time estimation of the vehicle sideslip angle through regression based on principal component analysis and neural networks
... Recently, neural network techniques have been vigorously applied to estimating various vehicle states [14]- [21]. Neural network structures can overcome the limitations of empirical methods because they can identify data characteristics that a human observer cannot detect [22]. ...
... Another effort attempted to solve the lateral state estimation problem of a preceding target vehicle using multiple neural networks, consisting of a nonlinear autoregressive exogenous model net, feedforward net, and Elman net [18]. In addition, hybrid approaches, combining a neural network with other methods, have been applied to estimate vehicle states [19]- [21]. Vehicle roll dynamics based unscented Kalman filter coupled with an artificial neural network provided a good estimation of vehicle roll angles [19]. ...
... A recurrent neural network combined with a vehicle kinematic model was trained using simulation data to estimate vehicle sideslip angles [20]. A principal component analysis was adopted for the preprocessing of input data of a neural network to estimate vehicle sideslip angles [21]. These methods reportedly lowered computational loads with increased accuracy. ...
This study employs a dual deep neural network (D-DNN) to accurately estimate the absolute longitudinal speed of a vehicle. Accuracy in speed estimation is crucial for vehicle safety, because longitudinal speed is a common parameter employed as a state variable in active safety systems such as anti-lock braking system and traction control system. In this study, DNNs are applied to determine the gain of an adaptive filter to estimate vehicle speed. The used data consists of longitudinal acceleration, wheel speed, filter gain, and estimated vehicle speed. The data generated from Carsim software are collected and preprocessed using a Simulink model. To acquire data with numerous wheel slip patterns, various acceleration and deceleration conditions are applied to four different road conditions. Though, it is challenging to achieve a single DNN model that is optimally cope with the various driving situations. Thus, we adopt two DNN models that were individually trained in low and high acceleration regions. The dual DNN model results in error reductions of 74% and 65%, compared with a single DNN and classical adaptive Kalman filter approaches, respectively.
... To improve estimation robustness, data preprocessing or vehicle state identification is considered. In this regard, Martino et al. employed the principle component analysis (PCA) to reduce the number of dimensions of raw training data [95]. Bonfitto et al. [96] presented a NN-based algorithm in tandem with road condition identification. ...
... In order to reduce the number of model input features and improve computational efficiency, a selection of signals with high correlation is necessary. The PCA method is used to verify whether the parameters have high relevance with sideslip angle [95]. In this work, since the PCA is not suitable for online estimation in the VCU, the covariance matrix is used to analyse the correlation between different parameters and vehicle sideslip angle, in which high eigenvalue indicates high correlation and vice versa. ...
Vehicle sideslip angle is a major indicator of dynamics stability for ground vehicles; but it is immeasurable with commercially-available sensors. Sideslip angle estimation has been the focus of intensive research in past decades, resulting in a rich library of related literature. This paper presents a comprehensive evaluation of state-of-the-art sideslip angle estimation methods, with the primary goal of quantitatively revealing their strengths and limitations. These include kinematics-, dynamics- and neural network-based estimators. A hardware-in-loop system is purposely established to examine their performance under four typical maneuvers. The results show that dynamics-based estimators are suitable at low vehicle velocities when tires operate in the linear region. In contrast, the kinematics-based methods yield superior estimation performance at high vehicle velocities, and the inclusion of dual GPS receivers is beneficial even when there is large disturbance to the steering angle. Of utmost importance, it is experimentally manifested that the neural network-based estimator can perform well in all maneuvers once the training datasets are properly selected.
... The side-slip angle is a variable notoriously hard to directly measure, so its availability for an actual control system would depend on the estimation of the variable based on other output signals. This estimation is not the focus of this paper, so it will be assumed that β is available with sufficiently high precision, as it is demonstrated to be achievable by some works present in the literature [28,29]. ...
In today’s automotive industry, electrification is a major trend. In-wheel electric motors are among the most promising technologies yet to be fully developed. Indeed, the presence of multiple in-wheel motors acting as independent actuators allows for the implementation of innovative active systems and control strategies. This paper analyzes different design possibilities for a torque vectoring system applied to an originally compact front-wheel drive hybrid electric vehicle with one internal combustion engine for the front axle and two added electric motors integrated in the wheels of the rear axle. A 14 degrees of freedom vehicle model is present o accurately reproduce the nonlinearities of vehicle dynamic phenomena and exploited to obtain high-fidelity numerical simulation results. Different control methods are compared, a PID, an LQR, and four different sliding mode control strategies. All controllers achieve sufficiently good results in terms of lateral dynamics compared with the basic hybrid version. The various aspects and features of the different strategies are analyzed and discussed. Chattering reduction strategies are developed to improve the performance of sliding mode controllers. For a complete overview, control systems are compared using a performance factor that weighs control accuracy and effort in different driving maneuvers, i.e., ramp and step steering maneuvers performed under quite different conditions ranging up to the limits.
... Different layouts of a neural network that estimates vehicle longitudinal velocity have been trained, tested and validated against actual measurements obtained from a fully instrumented car on four diverse tracks [45]. Several computational methods for vehicle longitudinal velocity estimation with diverse input dataset configurations and training parameters are presented and compared to the measured signal of interest. ...
Vehicle dynamics control systems have a fundamental role in smart and autonomous mobility, where one of the most crucial aspects is the vehicle body velocity estimation. In this paper, the problem of a correct evaluation of the vehicle longitudinal velocity for dynamic control applications is approached using a neural networks technique employing a set of measured samples referring to signals usually available on-board, such as longitudinal and lateral acceleration, steering angle, yaw rate and linear wheel speed. Experiments were run on four professional driving circuits with very different characteristics, and the vehicle longitudinal velocity was estimated with different neural network training policies and validated through comparison with the measurements of the one acquired at the vehicle’s center of gravity, provided by an optical Correvit sensor, which serves as the reference (and, therefore, exact) velocity values. The results obtained with the proposed methodology are in good agreement with the reference values in almost all tested conditions, covering both the linear and the nonlinear behavior of the car, proving that artificial neural networks can be efficiently employed onboard, thereby enriching the standard set of control and safety-related electronics.
... • Direct measure using a sideslip angle sensor-as in [45,46]-but this solution is usually very expensive and cumbersome to be installed in a road vehicle (usually used for research and testing purposes). ...
... The first sensor applied to the experimental setup is the Kistler S-Motion. It combines a complete inertial platform (with 3-axis accelerometers and gyroscopes) with a lamp module that can obtain a direct measure of the longitudinal and lateral velocities by processing a sequence of road images using the correvit principle [36,37]. To do so, the lamp module must be positioned at a 350±50 mm height facing the road, condition achieved in this case by a suction fixture in the rear door of the vehicles (Figure 13). ...
... The developed algorithm, based on the T.R.I.C.K. methodology described in [23], allows us to evaluate in a specifically dedicated on-board module the fundamental kinematic and dynamic quantities for the tire characterization in real time, starting from the experimental signals available within the vehicle CAN bus (Controller Area Network) and s-motion measurement or, as the case in exam, employing a specific set of sensors pre-configured on the vehicle. Such methodology also allows us to evaluate the potential of an estimation process in terms of tire interaction curves, such as in [48]. ...
In recent years, autonomous vehicles and advanced driver assistance systems have drawn a great deal of attention from both research and industry, because of their demonstrated benefit in reducing the rate of accidents or, at least, their severity. The main flaw of this system is related to the poor performances in adverse environmental conditions, due to the reduction of friction, which is mainly related to the state of the road. In this paper, a new model-based technique is proposed for real-time road friction estimation in different environmental conditions. The proposed technique is based on both bicycle model to evaluate the state of the vehicle and a tire Magic Formula model based on a slip-slope approach to evaluate the potential friction. The results, in terms of the maximum achievable grip value, have been involved in autonomous driving vehicle-following maneuvers, as well as the operating condition of the vehicle at which such grip value can be reached. The effectiveness of the proposed approach is disclosed via an extensive numerical analysis covering a wide range of environmental, traffic, and vehicle kinematic conditions. Results confirm the ability of the approach to properly automatically adapting the inter-vehicle space gap and to avoiding collisions also in adverse road conditions (e.g., ice, heavy rain).
... The performance of the monitoring system was verified with field tests. De Martino et al. [8] presented a method for real-time estimation of the vehicle sideslip angle based on principal component analysis and a neural network, the inputs of which were the steering wheel angle, lateral and longitudinal acceleration, wheel angular velocity, and yaw rate. The effectiveness of the method was verified by a comparison of the sideslip angle provided by an onboard optical Correvit sensor (Kistler Instrument Corp., USA). ...
This study aimed to develop a coach state estimation and prediction system to enhance driving safety. Different from existing vehicle stability estimation studies, the authors propose a hybrid method to estimate and predict the state of a coach in real time. First, the vehicle sideslip angle and yaw rate are estimated by a three‐degrees‐of‐freedom vehicle model combined with an extended Kalman filter (EKF) estimation algorithm. Then, a steering system is established that replaces the front‐wheel angle with the steering wheel input torque. Next, a seven‐degrees‐of‐freedom vehicle model analyses the effects of various driving influencing factors on the vehicle sideslip angle and the boundary of the stable region of the phase plane of the vehicle sideslip angle rate, and a boundary value parameter database is obtained. A back propagation neural network (BPNN) model is then established to obtain the boundary function parameter values under multifactor coupling conditions. Furthermore, an online prediction of the steering wheel input torque in a time series is done, and the prediction value is input to the steering system and neural network model. The effectiveness of the proposed method was evaluated via simulations based on MATLAB/Simulink and TruckSim software.
... signals that are implicitly correlated with vehicle sideslip angle are available from the Carsim. In order to reduce the number of model inputs and improve computational efficiency, the PCA method is used to select the most highly relevant signals[15]. The covariance matrix is used to analyze the correlation between different signals with vehicle sideslip angle, in which high eigenvalue indicates high correlation and vice versa. ...
In this paper, a time-delay neural network (TDNN) is proposed to estimate the sideslip angle under extreme maneuvers for in-wheel-motor-drive electric vehicles (IWMD EVs). The principle component analysis (PCA) method is first utilized for data preprocessing. Then a time delay module is introduced into the neural network model to improve its robustness. The estimated sideslip angle is further filtered by the Kalman filter. Finally, the proposed estimation scheme is verified via the comprehensive hardware-in-loop (HIL) tests, in which the proposed method can achieve high estimation accuracy.
div>To address the challenge of directly measuring essential dynamic parameters of vehicles, this article introduces a multi-source information fusion estimation method. Using the intelligent front camera (IFC) sensor to analyze lane line polynomial information and a kinematic model, the vehicle’s lateral velocity and sideslip angle can be determined without extra sensor expenses. After evaluating the strengths and weaknesses of the two aforementioned lateral velocity estimation techniques, a fusion estimation approach for lateral velocity is proposed. This approach extracts the vehicle’s lateral dynamic characteristics to calculate the fusion allocation coefficient. Subsequently, the outcomes from the two lateral velocity estimation techniques are merged, ensuring rapid convergence under steady-state conditions and precise tracking in dynamic scenarios. In addition, we introduce a tire parameter online adaptive module (TPOAM) to continually update essential tire parameters such as cornering stiffnesses, with its effectiveness demonstrated through DLC and slalom simulation tests. Using a dual extended Kalman filter (DEKF) observer, the article allows for joint estimation of vehicle states and tire parameters. Ultimately, we offer a cost-effective estimation method of vital dynamic vehicle parameters to support the motion control module in autonomous driving.</div
This paper presents a new velocity and pose sensor for a ground vehicle using modulated motion blur generated from an optical image modulator (OIM). The OIM repeatedly shakes the lens of a camera in the lateral direction of the vehicle body, which creates patterns of motion blur that are recorded in an image frame. The relationship between the blur pattern and vehicle velocity is determined using camera dynamics. The 2D velocity of the vehicle body is estimated by analyzing the motion blur pattern. The vehicle roll and pitch angle with respect to the ground are calculated by comparing the estimated average depths (heights) of the sections of a scene. We propose a signal processing and control system to precisely modulate the OIM, which is an electro-mechanical system consisting of a coil, magnet, spring and damper. In addition, we validate the influence of the control reference waveform through various experiments and confirm the robustness of the proposed system. The proposed system is suitable for many applications because it is insensitive to environmental changes and has a relatively low computation time while accurately measuring the speed.
The vehicle dynamics simulation and preview control require a dedicated vehicle model, such as multibody dynamics model. However, the multibody model has higher computational complexity which affects the response time of the vehicle controller. This issue can be alleviated by using a data-driven vehicle dynamics model due to its effective generalization and computational speed. In this work, we propose a data-driven modeling approach based on deep neural networks (DNNs) for computing and predicting the vehicle characteristics. The high-fidelity simulations of a validated vehicle multibody model are performed for data acquisition. This data is then used for training and testing the proposed model. The DNN inputs comprise the initial speed of the vehicle and the torque applied on front wheels to imitate vehicle acceleration and deceleration. The DNN outputs comprise the driving distance and the longitudinal velocity of the vehicle. The dynamics characteristics resulting from both the data-driven model and the multibody model are investigated and compared. Furthermore, the accuracy of the data-driven model is analyzed in terms of various error functions. The data-driven model is verified by using the results obtained from a commercial software package. The simulation results show that the data-driven vehicle model predicts the accurate velocity and driving distance in real-time. The data-driven model can be used for real-time simulation and preview control in autonomous vehicles.
Robust estimation of vehicle sideslip angle is essential for stability control applications. However, the direct measurement of sideslip angle is expensive for production vehicles. This paper presents a novel sideslip estimation algorithm which relies only on sensors available on passenger and commercial vehicles. The proposed method uses both kinematics and dynamics vehicle models to construct extended Kalman filter observers. The estimate relies on the results provided from the dynamics model observer where the tire cornering stiffness parameters are updated using the information provided from the kinematics model observer. The stability property of the proposed algorithm is discussed and proven. Finally, multiple experimental tests are conducted to verify its performance in practice. The results show that the proposed approach provides smooth and accurate sideslip angle estimation. In addition, our novel algorithm provides reliable estimates of bank angles and lateral acceleration sensor bias.
Extensive studies have been conducted on unmanned vehicle recently due to its bright prospect. However, few have studied the technology of driverless bus. In this paper, a path planning and navigation control system is designed for 12 meters length driverless electric bus. The global path planning is adopted by Dijkstra algorithm based on analysis tool of ArcGIS. For local planning, a triple-layer finite state machine is proposed to plan the driving maneuver, and it reduces computational complexity and improves driving effectiveness. Meanwhile, optimal trajectory is generated by curve fitting technique based on path planning and driving maneuver, which takes full account of the safety and kinetics for driverless bus. And the speed planning by expert rules are addressed to imitate an experienced driver. Furthermore, the trajectory tracking controller is devised by an adaptive rolling window. Moreover, a heading error compensation model and self-adaptive strategy of controller parameters are set up to improve the control precision. Finally, simulation by TruckSim and scenes experiments verify the validity and practicability of our proposed path planning and navigation control system for driverless bus.
Driving safety can be achieved by better understanding critical situations which may require the knowledge of interaction between vehicle tyres and the road surfaces. It is thus essential to have a good estimation of the tyre/road friction parameters in real-time. The paper deals with the trust-region based method for on-line estimation of tyre/road friction parameters. This method provides an appropriate modeling (of a vehicle and the tyre/road contact) to observe the tyre/road friction coefficients directly using measurable signals in real-time. In this work, we present a new LuGre model-based nonlinear least squares (NLLS) parameter estimation algorithm using vehicle dynamic to obtain the parameters of LuGre model based on recursive nonlinear optimization of the curve fitting errors. The proposed estimation method can also be utilized in large-scale problems with similar conditions. Very promising results have been obtained in real-time simulations for most of the driving and road situations.
Electric vehicles with multiple motors allow torque-vectoring, which generates a yaw moment by assigning different motor torques at the left and right wheels. This permits designing the steady-state cornering response according to several vehicle handling quality targets. For example, as widely discussed in the literature, to make the vehicle more sports-oriented, it is possible to reduce the understeer gradient and increase the maximum lateral acceleration with respect to the same vehicle without torque-vectoring. This paper focuses on the novel experimentally-based design of a reference vehicle understeer characteristic providing energy efficiency enhancement over the whole range of achievable lateral accelerations. Experiments show that an appropriate tuning of the reference understeer characteristic, i.e., the reference yaw rate of the torque-vectoring controller, can bring energy savings of up to ~11% for a case study fourwheel-drive electric vehicle demonstrator. Moreover, during constant speed cornering, it is more efficient to significantly reduce the level of vehicle understeer, with respect to the same vehicle with even torque distribution on the left and right wheels.
This paper presents some results on the development and testing of new solutions in the field of driving automation. The introduction of increasing levels of vehicle automation aimed at enhancing road safety requires a renewed approach to the research and development process and needs a multi-actor environment where the innovation can be tested. Indeed, vehicle automation spans several scientific disciplines and it is becoming exceedingly difficult and too costly for a single research innovation team to go in depth into all technologies and solutions. This is shifting the innovation process toward a multidisciplinary approach in which the only way to ensure an easy, rapid, efficient and scalable introduction of the required innovation is to adopt integrated and complex testing platforms for the simulation of automation solutions, based on a modular architecture, where independent components can be developed and then integrated and tested in a multi-actor environment. A platform for virtual testing is presented herein and employed to assess the performance of an integrated driving assistance solution based on computing appropriate surrogate measures of safety that allow for the transition between different automation logics in free-flow, car-following and emergency braking conditions.
A car-following model is formulated as a time-continuous dynamic process, depending on two parameters and two inputs. One of these inputs is the follower’s desired equilibrium spacing, assumed to exist and to be known. Another input is the speed of the lead vehicle. Given the formulation of the model, the contribution of these two inputs is separable from an analytical point of view. The proposed model is simple enough (while not being simplistic) to support real-time applications in the field of Advanced Driving Assistance Systems (ADAS). Starting from the equilibrium spacing it is possible to estimate the parameters of the model, allowing for a full identification procedure. The modelling framework was pre-validated against observed data from two different datasets, collected by means of two instrumented vehicles in independent experiments, carried out in Italy and the United Kingdom. The validation proved that the proposed car-following model gives good results not only around the desired equilibrium spacing, but also in general car-following conditions. Results are discussed against observed data in terms of values of obtained parameter, as well as with respect to the performance of the resulting dynamic process.
Accurate measurement of the vehicle sideslip angle is fundamental to improve reliability of the vehicle dynamics control systems focused on stability and developed both for safety and performance optimization. Many experimental procedures to estimate the vehicle sideslip angle have been proposed in the last years, mainly based on GPS, INS and physical models. The aim of this paper is to compare different methods to estimate sideslip angle employing an instrumented vehicle, equipped with a system for data acquisition and time-synchronized storage capabilities, a stand-alone GPS, a GPS aided MEMS-based Attitude and Heading Reference System (AHRS) and specific sensors to collect data on the steering wheel angle and on the position of brake, throttle and clutch pedals. Further information is collected by capturing the available data at the OBD port of the vehicle. Data acquisitions (from all sensors) are synchronized by means of an external triggering signal. After driving sessions performed with specific manoeuvres in order to highlight the main phenomena concerned with the dynamic behaviour of the vehicle, the different estimation procedures have been applied, discussing on the advantages and the degree of reliability of each one of them.
In the paper new structure elements have been developed and implemented in the already-existing TRT thermo-dynamic tyre model. The updated model aims to provide a complete tool to study and understand all the phenomena concerning the tyre in thermal transient conditions, since all the elements constituting its structure are finally modelled. The computational cost, connected to a more complex model to manage, was decreased by simplifying the mesh of the previous version of the model and, thus, by reducing the state vector length.
In this paper we propose a control architecture to stabilize a vehicle during cornering near the limit of lateral acceleration using the rear axle electric torque vectoring configuration of a hybrid vehicle. A vehicle model incorporating nonlinear tyre characteristics and coupling of tyre forces along longitudinal and lateral directions is used to calculate reference steady-state cornering conditions, as well as to design a linear controller with wheel slip ratio inputs. A backstepping controller then provides the necessary motor torques to achieve the wheel slip ratio requested by the linear controller. The controller provides stability of the lateral vehicle dynamics and regulates the longitudinal velocity to ensure feasibility of the reference trajectory.
Book available at HALF PRICE (50% off), shipping included worldwide till MARCH 31, 2016 here
http://www.springer.com/en/book/9789401785327 (Springer website)
THE MORE YOU KNOW VEHICLE DYNAMICS, THE MORE YOU'LL BE SURPRISED
Vehicle dynamics is often perceived as a quite intuitive subject. As a matter of fact, lots of people are able to drive a car. Nevertheless, without a rigorous mathematical formulation it is very difficult to truly understand the physical phenomena involved in the motion of a road vehicle.
In this book, mathematical models of vehicles are developed, always paying attention to state the relevant assumptions and to provide explanations for each step. This approach allows for a deep, yet simple, analysis of the dynamics of vehicles, without having to resort to foggy concepts. The reader will soon achieve a clear understanding of the subject, which will be of great help both in dealing with the challenges of designing and testing new vehicles and in tackling new research topics.
However, there is much more than that. Quite surprisingly, it is shown that several classical concepts, such as the understeer gradient or the roll axis, are either wrong or inadequate and need to be replaced.
The book covers handling and performance of both road and race cars. A new approach, called MAP (Map of Achievable Performance), is presented and thoroughly discussed. It provides a global and intuitive picture of the handling features of a vehicle. Moreover, the book also deals with several relevant topics in vehicle dynamics that have never been discussed before. Even very experienced people should find the book interesting and stimulating.
This new book is not a translation of the Italian Dinamica del Veicolo; by the same author. Actually, in some sense, this new book is totally different, with new topics and with new points of view for the topics covered in the Italian book as well.
In this paper an experimental test rig aimed to characterize mechanical properties of a pneumatic tyre, together with some results, is presented.
The objective is to determine tyre mechanical characteristics useful to physically model its behaviour; in particular: the normal interaction characteristic, the radial stiffness, the total stiffness and the longitudinal hysteretic cycles.
To this aim two different kind of tests have been executed: radial and longitudinal.
In the radial test the load is statically applied to the tyre, along the vertical direction, by means of an hydraulic press and it is measured together with the consequent radial deformation, so allowing the estimation of the tyre normal interaction characteristic and of its radial stiffness. Different radial tests can be conducted for an assigned tyre varying the inflation pressure.
The longitudinal tests are conducted applying, under an assigned constant vertical load, a variable horizontal strain to the tyre by means of a linear actuator, two profile rail guides and a system to transfer the horizontal motion to the contact patch of the tyre, opportunely placed on a moving steel plate placed on the two linear guide rails. During the tests the horizontal load and the resulting deformations are measured and acquired so allowing the estimation of tyre total stiffness and of its longitudinal hysteretic cycles.
Longitudinal tests can be conducted varying the assigned vertical load, the horizontal displacement law in terms of frequency and amplitude, the tyre inflation pressure.
All the different types of rim can be mounted on the test rig thanks to a universal quick flange.
In this article, a logic for vehicle dynamics control during partial braking while turning a corner is presented, which only requires knowledge of the instantaneous speed of the four wheels. For this reason, the proposed control algorithm can be adopted on all ABS equipped cars. A scheme of the simulation program for logic validation is described, which is constituted by a loop of software models of the principal vehicle subsystems which are singly illustrated. The proposed logic has been tested both in closed and open-loop maneuvers. The results are provided in the form of time histories of the principal analyzed quantities. The analysis of the results confirms the goodness of the proposed control strategy.
The objective of this article is to develop a vehicle sideslip observer that takes the nonlinearities of the system into account, both in the theoretical analysis and the design. The design goals include reduction of the computational complexity compared to the EKF, to make the observer suitable for implementation in the embedded hardware, and a reduction in the number of tuning parameters compared to the EKF. Design is based on a standard sensor configuration, and is subjected to the extensive testing in the realist conditions.
In the original version of the book, the Chapters 2, 3, 5, 6, 7, 8, 9 and 10 were revised.
The series Advances in Industrial Control aims to report and encourage technology transfer in control engineering. The rapid development of control technology impacts all areas of the control discipline. New theory, new controllers, actuators, sensors, new industrial processes, computer methods, new applications, new philosophies, .... , new challenges. Much of this development work resides in industrial reports, feasibility study papers and the reports of advanced collaborative projects. The series offers an opportunity for researchers to present an extended exposition of such new work in all aspects of industrial control for wider and rapid dissemination. Within the control community there has been much discussion of and interest in the new Emerging Technologies and Methods. Neural networks along with Fuzzy Logic and Expert Systems is an emerging methodology which has the potential to contribute to the development of intelligent control technologies. This volume of some thirteen chapters edited by Kenneth Hunt, George Irwin and Kevin Warwick makes a useful contribution to the literature of neural network methods and applications. The chapters are arranged systematically progressing from theoretical foundations, through the training aspects of neural nets and concluding with four chapters of applications. The applications include problems as diverse as oven tempera ture control, and energy/load forecasting routines. We hope this interesting but balanced mix of material appeals to a wide range of readers from the theoretician to the industrial applications engineer.
Electric vehicles with individually controlled drivetrains allow torque-vectoring, which improves vehicle safety and drivability. This paper investigates a new approach to the concurrent control of yaw rate and sideslip angle. The proposed controller is a simple single input single output (SISO) yaw rate controller, in which the reference yaw rate depends on the vehicle handling requirements, and the actual sideslip angle. The sideslip contribution enhances safety, as it provides a corrective action in critical situations, e.g., in case of oversteer during extreme cornering on a low friction surface. The proposed controller is experimentally assessed on an electric vehicle demonstrator along two maneuvers on surfaces with significantly varying tire-road friction coefficient. Different longitudinal locations of the sideslip angle used as control variable are compared during the experiments. Results show that: i) the proposed SISO approach provides significant improvements with respect to the vehicle without torque-vectoring, and the controlled vehicle with a reference yaw rate solely based on the handling requirements for high-friction maneuvering; and ii) the control of the rear axle sideslip angle provides better performance than the control of the sideslip angle at the center of gravity.
The paper discusses novel computationally efficient torque distribution strategies for electric vehicles with individually controlled drivetrains, aimed at minimising the overall power losses while providing the required level of wheel torque and yaw moment. Analytical solutions of the torque control allocation problem are derived and effects of load transfers due to moderate driving/braking and cornering conditions are studied and discussed in detail. Influences of different drivetrain characteristics on the front and rear axles are described. The analytical solution of the control allocation problem is experimentally validated along multiple driving cycles on an electric vehicle with four identical drivetrains. The results of the analytically-derived algorithm are contrasted with those from two other control allocation strategies, based on the off-line numerical solution of more detailed formulations of the control allocation problem (i.e., a multi-parametric non-linear programming problem). The experiments show that the computationally efficient analytical solution represents a very good compromise between energy efficiency, drivability and controller complexity.
This paper deals with the design of a brake caliper pressure controller for a conventional anti-lock braking system/electronic stability control system and the experimental validation of its tracking performances. The analysis of the hydraulic plant, carried out in part I of this two-part study, is here utilized to develop the control algorithm for an anti-lock braking system digital electro-hydraulic valves. The control strategy is based on a feed-forward and a proportional integral controller through pulse width modulation with a constant frequency and variable duty cycle. The feed-forward contribution requires modeling of the nonlinear open-loop system behavior which has been experimentally identified and described through two-dimensional maps: the inputs are the duty cycle applied to the electrovalves and the pressure drop across their orifice, while the output is the pressure gradient in the brake caliper. These maps, obtained for inlet and outlet valves, are used to set the feed-forward term. Finally a proportional integral controller is designed to reject external disturbances and compensate for model uncertainties. A brake system test rig, described in part I, is used for building inverse maps and validating the proposed control logic. Different reference pressure profiles are used to experimentally verify the control tracking performances.
The development of a real-time driving simulator involves highly complex integrated and interdependent subsystems that require a large amount of computational time. When advanced hardware is unavailable for economic reasons, achieving real-time simulation is challenging, and thus delays are inevitable. Moreover, computational delays in the response of driving simulator subsystems reduce the fidelity of the simulation. In this paper, we propose a technique to decrease computational delays in a driving simulator. We used approximation techniques, sensitivity analysis, decomposition, and sampling techniques to develop a surrogate-based vehicle dynamic model (SBVDM). This global surrogate model can be used in place of the conventional vehicle dynamic model to reduce the computational burden while maintaining an acceptable accuracy. Our results showed that the surrogate model can significantly reduce computing costs compared to the computationally expensive conventional model. In addition, the response time of the SBVDM is nearly five times faster than the original simulation codes. Also, as a method to reduce hardware cost, the SBVDM was used and the results showed that most of the responses were accurate and acceptable in relation to longitudinal and lateral dynamics. Based on the results, the authors suggested that the proposed framework could be useful for developing low-cost vehicle simulation systems that require fast computational output.
The paper deals with the elaboration of an Active Roll Control (ARC) oriented both on comfort and handling improvement. The ARC determines hydraulically the variation of the equivalent stiffness of the anti-roll bars. Thecontrol strategies conceived were extensively validated through road tests managed on an Alfa Romeo sedan. The first part of the paper deals with comfort improvement, mainly consisting in an absence of bar effect during straight-ahead travel and in a modification of the roll characteristic of the car. To increase driver's handling feeling, it was necessary to optimise the ratio between front and rear roll stiffness. This purpose can be reached through control strategies based exclusively on lateral acceleration. Some control strategy corrections were necessary to optimise roll damping and front/rear roll stiffness balancing. A control logic, based on yaw rate arid vehicle sideslip angle estimation, was introduced, with the main purpose of optimise yaw damping during transient manoeuvres. Road tests extensively managed validated the efficiency of active roll control strategies conceived. Finally the paper presents some example of integrations with other chassis control systems, like Vehicle Dynamics Control (VDC). The most important results obtained, both in simulation and during road tests, are presented and discussed.
With contributions from an internationally-renowned group of experts, this book uses a multidisciplinary approach to review recent developments in the field of smart sensor systems, providing complete coverage of all important system and design aspects, their building blocks and methods of signal processing. It examines topics over the whole range of sensor technology from the theory and constraints of basic elements, the applied techniques and electronic, up to the level of application-orientated issues. Developed as a complementary volume to 'Smart Sensor Systems' (Wiley 2008), which introduces the theoretical foundations, this volume focuses on practical applications, including: State-of-the-art techniques for designing smart sensors and smart sensor systems, with measurement techniques at system level, such as collaboration and trimming, and impedance-measurement techniques. Sensing elements and sensor systems for the measurement of mechanical quantities, and microarrays for DNA detection. Circuitdesign for sensor systems, such as the design of low-noise amplifiers, and measurement techniques at device level, such as dynamic offset cancellation and optical imagers. Implantable smart sensors for bio-medical applications and automotive sensors. A supplementary website hosts case studies and a solutions manual to the problems Smart Sensor Systems: Emerging Technologies and Applications will greatly benefit final year undergraduate and postgraduate students in the areas of electrical, mechanical and chemical engineering, and physics. Professional engineers and researchers in the microelectronics industry, including microsystem developers, will also find this a thorough and useful volume.
Electric vehicles with four individually controlled drivetrains are over-actuated systems and therefore the total wheel torque and yaw moment demands can be realized through an infinite number of feasible wheel torque combinations. Hence, the energy-efficient torque distribution among the four drivetrains of such vehicle configuration is crucial for reducing the drivetrains power losses and extending driving range. In this paper, the reference wheel torque distribution is formulated as the solution of a parametric optimization problem, depending on vehicle speed. An analytical solution is provided for the case of equal drivetrains on the front and rear axles, under the experimentally confirmed hypothesis that the drivetrain power losses are monotonically increasing with the torque demand. The provided easily implementable and computationally fast wheel torque distribution algorithm is validated by simulations and experiments on an electric vehicle demonstrator, along driving cycles and cornering maneuvers. The results show considerable energy savings compared to alternative torque distribution strategies.
This paper presents a semi-active differential, called MRF LSD (MagnetoRheological Fluid Limited Slip Differential) that allows to bias torque between the driving wheels. It is based on the magnetorheological (MR) fluid employment, that allows to change, in a controlled manner, the differential locking torque and, consequently, the torque bias ratio. The device is an adaptive one and allows to obtain an asymmetric torque distribution in order to improve vehicle handling. The device modelling and the control algorithm, realized for this activity, are described. The illustrated results highlight the advantages that are attainable regarding directional behaviour, stability and traction for a front wheel drive (FWD) vehicle. A comparison with a traditional passive limited slip differential has been conducted.
The most powerful engine, the most sophisticated aerodynamic devices or the most complex control systems will not improve vehicle performances if the forces exchanged with the road are not optimized by proper employment and knowledge of tires. The vehicle interface with the ground is constituted by the sum of small surfaces, wide about as one of our palms, in which tire/road interaction forces are exchanged. From this it is clear to see how the optimization of tire behavior represents a key-factor in the definition of the best setup of the whole vehicle.Nowadays, people and companies playing a role in automotive sector are looking for the optimal solution to model and understand tire's behavior both in experimental and simulation environments. The studies carried out and the tool developed herein demonstrate a new approach in tire characterization and in vehicle simulation procedures. This enables the reproduction of the dynamic response of a tire through the use of specific track sessions, carried out with the aim to employ the vehicle as a moving lab.The final product, named TRICK tool (Tire/Road Interaction Characterization and Knowledge), comprises of a vehicle model which processes experimental signals acquired from vehicle CAN bus and from sideslip angle estimation additional instrumentation. The output of the tool is several extra "virtual telemetry" channels, based on the time history of the acquired signals and containing force and slip estimations, useful to provide tire interaction characteristics. TRICK results can be integrated with the physical models developed by the Vehicle Dynamics UniNa research group, providing a multitude of working solutions and constituting an ideal instrument for the prediction and the simulation of the real tire dynamics.
Principal component analysis (PCA) is a mainstay of modern data analysis - a
black box that is widely used but (sometimes) poorly understood. The goal of
this paper is to dispel the magic behind this black box. This manuscript
focuses on building a solid intuition for how and why principal component
analysis works. This manuscript crystallizes this knowledge by deriving from
simple intuitions, the mathematics behind PCA. This tutorial does not shy away
from explaining the ideas informally, nor does it shy away from the
mathematics. The hope is that by addressing both aspects, readers of all levels
will be able to gain a better understanding of PCA as well as the when, the how
and the why of applying this technique.
This paper deals with the frictional behaviour of a tyre tread elementary volume in sliding contact with road asperities. Friction is supposed as composed by two main components: adhesion and deforming hysteresis. The target, fixed in collaboration with a motorsport racing team and with a tyre manufacturing company, is to provide an estimation of local grip for on-line analyses and real time simulations and to evaluate and predict adhesive and hysteretic frictional contributions arising at the interface between tyre tread and road. A way to approximate asperities, based on rugosimetric analyses on macro and micro scale, has been introduced. The adhesive component of friction has been estimated by means of a new approach based on two different models found in literature, whose parameters have been identified thanks to a wide experimental campaign previously carried out. The hysteretic component of friction has been estimated by means of an energy balance taking into account rubber viscoelastic behaviour, characterized by means of proper DMA tests, and internal stress / strain distribution due to indentation with road. Model results are finally shown and discussed and the validation experimental procedure is described. The correct reproduction of friction phenomenology and the model prediction capabilities are highlighted making particular reference to grip variability due to changes in working conditions.
Tyre and transmission characteristics affect engine-to-slip dynamics which are essential when it comes to traction control design. This paper uses different tyre/transmission physical models to assess the modelling features which are essential for an accurate description of the plant. Three different single-wheel models are presented and analysed in order to find the simplest, yet reliable, model for the description of the engine-to-slip dynamics. The models are a single-wheel model with a rigid tyre and a rigid transmission, a single-wheel model with a flexible tyre, and a single-wheel model which features a compliant tyre and a sprocket absorber. Analytic expressions for all models' transfer functions are given, and models are compared to each other and with road tests to prove their reliability. Finally, the effect of longitudinal acceleration on engine-to-slip dynamics is analysed and discussed.
Cited By (since 1996):3, Export Date: 15 October 2013, Source: Scopus
A Numerical–physical tyre model was developed . The whole model allows to obtain the road–tyre interactions so it can be used in vehicle dynamic simulations. In this article are presented its capabilities in normal interaction analysis.
The normal interaction, i.e. the relationship between the normal load and the normal deflection, influences the tangential (longitudinal plus lateral) one, which determines the vehicle handling behaviour. The parameters used in this model depend on the structure of the tyre and they can be measured on the real tyre.
The tyre has been schematized as composed by a flexible belt , the sidewalls and a rigid ring (Rim). The flexible belt is composed by a number of lumped masses linked by extensional and bending stiffnesses and dampers. The tyre model has been developed using the finite segment method. Using these method could be possible to include in the tyre simulations various non-linear structural effects due to large displacements and rotations. The model allows to simulate both steady state and transient conditions.
This paper deals with in curve vehicle lateral behaviour and is aimed to find out which vehicle physical characteristics affect significantly its stability. Two different analytical methods, one numerical (phase plane) and the other graphical (handling diagram) are discussed. The numerical model refers to the complete quadricycle, while the graphical one refers to a bicycle model. Both models take into account lateral load transfers and nonlinear Pacejka tyre road interactions. The influence of centre of mass longitudinal position, tyre cornering stiffness and front/rear roll stiffness ratio on vehicle stability are analyzed. The presented results are in good agreement with theoretical expectations about the above parameters influence, and show how some physical characteristics behave as saddle node bifurcation parameters.
Adaptive Cruise Control systems have been developed and introduced into the consumer market for over a decade. Among these systems, fully-adaptive ones are required to adapt their behaviour not only to traffic conditions but also to drivers’ preferences and attitudes, as well as to the way such preferences change for the same driver in different driving sessions. This would ideally lead towards a system where an on-board electronic control unit can be asked by the driver to calibrate its own parameters while he/she manually drives for a few minutes (learning mode). After calibration, the control unit switches to the running mode where the learned driving style is applied. The learning mode can be activated by any driver of the car, for any driving session and each time he/she wishes to change the current driving behaviour of the cruise control system.The modelling framework which we propose to implement comprises four layers (sampler, profiler, tutor, performer). The sampler is responsible for human likeness and can be calibrated while in learning mode. Then, while in running mode, it works together with the other modelling layers to implement the logic. This paper presents the overall framework, with particular emphasis on the sampler and the profiler that are explained in full mathematical detail. Specification and calibration of the proposed framework are supported by the observed data, collected by means of an instrumented vehicle. The data are also used to assess the proposed framework, confirming human-like and fully-adaptive characteristics.
The early evolution of inerual navigation, then seen as an almost-impossible, barely-affordable technology, was driven by the requirements of strategic missile guidance. Since then, several advances in instruments and system techniques have brought it to everyday use in aircraft, ships and land vehicles, whilst still remaining a fascinating and challenging field of engineering. The paper reviews the principles of inertial navigation, and charts its progress from the early gimballed systems, to today's strapdown systems using ring laser gyros. Progress is illustrated by Marconi's inertial products, and forecasts of future direction are presented.
Vehicle Dynamics and Control provides a comprehensive coverage of vehicle control systems and the dynamic models used in the development of these control systems. The control system topics covered in the book include cruise control, adaptive cruise control, ABS, automated lane keeping, automated highway systems, yaw stability control, engine control, passive, active and semi-active suspensions, tire models and tire-road friction estimation. In developing the dynamic model for each application, an effort is made to both keep the model simple enough for control system design but at the same time rich enough to capture the essential features of the dynamics.
Principal component analysis (PCA) is a multivariate technique that analyzes a data table in which observations are described by several inter‐correlated quantitative dependent variables. Its goal is to extract the important information from the table, to represent it as a set of new orthogonal variables called principal components, and to display the pattern of similarity of the observations and of the variables as points in maps. The quality of the PCA model can be evaluated using cross‐validation techniques such as the bootstrap and the jackknife. PCA can be generalized as correspondence analysis (CA) in order to handle qualitative variables and as multiple factor analysis (MFA) in order to handle heterogeneous sets of variables. Mathematically, PCA depends upon the eigen‐decomposition of positive semi‐definite matrices and upon the singular value decomposition (SVD) of rectangular matrices. Copyright © 2010 John Wiley & Sons, Inc.
This article is categorized under: Statistical and Graphical Methods of Data Analysis > Multivariate Analysis
Statistical and Graphical Methods of Data Analysis > Dimension Reduction
Neural networks are making their ways into various commercial products across many industries. As in aerospace, in automotive
industry they are not the main technology. Automotive engineers and researchers are certainly familiar with the buzzword,
and some have even tried neural networks for their specific applications as models, virtual sensors, or controllers (see,
e.g., [1] for a collection of relevant papers). In fact, a quick search reveals scores of recent papers on automotive applications
of NN, fuzzy, evolutionary and other technologies of computational intelligence (CI); see, e.g., [2], [3], [4]. However, such
technologies are mostly at the stage of research and not in the mainstream of product development yet. One of the reasons
is “black-box” nature of neural networks. Other, perhaps more compelling reasons are business conservatism and existing/legacy
applications (trying something new costs money and might be too risky) [5], [6].
The vehicle sideslip angle is one of the most important variable to evaluate vehicle stability during dynamic manoeuvres. In this paper a nonlinear estimator is proposed, which use measurements of lateral accleration, steering angle, yaw rate and longitudinal velocity as input signals and provide the sideslip angle estimate as output. The design of such an estimator is based on a recently proposed direct approach to the design of virtual sensor. The obtained estimator has been experimentally tested on a huge number of different manoeuvres showing quite good results in a large range of operation covering both the linear and the nonlinear behaviour of the car.
This paper proposes an effective model-based approach to estimate vehicle linear sideslip online via recursive least square method (RLS) with forgetting. In this approach, a Luenberger observer is first designed to estimate vehicle states, including vehicle sideslip. Two lumped vehicle parameters in this observer are updated recursively to minimize the discrepancy between the model used and the physical plant and any possible effects caused by external unknown disturbances, in particular, road surface. Computer simulation and in-vehicle testing have been conducted to verify the proposed approach with results indicating that the proposed approach is very effective and robust in estimating vehicle linear sideslip under various road surfaces
Real-time knowledge of the slip angle in a vehicle is useful in many active vehicle safety applications, including yaw stability control, rollover prevention, and lane departure avoidance. Sensors to measure slip angle, including two-antenna GPS systems and optical sensors, are too expensive for ordinary automotive applications. This paper develops a real-time algorithm for estimation of slip angle using inexpensive sensors normally available for yaw stability control applications. The algorithm utilizes a combination of model-based estimation and kinematics-based estimation. Compared with previously published results on slip angle estimation, this present paper compensates for the presence of road bank angle and variations in tire-road characteristics. The developed algorithm is evaluated through experimental tests on a Volvo XC90 sport utility vehicle. Detailed experimental results show that the developed system can reliably estimate slip angle for a variety of test maneuvers.
The application of a Moving Horizon Estimator (MHE) to the problem of vehicle side-slip angle estimation is studied. In particular, this work focuses on how MHEs can be designed in order to provide estimates of the variable of interest with guaranteed estimation error. The problem is solved using a nonlinear regression observer identified directly from data.
Automotive engines are multivariable system with severe non-linear dynamics, and their modelling and control are challenging tasks for control engineers. Current control of engine used look-up table combined with proportional and integral (PI) control and is not robust to system uncertainty and time varying effects. In this paper the model predictive control strategy is applied to engine air/fuel ratio control using neural network model. The neural network model uses information from multivariables and considers engine dynamics to do multi-step ahead prediction. The model is adapted in on-line mode to cope with system uncertainty and time varying effects. Thus, the control performance is more accurate and robust compared with non-adaptive model based methods. To speed up algorithm calculation, different optimisation algorithms are investigated and performance compared. Finally, the developed method is evaluated on a well-known engine benchmark, a simulated mean value engine model (MVEM). The simulation results demonstrate the effectiveness of the developed method.
A model of a neural system where a group of neurons projects to another group of neurons is discussed. We assume that a trace is the simultaneous pattern of individual activities shown by a group of neurons. We assume synaptic interactions add linearly and that synaptic weights (quantitative measure of degree of coupling between two cells) can be coded in a simple but optimal way where changes in synaptic weight are proportional to the product of pre-and postsynaptic activity at a given time. Then it is shown that this simple system is capable of “memory” in the sense that it can (1) recognize a previously presented trace and (2) if two traces have been associated in the past (that is, if trace f̄ was impressed on the first group of neurons and trace ḡ was impressed on the second group of neurons and synaptic weights coupling the two groups changed according to the above rule) presentation of f̄ to the first group of neurons gives rise to f̄ plus a calculable amount of noise at the second set of neurons. This kind of memory is called an “interactive memory” since distinct stored traces interact in storage. It is shown that this model can effectively perform many functions. Quantitative expressions are derived for the average signal to noise ratio for recognition and one type of association. The selectivity of the system is discussed. References to physiological data are made where appropriate. A sketch of a model of mammalian cerebral cortex which generates an interactive memory is presented and briefly discussed. We identify a trace with the activity of groups of cortical pyramidal cells. Then it is argued that certain plausible assumptions about the properties of the synapses coupling groups of pyramidal cells lead to the generation of an interactive memory.
Machines degrade as a result of aging and wear, which decreases performance reliability and increases the potential for faults and failures. The impact of machine faults and failures on factory productivity is an important concern for manufacturing industries. Economic impacts relating to machine availability and reliability, as well as corrective (reactive) maintenance costs, have prompted facilities and factories to improve their maintenance techniques and operations to monitor machine degradation and detect faults. This paper presents an innovative methodology that can change maintenance practice from that of reacting to breakdowns, to one of preventing breakdowns, thereby reducing maintenance costs and improving productivity. To analyze the machine behavior quantitatively, a pattern discrimination model (PDM) based on a cerebellar model articulation controller (CMAC) neural network was developed. A stepping motor and a PUMA 560 robot were used to study the feasibility of the developed technique. Experimental results have shown that the developed technique can analyze machine degradation quantitatively. This methodology could help operators set up machines for a given criterion, determine whether the machine is running correctly, and predict problems before they occur. As a result, maintenance hours could be used more effectively and productively.
The Global Positioning System (GPS) is a satellite-based navigation and time transfer system developed by the U.S. Department of Defense. It serves marine, airborne, and terrestrial users, both military and civilian. Specifically, GPS includes the Standard Positioning Service (SPS) which provides civilian users with 100 meter accuracy, and it serves military users with the Precise Positioning Service (PPS) which provides 20-m accuracy. Both of these services are available worldwide with no requirement for a local reference station. In contrast, differential operation of GPS provides 2- to 10-m accuracy to users within 1000 km of a fixed GPS reference receiver. Finally, carrier phase comparisons can be used to provide centimeter accuracy to users within 10 km and potentially within 100 km of a reference receiver. This advanced tutorial will describe the GPS signals, the various measurements made by the GPS receivers, and estimate the achievable accuracies. It will not dwell on those aspects of GPS which are well known to those skilled in the radio communications art, such as spread-spectrum or code division multiple access. Rather, it will focus on topics which are more unique to radio navigation or GPS. These include code-carrier divergence, codeless tracking, carrier aiding, and narrow correlator spacing.
Software-in-the-loop development and experimental testing of a semi-active magnetorheological coupling for 4wd on demand vehicles
- R Russo
- M Terzo
- F Timpone
R. Russo, M. Terzo, and F. Timpone, "Software-in-the-loop development
and experimental testing of a semi-active magnetorheological coupling
for 4wd on demand vehicles," in Proc Mini Conf Veh Sys Dyn Identif
Anomalies, 2008, pp. 73-82.
Sensor for determining movements of vehicle relative to reference surface
- M Becker
- K Kirstaetter
M. Becker and K. Kirstaetter, "Sensor for determining movements of
vehicle relative to reference surface," Germany Patent DE4 444 223A1,
december 13, 1994. [Online]. Available: http://documents.allpatents.
com/l/22235209/DE4444223A1
Software-in-the-loop development and experimental testing of a semi-active magnetorheological coupling for 4wd on demand vehicles
- russo