No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Reconfigurable Flight Control Using Nonlinear Dynamic Inversion with a Special Accelerometer Implementation
Abstract
This paper presents an approach to on-line control design for aircraft that have suffered either actuator failure, missing effector surfaces, surface damage, or any combination. The approach is based on a modified version of nonlinear dynamic inversion. The approach does not require a model of the baseline vehicle (effectors at zero deflection), but does require feedback of accelerations and effector positions. Implementation issues are addressed and the method is demonstrated on an advanced tailless aircraft. An experimental simulation analysis tool is used to directly evaluate the nonlinear system's stability robustness.
... By solving Eqs. (13), (16), (19), and (21) ...
... The WPID solution within the actuator constraints is given by [19,20] ...
... In previous works, the weights are unchanged and are determined either by maximum actuator position or by human experience [19][20][21][22][23][24][25] . In Eq. (26), the optimal object is an approximation of energy consumption, especially that of ballonets given in Eq. (21) of section B. The optimal solution is limited by constant weights. ...
A stratospheric airship flies at a working altitude of 20km when it takes off from the ground. During ascent and descent, the wind field and thermal environment are highly complex. The thermal environment affects altitude, whereas wind influences the horizontal position of the airship. At a low altitude, this horizontal position cannot be controlled by thrusts given the low thrust-to-weight ratio, especially under a large wind field. However, it may be controlled indirectly by the pitch angle during ascent and descent with a certain vertical velocity. This study therefore proposes ascending and descending schemes for a stratospheric airship based on the thermal model. In this model, altitude is determined by the net lift/weight, whereas the horizontal position is controlled by the thrust and pitch. The pitch angle is determined by ballonets and an elevator. To allocate pitch control between the ballonets and the elevator under different airspeeds, pseudo-inverse dynamics of varied weight are introduced. In horizontal position control, the method of chain allocation is then applied between a pitch angle and vectored thrust to control the position of a stratospheric airship during ascent/descent.
... In contrast to regular backstepping, this method is inherently implicit in the sense that desired closed-loop dynamics do not reside in some explicit model to be cancelled, but which results when the feedback loops are closed. Theoretical development of increments of nonlinear control action date back from the late nineties and started with activities concerning 'Implicit Dynamic Inversion' for DI-based flight control [24, 4], where the architectures considered in this paper were firstly described. Other designations for these developments found in the literature are 'Modified NDI' and 'Simplified NDI', but the designation 'Incremental NDI' is considered to describe the methodology and nature of these type of control laws better [9, 10, 11, 21] . ...
... Other designations for these developments found in the literature are 'Modified NDI' and 'Simplified NDI', but the designation 'Incremental NDI' is considered to describe the methodology and nature of these type of control laws better [9, 10, 11, 21] . INDI has been elaborated and applied theoretically in the past decade for flight control and space applications [21, 25, 4, 5, 6, 1]. The main motivation of this approach is to bring the implicitness of such sensorbased architectures with Lyapunov-based controller design such as backstepping for aerospace applications. ...
... Its design departure is from a stability and convergence viewpoint due to control Lyapunov function augmentations rather than forcing linear behaviour through conventional feedback linearization. Because of its advantage of stabilizing or tracking one or more loops within a single control command maintaining desired properties , the motivation for this approach also stems to the combined flexibility of this method over conventional approaches such as robust nonlinear dynamic inversion (NDI) [2, 3, 7, 12, 13, 15, 22, 23, 26, 27], and its adaptive [8, 19, 20] and incremental counterparts [1, 4, 5, 6, 9, 21, 24, 25]. For the discussion, we will consider physical systems or vehicle dynamics which are represented by the following strict-feedback second order cascaded form: ...
... The present paper presents a method that is not conservative, but still incorporates all uncertainties, by feeding back angular accelerations. This concept is previously described by [5,6]. The control systems showed good performance when subjected to aerodynamic model mismatch. ...
... However, the robustness properties of angular acceleration feedback are not explicitly described by the authors. Also, [5,6] assume that the angular accelerations are readily available from measurements. ...
... In [6] the concept of feeding back angular accelerations is derived by first rewriting the rotational dynamic equations of motion into an incremental form and then applying regular NDI, resulting in incremental nonlinear dynamic inversion (INDI). It should be noted that INDI has been referred to by [5] as simplified dynamic inversion. ...
This paper presents a flight control strategy based on nonlinear dynamic inversion. The approach presented, called incremental nonlinear dynamic inversion, uses properties of general mechanical systems and nonlinear dynamic inversion by feeding back angular accelerations. Theoretically, feedback of angular accelerations eliminates sensitivity to model mismatch, greatly increasing the robust performance of the system compared with conventional nonlinear dynamic inversion. However, angular accelerations are not readily available. Furthermore, it is shown that angular acceleration feedback is sensitive to sensor measurement time delays. Therefore, a linear predictive filter is proposed that predicts the angular accelerations, solving the time delay and angular acceleration availability problem. The predictive filter uses only references and measurements of angular rates. Hence, the proposed control method makes incremental nonlinear dynamic inversion practically available using conventional inertial measurement units.
... Incremental controllers arise from a simplified representation of the model dynamics: the Incremental Dynamics (ID) [5], [6]. From the ID formulation, control strategies have been proposed, such as the incremental nonlinear dynamics inversion (INDI) [7], the incremental backstepping (IBKS) [8], and the incremental sliding mode control (INDI-SMC) [9]. ...
... However, the state derivative includes quantities like the angular acceleration, which is not a standard measurement for aircraft. A second-order differentiator (SOD) may be included to mitigate the absence of state derivative measurements [6]. The SOD corresponds to a bandpass second-order filter providing state derivative estimation and filtering high-frequency noise [13], [14]. ...
Incremental control strategies have been proposed recently to reduce the dependence on the aircraft model. However, the technique still requires modeling the actuator effectiveness, which depends on inertial and aerodynamic parameters. Therefore, it is of the utmost importance to analyze the robustness of the strategy regarding these parameters. This work focuses on evaluating the robustness of an incremental backstepping strategy applied for automatic control of a Boeing 747 (B747). A sensitivity analysis is carried out to quantify the impact of inertial parameters and flight condition in the input effectiveness, whose results are used to define scheduling strategies for the incremental controller. A robustness analysis is then performed using a B747 simulator applying different maneuvers in several flight conditions. The flight performance is evaluated according to the military standard MIL-DTL-9490E. Results show that the proposed controller is robust to mismatches in the input effectiveness model, evidencing that even without using a scheduling strategy the controller is able to perform within the requirements specified in the standard.
... Acceleration sensors can detect such disturbance forces and torques before they propagate to lower order states and therefore can act to reject those disturbances more rapidly when incorporated in feedback control. Incremental dynamic inversion has been shown in simulation in [19,20] to reject disturbance dynamics, given measurements of the acceleration states. Application of arrays of accelerometers in past works has focused on gyro-free angular velocity sensing and estimation, in so-called gyro-free inertial navigation systems. ...
... This control strategy is similar to the incremental dynamic inversion control suggested in [19,20]. As in these previous works, our proposed control needs only an estimate of the acceleration state of the vehiclex a , a sufficiently high-fidelity estimate of the control authority B, and the actuator command generated by the attitude controller u r . ...
Rapid sensing of body motions is critical to stabilizing a flight vehicle in the presence of exogenous disturbances as well as providing high-performance tracking of desired control commands. This bandwidth requirement becomes more stringent as vehicle scale decreases. Many flying insects employ distributed networks of acceleration-sensitive sensors to provide information about body egomotion to rapidly detect forces and torques. In this work, a method for rapid sensing of force and torque using a distributed array of accelerometers, arbitrarily placed and rigidly affixed to a vehicle airframe, was developed. Simulations of the sensor array were performed to quantify the effects of sensor noise, sensor position error, and sensor number on acceleration state estimates. A hardware implementation of this distributed sensor array was designed and integrated into the avionics of a small quadrotor vehicle. The response of the array to induced acceleration stimuli was characterized. A linear state estimation matrix was derived from the calibration to directly estimate the total forces and torques exerted on the airframe. A force-Adaptive control law utilizing the force and torque estimates provided by the sensor network was implemented to improve tracking of reference states while rejecting exogenous force and torque disturbances. Successful rejection of disturbances in the form of internal actuator variation and external wind gusts was demonstrated on the quadrotor vehicle in flight. © Copyright 2015 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved.
... The incremental form of NDI, Incremental NDI or INDI, is less model dependent and more robust. It has been described in the literature since the late nineties [4] [5], sometimes referred to as simplified [6] or enhanced [7] NDI. Compared to NDI, instead of modeling the angular acceleration based on the state and inverting the actuator model to get the control input, the angular acceleration is measured and an increment of the control input is calculated based on a desired increment in angular acceleration. ...
... Moreover, disturbances cannot be predicted. Initially, a setup with multiple accelerometers was proposed by Ostroff and Bacon [5] to measure the angular acceleration. This setup has some drawbacks, because it is complex and the accelerometers are sensitive to structural vibrations. ...
Incremental nonlinear dynamic inversion is a sensor-based control approach that promises to provide high-performance nonlinear control without requiring a detailed model of the controlled vehicle. In the context of attitude control of micro air vehicles, incremental nonlinear dynamic inversion only uses a control effectiveness model and uses estimates of the angular accelerations to replace the rest of the model. This paper provides solutions for two major challenges of incremental nonlinear dynamic inversion control: how to deal with measurement and actuator delays, and how to deal with a changing control effectiveness. The main contributions of this article are 1) a proposed method to correctly take into account the delays occurring when deriving angular accelerations from angular rate measurements; 2) the introduction of adaptive incremental nonlinear dynamic inversion, which can estimate the control effectiveness online, eliminating the need for manual parameter estimation or tuning; and 3) the incorporation of the momentum of the propellers in the controller. This controller is suitable for vehicles that experience a different control effectiveness across their flight envelope. Furthermore, this approach requires only very coarse knowledge of model parameters in advance. Real-world experiments show the high performance, disturbance rejection, and adaptiveness properties.
Read More: http://arc.aiaa.org/doi/abs/10.2514/1.G001490
... The effect of choice of control variable weighting is demonstrated in these figures.Figure 6a shows responses for a control variable weighting of Q = diag([roll acceleration error weighting, pitch acceleration error weighting, yaw acceleration error weighting]) = diag([1,5,2]). As can be seen, the stability-axis roll rateFigure 6a – inner-loop p c , q c , and r c angular rate commands, Q = diag([1,5,2]).Figure 6b – inner-loop p c , q c , and r c angular rate commands, Q = diag([1,5,10]). ...
... V t c commands, Q = diag([1,5,2]). reference signal is more closely followed than the stability-axis yaw rate reference signal with low dutch roll damping as illustrated by the oscillatory sideslip response.Figure 6b shows responses for Q = diag([1,5,10]). This results in the stability-axis yaw rate reference signal being more closely followed than infigure 6a and a better damped sideslip response.Figure 7 shows time responses for outer-loop γ c ...
Micro aerial vehicles have been the subject of continued interest and development over the last several years. The majority of current vehicle concepts rely on rigid fixed wings or rotors. An alternate design based on an aeroelastic membrane wing has also been developed that exhibits desired characteristics in flight test demonstrations, competition, and in prior aerodynamics studies. This paper presents a simulation model and an assessment of flight control characteristics of the vehicle. Linear state space models of the vehicle associated with typical trimmed level flight conditions and which are suitable for control system design are presented as well. The simulation is used as the basis for the design of a measurement based nonlinear dynamic inversion control system and outer loop guidance system. The vehicle/controller system is the subject of ongoing investigations of autonomous and collaborative control schemes. The results indicate that the design represents a good basis for further development of the micro aerial vehicle for autonomous and collaborative controls research.
... To improve the control performance of aircraft with nonlinear models, many nonlinear control methods have been studied; backstepping and feedback linearization are two common methods [10]. Liu and Sang proposed both backstepping and sliding mode backstepping control to realize trajectory doi: 10.1016/j.neucom.2021.08.069 3 tracking control of stratospheric airships [11][12][13][14][15]. Bacon and Ostroff expanded dynamic inversion into an incremental form by introducing an angular acceleration feedback [16]. Incremental nonlinear dynamic inversion (INDI), a method in which the dynamics are written in an incremental form, is a sensor-based control method that does not rely on accurate aircraft model, and it can realize aircraft flight under certain uncertainties and structural faults [17]. ...
This paper presents an adaptive incremental nonlinear backstepping sliding-mode (INBSM) controller, for fault tolerant tracking control of a blended wing body (BWB) aircraft with unknown disturbances and actuator faults. The INBSM controller is based on a nonlinear dynamics model of the BWB aircraft. In addition, a radial basis function neural network disturbance observer (RBF-NNDO) is proposed to enhance the disturbance attenuation ability. A fault estimator is suggested to improve actuator fault tolerant control level. The closed-loop control system of the BWB aircraft is proved to be globally asymptotically stable using Lyapunov theory. Simulations of the combined NNDO-INBSM controller are presented and compared with both the INBSM design and an adaptive fuzzy controller. The results demonstrate an improved capability of the NNDO-INBSM control for the BWB aircraft to execute realistic attitude tracking missions, even in the presence of center of gravity movement, unknown disturbances, model uncertainties and actuator faults.
... is the available measurement (T s is the sampling time of the controller). Usually, angular acceleration measurements are used because of the model uncertainties in the rotational dynamics [22], but also other measurements can be used [12]. In order to reduce the necessary model knowledge, it is assumed that the term ?f ?x ...
In this paper, the design of an Incremental Backstepping controller for fault tolerant control of a hexacopter system is presented. The proposed controller compensates for the modeling mismatches including faults by relying on sensor measurements. Outdoor flight test results show the better performance of the Incremental Backstepping controller compared to the Backstepping controller. Furthermore, robustness against faults is demonstrated for the case of unknown control degradation within the propulsion system.
... Their simulation results showed an absence of oscillations in the transient response of the roll command. Waszak et al. [9] used a modified nonlinear dynamic inversion technique [10] that does not require the plant model. They developed a controller for the MAV in which the control input is less than the control variables. ...
This paper presents the implementation of a modified state observer-based adaptive dynamic inverse controller for the Black Kite micro aerial vehicle. The pitch and velocity adaptations are computed by the modified state observer in the presence of turbulence to simulate atmospheric conditions. This state observer uses the estimation error to generate the adaptations and, hence, is more robust than model reference adaptive controllers which use modeling or tracking error. In prior work, a traditional proportional-integral-derivative control law was tested in simulation for its adaptive capability in the longitudinal dynamics of the Black Kite micro aerial vehicle. This controller tracks the altitude and velocity commands during normal conditions, but fails in the presence of both parameter uncertainties and system failures. The modified state observer-based adaptations, along with the proportional-integral-derivative controller enables tracking despite these conditions. To simulate flight of the micro aerial vehicle with turbulence, a Dryden turbulence model is included. The turbulence levels used are based on the absolute load factor experienced by the aircraft. The length scale was set to 2.0 meters with a turbulence intensity of 5.0 m/s that generates a moderate turbulence. Simulation results for various flight conditions show that the modified state observer-based adaptations were able to adapt to the uncertainties and the controller tracks the commanded altitude and velocity. The summary of results for all of the simulated test cases and the response plots of various states for typical flight cases are presented.
The robustness of the incremental nonlinear dynamic inversion (INDI) technique depends on the accuracy of the feedback angular accelerations. However, angular accelerations are usually not readily available. They are obtained from differentiation and are sensitive to time delays and noise. These undesired effects decrease the robustness of the system to disturbances and model uncertainties. A novel INDI-based flight control strategy, named angular acceleration estimation-based INDI (EINDI), is proposed in this paper to solve the problem of acquiring accurate angular accelerations. The EINDI method combines the control surface deflections and an adaptive technique to estimate angular accelerations, which can reduce the effects of noise and time delays on angular accelerations, thereby ensuring the robustness of the system. Furthermore, stability analysis based on the Lyapunov theory demonstrated that the EINDI was robust to model uncertainties. Simulation results showed that EINDI was effective in reducing the influence of noise and time delays and overcoming the influence of the center of gravity (CG) changes and a series of model uncertainties.
Recently, the concept of incremental nonlinear dynamic inversion has seen an increasing adoption as an attitude control method for a variety of aircraft configurations. The reasons for this are good stability and robustness properties, moderate computation requirements and low requirements on modelling fidelity. While previous work investigated the robust stability properties of incremental nonlinear dynamic inversion, the actual closed-loop performance may degrade severely in the face of model uncertainty. We address this issue by first analysing the effects of modelling errors on the closed-loop performance by observing the movement of the system poles. Based on this, we analyse the neccessary modelling fidelity and propose simple modelling methods for the usual actuators found on small-scale electric aircraft. Finally, we analyse the actuator models using (flight) test data where possible.
Future launch vehicle concepts and technologies for expendable and reusable launch vehicles are currently investigated by the DLR research projects Akira and X-tras. In particular, the winged Liquid Fly-back Booster concept Lfbb based on an LOX/LH2 propellant combination for vertical takeoff and vertical landing (VTVL), as well as the delta-winged horizontal takeoff and horizontal landing (HTHL) concept Aurora based on an LOX/Kerosene propellant combination are considered in these projects. Because of the complexity and risks involved in on-line trajectory optimization, off-line reference trajectories are still considered important for tracking purposes. In that sense, the goal of this paper is to investigate an off-line and general-purpose guidance and control (G&C) architecture for preliminary studies of reusable launch vehicles. This is done by using trajectory optimization combined with Modelica models for the generation of optimal guidance commands, and then trajectory tracking is performed by means of inner-loop feedback controls in terms of nonlinear dynamic inversion with prescribed desired dynamics. We showcase the advantages of this baseline G&C architecture in terms of early stability and controllability aspects during the preliminary design studies of an example configuration of a reusable launch vehicle investigated in the context of the research projects above mentioned.
Previous results reported in the robotics literature show the relationship between $\textit{time-delay control}$ (TDC) and $\textit{proportional-integral-derivative control}$ (PID). In this paper, we show that $\textit{incremental nonlinear dynamic inversion}$ (INDI) $-$ more familiar in the aerospace community $-$ are in fact equivalent to TDC. This leads to a meaningful and systematic method for PI(D)-control tuning of robust nonlinear flight control systems via INDI. We considered a reformulation of the plant dynamics inversion which removes effector blending models from the resulting control law, resulting in robust model-free control laws like PI(D)-control.
In general, an airship is equipped with hybrid-heterogeneous actuators, aerodynamic control surfaces, the vectored thrusts, and buoyant ballonets; however, moving-mass control is still introduced to a stratospheric airship for its special working condition of low atmospheric density and low airspeed. Thus, the composite control of hybrid. heterogeneous actuators is the primary object in controller design for stratospheric airship. The dynamic equation of airship is derived using the Newton Euler method, and the mechanism of moving-mass control is investigated. The control capabilities between the moving mass and the aerodynamic control surface are compared. The weighted generalized inverse (WGI) is used to design the nonlinear composite controller, where the authority of every actuator can be decided by setting the corresponding value of the control efficiency weighted matrix; thus, the control law is unchanged under different actuator configurations. Using the stratospheric airship as an example, the aerodynamic control surfaces, moving mass, and vectored thrust are combined into the composite control system, and the simulation of altitude tracking is provided. The results of the analysis show that the movements of the moving mass change the pitch and roll angles of the airship such that the passive gliding movement can be realized in the same manner as a glider in water. The moving-mass ratio and the displacement of the moving masses are the primary factors affecting the moving-mass attitude control. The moving-mass attitude motion is unaffected by airspeed; hence, it has a strong adaptability. The WGI achieves a good distribution and reconfiguration among these heterogeneous actuators and maintains the minimum control energy, thereby enhancing the reliability of the control system.
A control allocation based reconfigurable flight controller is presented for aircraft with multiple control effectors because of the redundancy effectors feature. The main idea is to use the remaining control effectors to cancel or compensate the effects of the failure effectors. The control law was designed by trajectory linearization control method, which can provides robust performance at all stages of flight. And the desired moments are the outputs of the trajectory linearization controller. A novel optimal control allocation method, named bases sequence optimal control allocation (BSOCA) method, is proposed to distribute the deflections of control effectors of the aircraft to generate desired moments. A control effectors management system based on BSOCA method is presented to achieve flight reconfiguration under effector failures. The simulation results show that the control allocation based reconfigurable flight controller can adapt the effector failures and maintain stability and acceptable handling qualities. Copyright © 2009 by the American Institute of Aeronautics and Astronautics, Inc.
1. Abstract The objective of this paper is ,to report the results from the research being conducted in reconfigurable flight controls at NASA Ames. A studywas,conducted with three NASA Dryden test pilots to evaluate two, approaches of reconfiguring an aircraft’s control system when ,failures occur in ,the control surfaces and engine. NASA Ames is investigating both a Neural Generalized Predictive Control scheme and ,a Neural Network based Dynamic Inverse controller. This paper highlights the Predictive Control scheme where a simple augmentation to reduce zerostead y-state error ledto the neural network predictor modelbecom,ingre dundant for the task. Instead of using a neural network predictor model, a nominal single point linear model was usedan d thena ugmented withan er rorco rrector. T his paper shows ,that the Generalized Predictive Controller and the Dynamic Inverse Neural Network controller perform equally well at reconfiguration, but with less rate requirements from the actuators. Also presented areth e pilot ratings for each controller for various failure scenarios and ,two samples of ,the required control actuation during reconfiguration. Finally, the paper concludes by, stepping through the ,Generalized ,Predictive Control’s reconfiguration process for an elevator failure.
This survey of model-based fault diagnosis focuses on those methods that are appli-cable to aerospace systems. To highlight the characteristics of aerospace models, generic non-linear dynamical modelling from flight mechanics is recalled and a unifying representation of sensor and actuator faults is presented. An extensive bibliographical review supports a descrip-tion of the key points of fault detection methods that rely on analytical redundancy. The approaches that best suit the constraints of the field are emphasized and recommendations for future developments in in-flight fault diagnosis are provided.
In order to solve the robustness problem of regular explicit nonlinear dynamic inversion (NDI) control law, the incremental version NDI control law is designed via the implicit dynamic inversion (DI) method and singular perturbation theory, in which the state acceleration (derivative) is feedback to measure the effects of the aircraft model perturbation and the current position information of the effector is needed. The whole aerodynamic model is unnecessary because there is no nonlinear direct state feedback base on the nominal aircraft model, and as a result, the sensitivity to aircraft model is decreased. The control structure is simple, re-allocation of control power and the reconfiguration of control law can be achieved conveniently and rapidly. The control law design and computer simulation for an aircraft with aerodynamic perturbation demonstrate that the robustness of the control law is increased.
Autonomy in accomplishing a planned flight mission is a crucial requirement for unmanned aerial vehicles. To achieve this goal, the flight control system of the aircraft must feature a high level of robustness against model and parameter uncertainties as well as significant adaptability in order to be able to cope with disturbances and failure conditions. Furthermore, adherence to the operational envelope of the aircraft must be ensured. An adaptive control concept, based on nonlinear dynamic inversion is presented, that meets the requirements stated above and allows tracking complex, three-dimensional trajectories at a high bandwidth and with high accuracy. The system allows the full exploitation of the physical capabilities of the airframe and accounts for saturation effects in control deflections and rates. The assessment of the system is performed on the basis of nonlinear simulations utilizing a complex simulation model.
A generic control system framework for both real-time and batch six-degree-of-freedom simulations is presented. This framework uses a simplified dynamic inversion technique to allow for stabilization and control of any type of aircraft at the pilot interface level. The simulation, designed primarily for the real-time simulation environment, also can be run in a batch mode through a simple guidance interface. Direct vehicle-state acceleration feedback is required with the simplified dynamic inversion technique. The estimation of surface effectiveness within real-time simulation timing constraints also is required. The generic framework provides easily modifiable control variables, allowing flexibility in the variables that the pilot commands. A direct control allocation scheme is used to command aircraft effectors. Primary uses for this system include conceptual and preliminary design of aircraft, when vehicle models are rapidly changing and knowledge of vehicle six-degree-of-freedom performance is required. A simulated airbreathing hypersonic vehicle and simulated high-performance fighter aircraft are used to demonstrate the flexibility and utility of the control system.
In this paper, we describe development and analysis of an active
fault detection and isolation system for the commuter and business
aircraft. To accommodate faults wherever possible, we develop an
algorithm that can reliably detect and isolate system faults with
minimal disruption of normal aircraft operations. We define several
candidate fault scenarios (e.g., aircraft icing, faults of control
surface actuators, stuck or floating control surfaces, etc.) that are
common occurrences, and construct a jet aircraft model, suitable for
simulation studies. Aircraft faults are detected and isolated using a
hierarchy of techniques. Successive layers in the hierarchy are
increasingly invasive, higher layers being invoked only when lower
layers indicate a potential problem
During the summer of 1996 a series ofight tests demonstrated a new indirect-adaptive approach to recon- ®gurableight control known as the self-designing con- troller (SDC). The SDC achieves improved, appropriately decoupled responses during arbitrary effector or airframe impairment scenarios, and successful SDCight tests cul- minated with smooth landing of the VISTA/F-16 in cross- wind conditions with a (simulated) missing primary con- trol surface (left horizontal tail). The SDC couples model- following receding-horizon optimal control with an on-line parameter identi®cation (ID) algorithm designed to provide smooth, accurateestimatesofpossiblytime-varyingsystem parameters, even under conditions of low excitation. The adaptive model-following approach is designed to reduce control law development costs and improve system perfor- mance in the presence of gradual or abrupt changes, includ- ing unforeseen events. This paper provides (1) a brief sum- maryoftheSDCalgorithms,(2)adiscussionofSDCimple- mentation on the VISTA/F-16 ¯ight control hardware, (3) a summary ofight test results, and (4) suggestions for fur- ther research in recon®gurable/adaptive controls.
The theoretical development of a direct adaptive tracking control architecture using neural networks is presented. Emphasis is placed on utilization of neural networks in a flight control architecture based on feedback linearization of the aircraft dynamics, Neural networks are used to represent the nonlinear inverse transformation needed for feedback linearization. Neural networks may be first trained offline using a nominal mathematical model, which provides an approximate inversion that can accommodate the total flight envelope, Neural networks capable of on-line learning are required to compensate for inversion error, which may arise from imperfect modeling, approximate inversion, or sudden changes in aircraft dynamics. A stable weights adjustment rule for the on-line neural network is derived. Under mild assumptions on the nonlinearities representing the inversion error, the adaptation algorithm ensures that all of the signals in the hoop are uniformly bounded and that the weights of the on-line neural network tend to constant values. Simulation results far an F-18 aircraft model are presented to illustrate the performance of the on-line neural network based adaptation algorithm.
Traditional autopilot design for guided munitions requires an accurate aerodynamic model and relies on a gain schedule to account for system nonlinearities. This paper presents an approach that simplifies the autopilot design process by combining an Inverting controller designed at a single flight condition with an on-line neural network to account for errors that arise because of the approximate inversion. This eliminates the need for an extensive design process and also the requirement for accurate aerodynamic data, which can be especially critical at high angles of attack or in other regimes at which the aerodynamics become highly nonlinear. The choice of the inversion process itself has been found to be critical in the implementation and is therefore discussed at length. Finally, results from an application of this approach to a full nonlinear six-degree-of-freedom guided munition simulation are presented.
This report describes the Phase 2 portion of a joint U.S. Air Force U.S. Navy sponsored investigation of innovative aerodynamic control concepts for fighter aircraft without vertical tails. The Phase 1 work consisted of effector performance prediction, vehicle integration assessment, and transition studies for numerous control effector concepts. The Phase 2 effort is an extension of the Phase 1 work and consists mainly of three wind tunnel tests to acquire more information about the aerodynamic performance of all moving wing tips (AMTs) and spoiler slot deflectors (SSDs). These tests clearly showed both the AMT and SSD can produce significant yaw control power and are strong control effector candidates for tailless fighters. Predicted vehicle dynamics from Phase 1 are compared with predictions developed utilizing the new aerodynamic data acquired in Phase 2.
An indirect adaptive control system approach is presented and demonstrated via the nonlinear six degree of freedom simulation of a tailless fighter aircraft. A summary of the tailless aircraft configuration is given. The modular control system architecture is described and interactions between the modules are identified and illustrated using nonlinear simulation results. Mitigation approaches for addressing the adverse interactions are presented and illustrated using nonlinear simulation results. Simulation results that demonstrate the reconfiguration capability are also given. Copyright © 1999 John Wiley & Sons, Ltd.
A new control design methodology is developed: Stochastic Optimal Feedforward and Feedback Technology (SOFFT). Traditional design techniques optimize a single cost function (which expresses the design objectives) to obtain both the feedforward and feedback control laws. This approach places conflicting demands on the control law such as fast tracking versus noise atttenuation/disturbance rejection. In the SOFFT approach, two cost functions are defined. The feedforward control law is designed to optimize one cost function, the feedback optimizes the other. By separating the design objectives and decoupling the feedforward and feedback design processes, both objectives can be achieved fully. A new measure of command tracking performance, Z-plots, is also developed. By analyzing these plots at off-nominal conditions, the sensitivity or robustness of the system in tracking commands can be predicted. Z-plots provide an important tool for designing robust control systems. The Variable-Gain SOFFT methodology was used to design a flight control system for the F/A-18 aircraft. It is shown that SOFFT can be used to expand the operating regime and provide greater performance (flying/handling qualities) throughout the extended flight regime. This work was performed under the NASA SBIR program. ICS plans to market the software developed as a new module in its commercial CACSD software package: ACET.
This paper describes a neural network based direct adaptive control approach to the problem of reconfigurable flight control. A Tailless fighter aircraft configuration with multiple and redundant control actuation devices is used to illustrate the level to which handling qualities can be maintained in the presence of large scale failures in the actuation channels. Of significance here is the speed with which recovery and maintenance of handling qualities can take place. The main advantage lies in eliminating the need for parameter identification during the recovery phase, and limiting the potential need for parameter identification to the problem of control allocation following the failure. A second by-product of this work is that the need for an accurate aerodynamic data base for purposes of flight control design can be significantly reduced. Moreover, the need for extensive off-line analysis, in-flight tuning and validation of gain schedules, and contingency coding necessary to handl...
This paper describes a general form of nonlinear dynamic inversion control for use in a generic nonlinear simulation to evaluate candidate augmented aircraft dynamics. The implementation is specifically tailored to the task of quickly assessing an aircraft's control power requirements and defining the achievable dynamic set. The achievable set is evaluated while undergoing complex mission maneuvers, and perfect tracking will be accomplished when the desired dynamics are achievable. Variables are extracted directly from the simulation model each iteration, so robustness is not an issue. Included in this paper is a description of the implementation of the forces and moments from simulation variables, the calculation of control effectiveness coefficients, methods for implementing different types of aerodynamic and thrust vectoring controls, adjustments for control effector failures, and the allocation approach used. A few examples illustrate the perfect tracking results obtained. Introdu...
This paper contains a study of two methods for use in a generic nonlinear simulation tool that could be used to determine achievable control dynamics and control power requirements while performing perfect tracking maneuvers over the entire flight envelope. The two methods are NDI (nonlinear dynamic inversion) and the SOFFT (Stochastic Optimal Feedforward and Feedback Technology) feedforward control structure. Equivalent discrete and continuous SOFFT feedforward controllers have been developed. These equivalent forms clearly show that the closed-loop plant model loop is a plant inversion and is the same as the NDI formulation. The main difference is that the NDI formulation has a closed-loop controller structure whereas SOFFT uses an open-loop command model. Continuous, discrete, and hybrid controller structures have been developed and integrated into the formulation. Linear simulation results show that seven different configurations all give essentially the same response, with the NDI...
A method for real-time estimation of parameters in a linear dynamic state space model was developed and studied. The application is aircraft dynamic model parameter estimation from measured data in flight for indirect adaptive or reconfigurable control. Equation error in the frequency domain was used with a recursive Fourier transform for the real-time data analysis. Linear and nonlinear simulation examples and flight test data from the F-18 High Alpha Research Vehicle (HARV) were used to demonstrate that the technique produces accurate model parameter estimates with appropriate error bounds. Parameter estimates converged in less than 1 cycle of the dominant dynamic mode natural frequencies, using control surface inputs measured in flight during ordinary piloted maneuvers. The real-time parameter estimation method has low computational requirements, and could be implemented aboard an aircraft in real time. Nomenclature A,B,C,D system matrices EYd expectation operator g acceleration ...