[Show abstract][Hide abstract] ABSTRACT: Support vector machine (SVM) is a popular machine learning technique and its variant least squares support vector regression (LS-SVR) is effective for image denoising. However, conventional LS-SVR does not fully consider the sampling distribution of noisy images, which may degrade the performance of the algorithm. In this paper, we propose a new fuzzy density weight SVR (FDW-SVR) denoising algorithm, which assigns fuzzy priority to each sample according to its density weight. FDW is designed to estimate the joint probability density function via the fuzzy theory based on the pixel density and neighborhood density. Extensive experimental results show that FDW-SVR is superior to those state-of-the-art denoising techniques in light of both subjective and objective evaluations.
[Show abstract][Hide abstract] ABSTRACT: This paper mainly aims at the problem of adaptive quantized control for a class of uncertain nonlinear systems preceded by asymmetric actuator backlash. One challenging problem that blocks the construction of our control scheme is that the real control signal is wrapped in the coupling of quantization effect and nonsmooth backlash nonlinearity. To resolve this challenge, this paper presents a two-stage separation approach established on two new technical components, which are the approximate asymmetric backlash model and the nonlinear decomposition of quantizer, respectively. Then the real control is successfully separated from the coupling dynamics. Furthermore, by employing the neural networks and adaptation method in control design, a quantized controller is developed to guarantee the asymptotic convergence of tracking error to an adjustable region of zero and uniform ultimate boundedness of all closed-loop signals. Eventually, simulations are conducted to support our theoretical results.
Full-text · Article · Dec 2015 · IEEE transactions on neural networks and learning systems
[Show abstract][Hide abstract] ABSTRACT: This paper presents a novel fuzzy adaptive controller for controlling a class of dead-zone output nonlinear systems with time delays. A new approximate model is first designed to describe a special dead-zone phenomenon encountered by the output mechanism of nonlinear systems, and the proposed smooth model can be conveniently fused with available adaptive fuzzy control techniques. In addition, the coupling effect that the dead zone output and the time-delayed states coexist in a common coupling function makes the tracking control design more complicated. To further address this difficulty, a compensation method fusing mean-value theorem with Lyapunov-Krasovskii function is presented in this paper. By using the proposed output deadzone model, and based on Lyapunov synthesis, a new optimized algorithm is developed to guarantee the prescribed convergence of tracking error, and the boundedness of all the signals in the closed-loop systems. Simulations have been implemented to verify the performance of the proposed fuzzy adaptive controller.
No preview · Article · Dec 2015 · IEEE Transactions on Fuzzy Systems
[Show abstract][Hide abstract] ABSTRACT: Conventional machine learning methods such as neural network (NN) uses empirical risk minimization (ERM) based on infinite samples, which is disadvantageous to the gait learning control based on small sample sizes for biped robots walking in unstructured, uncertain and dynamic environments. Aiming at the stable walking control problem in the dynamic environments for biped robots, this paper puts forward a method of gait control based on support vector machines (SVM), which provides a solution for the learning control issue based on small sample sizes. The SVM is equipped with a mixed kernel function for the gait learning. Using ankle trajectory and hip trajectory as inputs, and the corresponding trunk trajectory as outputs, the SVM is trained based on small sample sizes to learn the dynamic kinematics relationships between the legs and the trunk of the biped robots. Robustness of the gait control is enhanced, which is propitious to realize the stable biped walking, and the proposed method shows superior performance when compared to SVM with radial basis function (RBF) kernels and polynomial kernels, respectively. Simulation results demonstrate the superiority of the proposed methods.
No preview · Article · Oct 2015 · Applied Soft Computing
[Show abstract][Hide abstract] ABSTRACT: This paper presents a saturated Nussbaum function based approach for robotic systems with unknown actuator dynamics. To eliminate the effect of the control shock from the traditional Nussbaum function, a new type of the saturated Nussbaum function is developed with the idea of time-elongation. Moreover, by exploiting properties of the proposed Nussbaum function, a promising theorem is established to deal with unknown multiple actuator nonlinearities. In what follows, the proposed theorem is integrated with the adaptive control technique such that the stability analysis of the robotic system is completed. It thus guarantees that the state of the robotic system asymptotically converges to the desired trajectory. Finally, comparative studies are carried out to validate the effectiveness and the superiority of the proposed approach.
No preview · Article · Sep 2015 · Cybernetics, IEEE Transactions on
[Show abstract][Hide abstract] ABSTRACT: To realize the isolation from the deleterious vibration source, it is essential to model the vibration's physical characteristic and then compensate for it in the control scheme. However, most current research studies of active isolator treat the vibration as bounded lumped blocks or linear fashions in the system analysis. Therefore, it unavoidably results in unsatisfying isolation in practical applications. In this paper, the modeling technique on multiple frequency vibrations is first developed to pave the way for the controller formulation. Subsequently, a novel adaptive compensation network is constructed, which aims to compensate for the vibration's physical property. In what follows, an adaptive neural controller is proposed for stabilizing the active isolation system. Guaranteed by the Lyapunov method, all signals in the closed-loop system are kept stable during the vibration suppression. Finally, the comparative results are presented to validate the proposed scheme's effectiveness.
No preview · Article · Aug 2015 · IEEE Transactions on Control Systems Technology
[Show abstract][Hide abstract] ABSTRACT: This paper presents an adaptive control approach to bridge the gap between unknown time-varying actuator nonlinearities and identical control directions in the area of robotic systems. Technically, a novel Nussbaum gain and its properties are first proposed to pave the way for the formulation of a more advanced tool. Subsequently, a newly developed theorem is summarized and integrated into the stability analysis such that all control units automatically and simultaneously estimate the unknown actuator nonlinearities and directions. Finally, it is rigorously proved that robotic systems asymptotically converge to the desired trajectories.
No preview · Article · Aug 2015 · Nonlinear Dynamics
[Show abstract][Hide abstract] ABSTRACT: Due to the fact that backlash nonlinearity is widespread in actuators, it is impossible to ignore its existence and achieve excellent and desirable mechanical system performance. In this paper, the problem of the humanoid robot grasping a common object with unknown actuator backlash is investigated. To tackle the nonsmooth backlash nonlinearity, a smooth adaptive backlash inverse is incorporated to compensate the line-segment effect. Moreover, a decentralized robust fuzzy adaptive control is constructed and developed to guarantee the object's motion and internal forces converge to the predefined values. The stabilities of the signals in the closed-loop system are proven by utilizing the Lyapunov method. In the end, experiments and simulations involving humanoid robot manipulation are conducted to validate the effectiveness of the proposed algorithms.
No preview · Article · Jun 2015 · IEEE Transactions on Fuzzy Systems
[Show abstract][Hide abstract] ABSTRACT: Hand tremors may cause some blemishes in precision and stability of a minimally invasive surgery (MIS). To track the tremor signals accurately, there are two main problems left to be settled. First, it is not practical to collect the sample data of tremor in large scale in practical applications. To deal with the hand tremors, a learning method based on small samples sizes and high dimensional input space is needed. Second, the hand tremors have time-varying characteristics. This fact is neglected by traditional learning methods, which could lead to imprecision and instability of a MIS. In this work, a time-sequence-based fuzzy support vector machine adaptive filter (TSF-SVMAF) for tremor cancelling is proposed. The proposed method is based on support vector machine and time series. It is suitable for solving the problem that the inputs are time-varying and the samples are small-scale. To cancel the time-varying hand tremors, different learning-weight-functions are designed for tremor signals with different frequencies. From the simulation results, compared with the existing methods such as back propagation (BP), weighted-frequency Fourier combiner (WFLC) and bandlimited multiple Fourier linear combiner (BMFLC), the proposed method has better performance when learning the time-varying hand tremors with small sample sizes.
No preview · Article · Apr 2015 · International Journal of Systems Science
[Show abstract][Hide abstract] ABSTRACT: This paper addresses the problem of adaptive neural output-feedback control for a class of special nonlinear systems with the hysteretic output mechanism and the unmeasured states. A modified Bouc-Wen model is first employed to capture the output hysteresis phenomenon in the design procedure. For its fusion with the neural networks and the Nussbaum-type function, two key lemmas are established using some extended properties of this model. To avoid the bad system performance caused by the output nonlinearity, a barrier Lyapunov function technique is introduced to guarantee the prescribed constraint of the tracking error. In addition, a robust filtering method is designed to cancel the restriction that all the system states require to be measured. Based on the Lyapunov synthesis, a new neural adaptive controller is constructed to guarantee the prescribed convergence of the tracking error and the semiglobal uniform ultimate boundedness of all the signals in the closed-loop system. Simulations are implemented to evaluate the performance of the proposed neural control algorithm in this paper.
No preview · Article · Apr 2015 · IEEE transactions on neural networks and learning systems
[Show abstract][Hide abstract] ABSTRACT: The control problem of the dual arm robot rigidly grasping an object with unknown nonsymmetric deadzone input is investigated in this paper. Due to the factor that the deadzone nonlinearity is widespread in the actuators, a smooth inverse adaptive deadzone is incorporated to minimize the effect of the deadzone nonlinearity in the dual arm robot system to guarantee a high accuracy tracking. Since these type of robots are usually applied in complex environments, a multi-input multi-output fuzzy logic unit is adopted to approximate the manipulator’s dynamics to achieve a accuracy tracking performances. Moreover, a decentralized robust fuzzy adaptive control scheme is constructed to make the motion and internal forces track a reference trajectories in the presence of parameters uncertainty and external disturbance. By using the Lyapunov method, the stability of the signals in the closed-loop system is proved. Simulation result demonstrates that the proposed controller is effective to the dual arm robot system with deadzone nonlinearity.
No preview · Article · Apr 2015 · Nonlinear Dynamics
[Show abstract][Hide abstract] ABSTRACT: This paper presents a novel adaptive controller for controlling an autonomous helicopter with unknown inertial matrix to asymptotically track the desired trajectory. To identify the unknown inertial matrix included in the attitude dynamic model, this paper proposes a new structural identifier that differs from those previously proposed in that it additionally contains a neural networks (NNs) mechanism and a robust adaptive mechanism, respectively. Using the NNs to compensate the unknown aerodynamic forces online and the robust adaptive mechanism to cancel the combination of the overlarge NNs compensation error and the external disturbances, the new robust neural identifier exhibits a better identification performance in the complex flight environment. Moreover, an optimized algorithm is included in the NNs mechanism to alleviate the burdensome online computation. By the strict Lyapunov argument, the asymptotic convergence of the inertial matrix identification error, position tracking error, and attitude tracking error to arbitrarily small neighborhood of the origin is proved. The simulation and implementation results are provided to evaluate the performance of the proposed controller.
No preview · Article · Mar 2015 · IEEE transactions on neural networks and learning systems
[Show abstract][Hide abstract] ABSTRACT: In this paper, a fuzzy adaptive approach for stochastic strict-feedback nonlinear systems with quantized input signal is developed. Compared with the existing research on quantized input problem, the existing works focus on quantized stabilization, while this paper considers the quantized tracking problem, which recovers stabilization as a special case. In addition, uncertain nonlinearity and the unknown stochastic disturbances are simultaneously considered in the quantized feedback control systems. By putting forward a new nonlinear decomposition of the quantized input, the relationship between the control signal and the quantized signal is established, as a result, the major technique difficulty arising from the piece-wise quantized input is overcome. Based on fuzzy logic systems' universal approximation capability, a novel fuzzy adaptive tracking controller is constructed via backstepping technique. The proposed controller guarantees that the tracking error converges to a neighborhood of the origin in the sense of probability and all the signals in the closed-loop system remain bounded in probability. Finally, an example illustrates the effectiveness of the proposed control approach.
No preview · Article · Mar 2015 · Cybernetics, IEEE Transactions on
[Show abstract][Hide abstract] ABSTRACT: This paper focuses on a problem of adaptive control for a class of nonlinear strict-feedback systems with a fuzzy dead zone and immeasurable states. By using the adaptive backstepping technique, an adaptive fuzzy output-feedback controller is constructed. The proposed control method requires only one adaptive law for an nth-order system. Compared with the conventional deterministic dead-zone models in previous articles, the main advantage of this paper is that the proposed dead-zone model is uncertain and fuzzy. By defuzzifying for fuzzy dead zone ~Γ(u) and employing an integrated design, an integrated fuzzy controller is constructed. It is proved that, even though the dead-zone input ~Γ(u) is fuzzy, the integrated fuzzy controller can make the closed-loop system semiglobally uniformly ultimately bounded and the tracking error converge to a small neighborhood of the origin. Finally, simulation results are provided to show the effectiveness of the proposed approach.
No preview · Article · Feb 2015 · IEEE Transactions on Fuzzy Systems
[Show abstract][Hide abstract] ABSTRACT: Support vector machines are the popular machine learning techniques. Its variant least squares support vector regression (LS-SVR) is effective for image denoising. However, the fitting of the samples contaminated by noises in the training phase will result in the fact that LS-SVR cannot work well when noise level is too far from it or noise density is high. Type-2 fuzzy sets and systems have been shown to be a more promising method to manifest the uncertainties. Various noises would be taken as uncertainties in scene images. By integrating the design of learning weights with type-2 fuzzy sets, a systematic design methodology of interval type-2 fuzzy density weighted support vector regression (IT2FDW-SVR) model for scene denoising is presented to address the problem of sample uncertainty in scene images. A novel strategy is used to design the learning weights, which is similar to the selection of human experience. To handle the uncertainty of sample density, interval type-2 fuzzy logic system (IT2FLS) is employed to deduce the fuzzy learning weights (IT2FDW) in the IT2FDW-SVR, which is an extension of the previously weighted SVR. Extensive experimental results demonstrate that the proposed method can achieve better performances in terms of both objective and subjective evaluations than those state-of-the-art denoising techniques.