Variational Bayesian based adaptive PDA filter in scenarios with unknown detection probability and heavy-tailed process noise

For target tracking systems, the probability of detecting a target is difficult to determine, and the process noise often has non-Gaussian heavy-tailed characteristics owing to interference from outliers. To address the issues associated with single target tracking within clutters in scenarios with an unknown detection probability and heavy-tailed process noise, this paper presents a variational Bayesian-based adaptive probabilistic data association filter (VB-APDAF). The beta distribution, Pearson type VII distribution and multinomial distribution are used to model the detection probability, the process noise, and the association events, respectively. To guarantee the conjugation, a novel parameter estimation strategy is employed. In this strategy, the previous state is introduced in the state update process to construct the joint probability density function of parameters to be estimated and data set. The VB framework is used to estimate the target state, detection probability, and associated events. An experiment was performed under simulated conditions to demonstrate the effectiveness of the proposed filter.     

Bias-compensated affine-projection-like algorithm based on maximum correntropy criterion for robust filtering

This article proposes an affine-projection-like maximum correntropy (APLMC) algorithm for robust adaptive filtering. The proposed APLMC algorithm is derived by using the objective function based on the maximum correntropy criterion (MCC), which can availably suppress the bad effects of impulsive noise on filter weight updates. But the overall performance of the APLMC algorithm may be decreased when the input signal is polluted by noise. To compensate for the deviation of the APLMC algorithm in the input noise interference environment, the bias compensation (BC) method is introduced. Therefore, the bias-compensated APLMC (BC-APLMC) algorithm is presented. Besides, the convergence of the BC-APLMC algorithm in the mean and the mean square sense is studied, which provides a constraint range for the step-size. Computer simulation results show that the APLMC, and BC-APLMC algorithms are valid in acoustic echo cancellation and system identification applications. It also shows that the proposed algorithms are robust in the presence of input noise and impulse noise.     

Robust tracking control for quadrotor with unknown nonlinear dynamics using adaptive neural network based fractional-order backstepping control

This work aims to design a neural network-based fractional-order backstepping controller (NNFOBC) to control a multiple-input multiple-output (MIMO) quadrotor unmanned aerial vehicle (QUAV) system under uncertainties and disturbances and unknown dynamics. First, we investigated the dynamic of QUAV composed of six inter-connected nonlinear subsystems. Then, to increase the convergence speed and control precision of the classical backstepping controller (BC), we design a fractional-order BC (FOBC) that provides further degrees of freedom in the control parameters for every subsystem. Besides, designing control is a challenge as the FOBC requires knowledge of accurate mathematical model and the physical parameters of QUAV system. To address this problem, we propose an adaptive approximator that is a radial basis function neural network (RBFNN) included in FOBC to fix the unknown dynamics problem which results in the new approach NNFOBC. Furthermore, a robust control term is introduced to increase the tracking performance of a reference signal as parametric uncertainties and disturbances occur. We have used Lyapunov's theorem to derive adaptive laws of control parameters. Finally, the outcoming results confirm that the performance of the proposed NNFOBC controller outperforms both the classical BC , FOBC and a neural network-based classical BC controller (NNBC) and under parametric uncertainties and disturbances.     

Robust adaptive visual tracking control for uncertain robotic systems with unknown dead-zone inputs

This paper is concerned with the image-based visual servoing (IBVS) control for uncalibrated camera-robot system with unknown dead-zone constraint, where the uncertain kinematics and dynamics are also considered. The control implementation is achieved by constructing a smooth inverse model for dead-zone-input to eliminate the nonlinear effect resulting from the actuator constraint. A novel adaptive algorithm, which does not require a priori knowledge of the parameter intervals of dead-zone model, is proposed to update the parameter values online, and the dead-zone slopes are not required the same. Furthermore, to accommodate the uncertainties of uncalibrated camera-robot system, adaptation laws are developed to estimate the uncertain parameters, simultaneously avoiding singularity of the image Jacobian matrix. With the full consideration of unknown dead-zone constraint and system uncertainties, an adaptive robust visual tracking control scheme together with dead-zone compensation is subsequently established such that the image tracking error converges to the origin. Based on a 3-DOF manipulator, simulations are conducted to verify the tracking performance of the proposed controller.     

Three-stage Kalman filter for state and fault estimation of linear stochastic systems with unknown inputs

The paper studies the problem of simultaneously estimating the state and the fault of linear stochastic discrete-time varying systems with unknown inputs. The fault and the unknown inputs affect both the state and the output. However, if the dynamical evolution models of the fault and the unknown inputs are available the filtering problem will be solved by the Optimal three-stage Kalman Filter (OThSKF). The OThSKF is obtained after decoupling the covariance matrices of the Augmented state Kalman Filter (ASKF) using a three-stage U–V transformation. Nevertheless, if the fault and the unknown inputs models are not perfectly known the Robust three-stage Kalman Filter (RThSKF) will be applied to give an unbiased minimum-variance estimation. Finally, a numerical example is given in order to illustrate the proposed filters.     

基于粒子滤波的无线传感器网络目标跟踪

本文首先介绍了PF(Particle Filter)和UPF(Unscented Particle Filter)的基本原理,然后针对无线传感器网络(WSN)目标跟踪这一应用方向,采用等级网络结构,参考分布式粒子滤波算法,将UPF应用于WSN单目标跟踪以提高网络跟踪精度,仿真证明UPF较PF在跟踪精度上确实有明显的提高。     

Finite-time command filter-based adaptive fuzzy tracking control for stochastic nonlinear induction motors systems with unknown backlash-like hysteresis

In this article, an adaptive fuzzy control method is proposed for induction motors (IMs) drive systems with unknown backlash-like hysteresis. First, the stochastic nonlinear functions existed in the IMs drive systems are resolved by invoking fuzzy logic systems. Then, a finite-time command filter technique is exploited to overcome the obstacle of “explosion of complexity” emerged in the classical backstepping procedure during the controller design process. Meanwhile, the effect of the filter errors generated by command filters is decreased by utilizing corresponding error compensating mechanism. To cope with the influence of backlash-like hysteresis input, an auxiliary system is constructed, in which the output signal is applied to compensate the effect of the hysteresis. The finite-time control technology is adopted to accelerate the response speed of the system and reduce the tracking error, and the stochastic disturbance and backlash-like hysteresis are considered to improve control accuracy. It’s shown that the tracking error can converge to a small neighborhood around the origin in finite-time under the constructed controller. Finally, the availability of the presented approach is validated through simulation results.     

Finite-time adaptive neural prescribed tracking control of stochastic nonlinear systems with multiple power terms and unknown time-varying powers

This research addresses the problem of finite-time tracking error constrained control for a class of non-strict stochastic nonlinear systems with unknown time-varying powers and multiple power terms. Based on the conversion from constrained tracking error to an unconstrained signal with the same effect, by adopting the backstepping technique together with adaptive neural network control, a controller with upper and lower time-varying power bounds is designed to meet the prescribed performance control scheme in finite-time. Finally, two simulation examples are shown to verify the effectiveness of the commendatory control method.     

Adaptive optimal tracking controls of unknown multi-input systems based on nonzero-sum game theory

This paper focuses on the optimal tracking control problem (OTCP) for the unknown multi-input system by using a reinforcement learning (RL) scheme and nonzero-sum (NZS) game theory. First, a generic method for the OTCP of multi-input systems is formulated with steady-state controls and optimal feedback controls based on the NZS game theory. Then a three-layer neural network (NN) identifier is introduced to approximate the unknown system, and the input dynamics are obtained by using the derivative of the identifier. To transform the OTCP into a regulation optimal problem, an augmentation of the multi-input system is constructed by using the tracking error and the commanded trajectory. Moreover, we use an NN-based RL method to online learn the optimal value functions of all the inputs, which are then directly used to calculate the optimal tracking controls. All the NN weights are tuned synchronously online with a newly introduced adaptation based on the estimation error. The convergence of the NN weights and the stability of the closed-loop system are analyzed. Finally, a two-motor driven servo system and another nonlinear system are presented to illustrate the feasibility of the algorithm for both linear and nonlinear multi-input systems.     

Hierarchical recursive least squares parameter estimation of non-uniformly sampled Hammerstein nonlinear systems based on Kalman filter

This paper focuses on parameter estimation problems for non-uniformly sampled Hammerstein nonlinear systems. By combining the lifting technique and state space transformation, we derive a nonlinear regression identification model with different input and output updating rates. Furthermore, the unmeasurable state vector is estimated by Kalman filter, and by using the hierarchical identification principle, we develop a hierarchical recursive least squares algorithm for estimating the unknown parameters of the identification model. Finally, illustrative examples are given to indicate that the proposed algorithm is effective.     

Tracking control based on adaptive Bernstein polynomial approximation for a class of unknown nonlinear dynamic systems

In this article, an adaptive tracking control approach using Bernstein polynomial approximation is firstly proposed for an unknown nonlinear dynamic system. Bernstein polynomial approximation aims to compensate the unknown nonlinear dynamic function. However, if Bernstein theorem is directly used, the Bernstein polynomial's coefficients need to be derived from the system dynamic function. Nevertheless, the dynamic function is presumed to be unknown, hence the polynomial approximation still cannot be used for designing this control. In order to obtain the available function approximation, adaptive strategy is considered to estimate these coefficients. Finally, by learning from the classical adaptive algorithm, the undetermined coefficient problem is addressed, so that the valid tracking control is found for the unknown nonlinear dynamic system. According to Lyapunov stability analysis and simulation experiment, it is concluded that the new adaptive scheme can realize the control objective.     

Online adaptive policy iteration based fault-tolerant control algorithm for continuous-time nonlinear tracking systems with actuator failures

In this paper, a novel tracking control scheme for continuous-time nonlinear affine systems with actuator faults is proposed by using a policy iteration (PI) based adaptive control algorithm. According to the controlled system and desired reference trajectory, a novel augmented tracking system is constructed and the tracking control problem is converted to the stabilizing issue of the corresponding error dynamic system. PI algorithm, generally used in optimal control and intelligence technique fields, is an important reinforcement learning method to solve the performance function by critic neural network (NN) approximation, which satisfies the Lyapunov equation. For the augmented tracking error system with actuator faults, an online PI based fault-tolerant control law is proposed, where a new tuning law of the adaptive parameter is designed to tolerate four common kinds of actuator faults. The stability of the tracking error dynamic with actuator faults is guaranteed by using Lyapunov theory, and the tracking errors satisfy uniformly bounded as the adaptive parameters get converged. Finally, the designed fault-tolerant feedback control algorithm for nonlinear tracking system with actuator faults is applied in two cases to track the desired reference trajectory, and the simulation results demonstrate the effectiveness and applicability of the proposed method.     

Path tracking of unmanned agricultural tractors based on a novel adaptive second-order sliding mode control

Unmanned tractors are widely adopted in agricultural operations as autonomous driving technology progresses. The current path tracking control methods are limited by the unstructured farmland, the accuracy and anti-interference ability needed to be improved. This paper presents a novel adaptive second-order sliding mode (ASOSM) control method to tackle the aforementioned problems in practical implementation. First, we introduce a preview lateral offset model based on the preview kinematic and tractor dynamic model, which helps solve the under-actuated problem in path tracking. Then, the ASOSM controller is designed using the revamped adding a power integrator (API) and adaptive mechanism, which ensures that the sliding variable is converged to zero within the finite time. Meanwhile, the chattering problem in traditional sliding mode control is relieved. Finally, a high-fidelity and full-car model is established under Simulink/Carsim environment, and comparative simulations conrm the superiority of the designed control method.     

Command filter based adaptive DSC of uncertain stochastic nonlinear systems with input delay and state constraints

In this paper, a command filter based dynamic surface control (DSC) is developed for stochastic nonlinear systems with input delay, stochastic unmodeled dynamics and full state constraints. An error compensation system is designed to constrain the filtering error caused by the first-order filter in the traditional dynamic surface design. On this basis, the stability proof of DSC for stochastic nonlinear systems based on command filter is proposed. The definition of state constraints in probability is presented, and the problem of stochastic full state constraints is solved by constructing a group of coordinate transformations with nonlinear mappings. The Pade approximation is adopted to deal with input delay. The stochastic unmodeled dynamics is considered, which is processed by utilizing the property of stochastic input-to-state stability (SISS) and changing supply function. All the signals of the system are proved to be semi-globally uniformly ultimately bounded (SGUUB) in probability, and the full state constraints are not violated. The two simulation examples also verify the effectiveness of the proposed adaptive DSC scheme.     

Barrier Lyapunov function based adaptive region tracking control for underwater vehicles with thruster saturation and dead zone

This paper proposes a new adaptive region tacking control scheme with nonlinear error transformation for underwater vehicles based on barrier Lyapunov functions. In the new scheme, a redefinition of the tracking error is given by introducing nonlinear error transformation in prescribed performance control. Although the results created by the new scheme indicate a slight decrease in the tracking precision, the real tracking error will be still kept within the prescribed performance functions, while the control signals also become smoother, compared with the original prescribed performance control scheme. Then an approximation form of the control input with constraints, together with an improved Nussbaum function, is designed to derive the control law for underwater vehicles with thruster saturation and dead zone. Furthermore, a new velocity error variable is given by introducing an auxiliary variable to compensate the effect from thruster saturation. Finally, it is proved that the nonlinear system is semi-global practical finite-time stable and the tracking error is always kept within the prescribed boundaries. The effectiveness of the proposed region tracking control scheme is validated through simulation-based case studies on an underwater vehicle with measurement noise.     

Finite impulse response filter based fault estimation with computational efficiency for linear discrete time-varying systems subject to multiplicative noise

The robust fault estimation problem for linear discrete time-varying (LDTV) systems subject to multiplicative noise is investigated by means of finite impulse response (FIR) filter. A novel analytical redundancy, expressed via all states of the previous time window, is originally established to construct the fault estimator. To ensure the satisfactory fault estimation accuracy in stochastic sense under the interference of random uncertainty, a new performance index in forms of matrix trace function is proposed. An easy-to-check necessary and sufficient condition is presented to obtain the optimal filter gain via minimizing the performance index at each time instant. It is analytically demonstrated that, the newly proposed fault estimation algorithm enjoys obvious computational advantages in updating the filter gain, especially as the length of the time window increases for time-varying systems. Simulation results are finally provided to verify its feasibility and superiority.     

Compound tracking control based on MPC for quadrotors with disturbances

In this paper, a compound control strategy is proposed to realize the trajectory tracking task of quadrotors under operating constraints and disturbances. Disturbances caused by model uncertainties, environmental noises, and measurement disturbances are divided into matched disturbances and unmatched ones, which are compensated and suppressed separately by using two control components. The integral sliding mode control component is designed to actively reject the matched disturbances, and the control system is then transformed into an equivalent control system subject to equivalent disturbances only related to the unmatched disturbances. The remaining equivalent disturbances are treated by a robust model predictive control component based on the idea of constraints tightening, which minimizes the tracking error in an optimization framework and takes both state and input constraints into account explicitly. The derived compound control strategy is based on these two control components. Conditions are provided to guarantee the robust constraint satisfaction, recursive feasibility and closed-loop stability of the tracking error system. An illustrative example on the quadrotors shows the efficiency and robustness of this compound tracking control algorithm.     

基于相关距离Fisher比的混合参数用于说话人识别

为了得到更具区分性的特征参数,采用改进的MFCC提取方法,即低方差性的多窗谱估计MFCC,并在其基础上引入了短时TEO能量和ΔMFCC动态特征参量组合特征进行说话人识别。由于直接将两者进行组合会造成维度过高,计算复杂度增加,为此提出了相关距离Fisher比来对特征参数进行加权和维度筛选,最后送入GMM-UBM分类器模型进行识别。实验表明,改进的混合特征参数相较于单一的特征参量,具备更好的识别能力,使得识别率有一定程度的提高。     

Random-valued impulse noise removal based on component-wise noise detector with auto-tuning function and vector median interpolation

We present a novel switching median filter for the removal of random-valued impulse noise on a color image. The present filter consists of a new noise detector with auto-tuning function, and a vector type noise remover avoiding the occurrence of the color artifacts after the noise removal. The detector has a parameter, which is automatically tuned only by using the distribution information of the noise signal without any training signals, and detects the pixel corrupted by the random-valued impulse noise with the parameter. And the detector operates in each channel independently. As a noise remover, the ordinary vector median filter is employed to interpolate only the pixel which is detected as noisy one by the detector. Through the experiments using some digital color images, the validity and the effectiveness of the present method are verified.     

Removing the feasibility conditions on adaptive fuzzy decentralized tracking control of large-scale nonlinear systems with full-state constraints

This work is dedicated to solving the adaptive fuzzy decentralized tracking control issue of large-scale nonlinear systems with full-state constraints. Different with barrier Lyapunov function, the main difference is that a novel nonlinear state-dependent function (NSDF) is introduced to prevent the state constraints being overstepped. Based on NSDF, the necessary feasibility conditions for virtual controllers are completely removed. Then, the prior knowledge of the unknown virtual control coefficients is no longer required since the original system is transformed via the new affine variable. Under the control strategy, three objectives on system performance are achieved: (a) all signals of the closed-loop system are bounded; (b) the subsystem output closely tracks the reference trajectory and original error is ultimately uniformly bounded; (c) the full-state constraints are not violated for all the time. At the end, two simulation examples are shown to verify the effectiveness of the control method.