Investigating the issue of intermittent actuator faults for single-joint robots is both theoretical and practical. In this paper, we develop a prescribed performance adaptive output feedback control scheme for single-joint robots with intermittent actuator faults. Firstly, the estimated parameters in the controller are updated by the projection operator to ensure the boundedness of parameter estimates. Secondly, a modified Lyapunov function is developed to prove the closed-loop stability, proving that the tracking error is always kept within the predefined domain provided by the given performance function. Meanwhile, better tracking performance can be obtained if the jump amplitude of the Lyapunov function is smaller and the minimum fault time interval is longer. Finally, simulation results show the effectiveness of the proposed control scheme.

It is both challenging and meaningful to investigate the control issue for uncertain manipulators with intermittent actuator faults. To solve this problem, an adaptive fractional sliding mode fault-tolerant control scheme is proposed. Within this scheme, a projection operator is used to change the adaptive parameters in controller and a piecewise continuous Lyapunov function is proposed to prove the closed-loop stability. It proves that the boundedness of all closed-loop signals are guaranteed in the case of intermittent actuator faults (IAFs) and the control goal is achieved. Furthermore, when the jump amplitude of the unknown fault parameters is smaller and the time interval between two consecutive actuator faults is longer, the tracking error is smaller. Finally, simulation studies are done to verify the feasibility of the developed control scheme.

控制系统的执行器经常发生各种未知的间歇性故障.如何有效地处理这些故障对系统的影响是一个难题.针对一类不确定严格反馈非线性系统,提出一种自适应CFB(Command filtered backstepping)控制方案解决了间歇性执行器故障的补偿问题.利用神经网络逼近控制器中的未知函数,并采用投影算子实时在线更新控制器中的估计参数使得参数估计值随着故障次数的累积而不断增加的问题被消除.提出改进的Lyapunov函数证明了所提出的方案能够保证所有闭环信号的有界性,同时建立了跟踪误差与Lyapunov函数跳变幅度,最小故障时间间隔,设计参数之间的关系.如果Lyapunov函数的跳变幅度越小以及两个连续故障之间的时间间隔越长,系统的稳态跟踪指标越好.通过迭代计算建立了暂态跟踪误差指标的均方根型界.该界表明了通过选择恰当的设计参数,可改善系统的暂态指标.仿真结果表明了所提方案的有效性.

Due to the existing effects of intermittent jumps of unknown parameters during operation, effectively establishing transient and steady-state tracking performances in control systems with unknown intermittent actuator faults is very important. In this article, two prescribed performance adaptive neural control schemes based on command-filtered backstepping are developed for a class of uncertain strict-feedback nonlinear systems. Under the condition of system states being available for feedback, the state feedback control scheme is investigated. When the system states are not directly measured, a cascade high-gain observer is designed to reconstruct the system states, and in turn, the output feedback control scheme is presented. Since the projection operator and modified Lyapunov function are, respectively, used in the adaptive law design and stability analysis, it is proven that both schemes can not only ensure the boundedness of all closed-loop signals but also confine the tracking errors within prescribed arbitrarily small residual sets for all the time even if there exist the effects of intermittent jumps of unknown parameters. Thus, the prescribed system transient and steady-state performances in the sense of the tracking errors are established. Furthermore, we also prove that the tracking performance under output feedback is able to recover the tracking performance under state feedback as the observer gain decreases. Simulation studies are done to verify the effectiveness of the theoretical discussions.

Effectively compensating unknown intermittent actuator faults in uncertain decentralized nonlinear systems is a very difficult problem, and very few results have been obtained. In this article, to address this issue, an adaptive neural output feedback compensation control scheme based on command-filtered backstepping is developed. First, we design a bank of observers to estimate the system states and utilize neural networks with random hidden nodes to approximate the unknown functions of these observers. Second, a smooth projection algorithm is used to online update estimated parameters in the controllers such that the possible ceaseless increase in the estimated parameters caused by intermittent actuator faults can be eliminated. Due to the presence of intermittent jumps of unknown parameters, a modified Lyapunov function is developed to analyze the system stability. It is proved that the boundedness of all closed-loop system signals is ensured and the ultimate bound of the tracking error depends on design parameters, adjustable jumping amplitude of Lyapunov function, and minimum fault time interval. Third, by analyzing the system transient performance, the peaking phenomenon at the starting instant of the system operation can be removed, and a root mean square type of bound is established to illustrate that the transient tracking error performance is tunable by design parameters. Finally, simulations studies are done to illustrate the effectiveness of the theoretical results.

Actuator faults have been thought to be one of the most main challenges to be addressed because the system stability can be seriously deteriorated by incorrect actuator actions. Therefore, in this paper, an adaptive neural fault tolerant control (FTC) method is developed for a class of multiple-input-multiple-output (MIMO) uncertain nonlinear systems which are subject to actuator faults and external disturbances. The proposed method utilizes extreme learning machine (ELM) neural networks which are of random hidden nodes to approximate the uncertain modules of systems due to their universal approximation property. Meanwhile, our method does not require prior knowledge of the considered systems outlined, and can effectively compensate for the effects of the actuator faults and external disturbances. Moreover, it is proved that all signals in the closed-loop system remain uniformly ultimately bounded. Finally, simulation results demonstrate the effectiveness of the proposed FTC method.

In this paper, an adaptive control scheme based on command-filtered backstepping technique is developed for a class of uncertain multiple-input-multiple-output (MIMO) strict-feedback nonlinear systems. Within this scheme, extreme learning machine (ELM) with random hidden nodes is used in the controller to approximate unknown functions, and a smooth projection algorithm is adopted to adjust online estimated parameters in the controller such that the boundedness of the parameter estimates can be ensured. Furthermore, some stale command filters are designed to produce virtual control signals and their derivations such that the analytic calculation of the partial derivatives of virtual control signals is removed. Also. some other stable filters are proposed to generate compensating signals of above-mentioned filtered errors to compute compensated tracking errors. It is proved that the proposed control scheme can guarantee the boundedness of all signals in the closed-loop system. Finally, the proposed control scheme is applied to control wind turbine and the simulation studies illustrate the theoretic results obtained.

In real applications, actuators of control systems frequently encounter unknown intermittent faults during operation while effectively handing such faults is still a challenge. In this paper, an adaptive neural output feedback fault tolerant control (FTC) scheme based on the command filtered backstepping is developed for a class of uncertain nonlinear systems to address this challenge. In this scheme, a stable nonlinear observer is designed to estimate the system states and neural networks with random hidden nodes are utilized in this observer to approximate unknown functions. A projection algorithm is adopted to estimate system unknown parameters such that the boundedness of parameter estimates is guaranteed. It is proved that the boundedness of all signals in the closed-loop system can be ensured by the proposed modified Lyapunov function. Also the ultimate bound of the tracking error depends on design parameters, adjustable jumping amplitude of Lyapunov function and minimum fault time interval. A truncated L 2 bound is established by iterative calculation to illustrate that the transient tracking error performance is determined by design parameters in the controller and observer. Applications on two simulation examples validate the effectiveness of the proposed scheme.

Rolling bearings are one of the most commonly used components in rotating machinery which is mainly operated in complex working environment. Therefore, it is of great theoretical value and practical significance to study the state monitoring and fault diagnosis technology of rolling bearing to avoid sudden accidents and make a better system maintenance. In this paper, we propose a one-dimensional convolutional neural network to identify rolling bearing fault. Furthermore, we adopt a novel activation function: exponential linear units in the task of rolling bearing fault diagnosis. Simulation results show that one-dimensional convolutional neural network has a prominent generalization ability and high accuracy rate. Exponential linear units can make neural network more robust and stable when we diagnose the rolling bearing fault.

Deep learning has been successfully applied to the field of fault diagnosis in recent years. Due to the advantages of deep belief network (DBN) in fitting nonlinear complex systems and the ability of wavelet analysis in time-frequency analysis, in this paper, an improved fault diagnosis method based on a deep wavelet neural network (DWNN), which combines the DBN with morlet activation functions, is proposed for fault diagnosis of reciprocating compressor. A five-layer DBN using sigmoid, tanh, rectified linear unit (ReLU) and morlet wavelet functions as the activation functions of hidden layers separately is proposed for fault diagnosis of reciprocating compressor. As the contrast, a three-layer back propagation neural network (BPNN) using the same four activation functions separately is proposed for fault diagnosis of reciprocating compressor. The experimental results show that, the fault diagnosis rate of five-layer DBN is higher than the three-layer BPNN. The method based on DWNN can make the fault diagnosis rate reach 100% within short time. Compared with using other activation functions, the DWNN architecture requires less epochs to train the model.

Actuator, as a key component of control system, whose faults detection and diagnosis (FDD) is a complex problem due to system modeling uncertainty, so it is essential to propose advanced FDD scheme that accurately detects the faults. In this paper, we develop an actuator FDD scheme for a class of uncertain nonlinear systems based on extreme learning machine (ELM). In ELM, all parameters of hidden layer nodes need not be adjusted during learning, which may simply be assigned with random values, and the output weights only need to be adjusted. Within this scheme, two ELMs are employed to learn the unknown system function and unknown fault function. Firstly, a stable adaptive observer is designed to monitor faults in an online manner. Secondly, adaptive threshold is designed to make the fault detection, and deviation between the actual and the estimated system is known as residual. If the residual exceeds threshold at finite time denotes a fault occurrence. Different from the existing schemes, online computational efficiency and learning speed are improved considerably because ELM is introduced in this FDD scheme. Finally, a single-link robotic arm will be employed in simulation to illustrate the effectiveness of the proposed FDD scheme.

针对刚性机械臂系统的控制问题,提出基于极限学习机(ELM)的自适应神经控制算法.极限学习机随机选择单隐层前馈神经网络(SLFN)的隐层节点及其参数,仅调整其网络的输出权值,以极快的学习速度获得良好的推广性.采用李亚普诺夫综合法,使所提出的ELM控制器通过输出权值的自适应调整能够逼近系统的模型不确定性部分,从而保证整个闭环控制系统的稳定性.将该自适应神经控制器应用于2自由度平面机械臂控制中,并与现有的径向基函数(RBF)神经网络自适应控制算法进行比较.实验结果表明,在同等条件下,ELM控制器具有良好的跟踪控制性能,表明了所提出控制算法的有效性.

Based on extreme learning machine ( ELM ) , two adaptive neural control algorithms for rigid arm robot system were presented. ELM for signle-hidden layer feedforward neural networks(SLFNs), which randomly chooses hidden node parameters and analytically determines the output weights of SLFNs, tends to provide good generalized performance at extremely fast learning speed. Within these adaptive control algorithms, ELM was employed to approximation the plant’ s unknown nonlinear function and ro-bust control term was used to compensate for approximation error. Parameter adaptive laws and robust control term of ELM controllers were derived based on Lyapunov stability analysis so that global stability and asymptotic convergence to zero of tracking errors can be guaranteed. Futhermore, two adaptive con-trollers do not depend on any parameter initialization conditions and relax the requirement of bounding pa-rameter values. The proposed adaptive ELM control algorithms were then applied to a tracking control in-stance for two-link rigid arm robot and compared with existing radial basis function( RBF) neural control algorithms. Simulation results show that ELM controllers have good tracking performance and demonstrate the effectiveness of the proposed control algorithms.

Based on the extreme learning machine(ELM), a robust adaptive neural control method for a class of multiple-input-multiple-output(MIMO) affine nonlinear dynamic systems is presented. ELM for single-hidden layer feedforward networks(SLFNs), which randomly chooses hidden node parameters and analytically determines the output weights of SLFNs, shows good generalized performance at extremely fast learning speed. The proposed control scheme utilizes the ELM to approximate the plant’s unknown nonlinear terms. Meanwhile, output weights of ELM, unknown upper bound values of approximation errors and external disturbances can be online estimated through parameter adaptive laws by using Lyapunov stability analysis, so that semi-global uniform ultimare boundedness of all signals in the closed-loop system can be guaranteed. Finally, simulation results show the effectiveness of the proposed adaptive ELM control.