Adaptive recurrent neural network control using a structure adaptation algorithm

This paper proposes an adaptive recurrent neural network control (ARNNC) system with structure adaptation algorithm for the uncertain nonlinear systems. The developed ARNNC system is composed of a neural controller and a robust controller. The neural controller which uses a self-structuring recurrent neural network (SRNN) is the principal controller, and the robust controller is designed to achieve L ² tracking performance with desired attenuation level. The SRNN approximator is used to online estimate an ideal tracking controller with the online structuring and parameter learning algorithms. The structure learning possesses the ability of both adding and pruning hidden neurons, and the parameter learning adjusts the interconnection weights of neural network to achieve favorable approximation performance. And, by the L ² control design technique, the worst effect of approximation error on the tracking error can be attenuated to be less or equal to a specified level. Finally, the proposed ARNNC system with structure adaptation algorithm is applied to control two nonlinear dynamic systems. Simulation results prove that the proposed ARNNC system with structure adaptation algorithm can achieve favorable tracking performance even unknown the control system dynamics function.     

并联混合有源电力滤波器自适应L2增益鲁棒控制

为了抑制外界未知扰动和参数摄动对并联混合有源电力滤波器(SHAPF) 系统性能的影响, 提出一种新型的自适应L₂ 增益鲁棒控制策略. 首先建立含有扰动和参数摄动的SHAPF 欧拉-拉格朗日(EL) 数学模型, 得到了SHAPF 在dq 坐标系下的误差动态模型; 然后通过构造适当的Lyapunov 函数设计参数自适应控制率, 实现了对系统参数摄动的补偿, 进而利用阻尼注入方法设计系统的L₂增益鲁棒控制器, 以保证闭环系统的 gamma  耗散性. 仿真实验验证了所提出策略的正确性和有效性.

     

不确定串联非线性系统H∞鲁棒自适应控制

针对一类含有未知参数和干扰的非最小相位串联非线性系统,结合H_∞控制和自适应控制方法并利用李雅普诺夫函数递推设计方法设计了状态反馈H_∞自适应控制器,避免了求解Hamilton-Jacobi-Isaacs不等式设计控制器的困难.该控制器不仅保证闭环系统ISS(input-to-state)稳定,而且使得系统对于所有允许的参数不确定从干扰输入到可控输出的L₂增益不大于给定的值.最后,给出了一个仿真例子,仿真结果充分表明了所设计的控制器的可行性和有     

Design of intelligent power controller for DC–DC converters using CMAC neural network

DC–DC converters are the devices which can convert a certain electrical voltage to another level of electrical voltage. They are very popularly used because of the high efficiency and small size. This paper proposes an intelligent power controller for the DC–DC converters via cerebella model articulation controller (CMAC) neural network approach. The proposed intelligent power controller is composed of a CMAC neural controller and a robust controller. The CMAC neural controller uses a CMAC neural network to online mimic an ideal controller, and the robust controller is designed to achieve L ₂ tracking performance with desired attenuation level. Finally, a comparison among a PI control, adaptive neural control and the proposed intelligent power control is made. The experimental results are provided to demonstrate the proposed intelligent power controller can cope with the input voltage and load resistance variations to ensure the stability while providing fast transient response and simple computation.     

Robust neural network control system design for linear ultrasonic motor

Linear ultrasonic motor (LUSM) has much merit, such as high precision, fast control dynamics and large driving force, etc.; however, the dynamic characteristic of LUSM is nonlinear and the precise dynamic model of LUSM is difficult to obtain. To tackle this problem, this study presents a robust neural network control (RNNC) system for LUSM to track a reference trajectory with L ₂ robust tracking performance. The developed RNNC system is composed of a neural network controller and a robust controller. The neural network controller is the principal controller used to mimic an ideal controller and the robust controller is adopted to achieve L ₂ robust tracking performance. The developed RNNC system is then applied to control an LUSM. Experimental results show that the developed RNNC system can achieve favorable tracking performance with unknown of LUSM model.     

Adaptive dynamic CMAC neural control of nonlinear chaotic systems with L2 tracking performance

The advantage of using cerebellar model articulation control (CMAC) network has been well documented in many applications. However, the structure of a CMAC network which will influence the learning performance is difficult to select. This paper proposes a dynamic structure CMAC network (DSCN) which the network structure can grow or prune systematically and their parameters can be adjusted automatically. Then, an adaptive dynamic CMAC neural control (ADCNC) system which is composed of a computation controller and a robust compensator is proposed via second-order sliding-mode approach. The computation controller containing a DSCN identifier is the principal controller and the robust compensator is designed to achieve L₂ tracking performance with a desired attenuation level. Moreover, a proportional–integral (PI)-type adaptation learning algorithm is derived to speed up the convergence of the tracking error in the sense of Lyapunov function and Barbalat’s lemma, thus the system stability can be guaranteed. Finally, the proposed ADCNC system is applied to control a chaotic system. The simulation results are demonstrated that the proposed ADCNC scheme can achieve a favorable control performance even under the variations of system parameters and initial point.     

Robust Neural Network Controller Design For A Biaxial Servo System

A robust control method for synchronizing a biaxial servo system motion is proposed in this paper. A new neural network based cross‐coupled control and neural network techniques are used together to cancel out the skew error. In the proposed control scheme, the conventional fixed gain PID cross‐coupled controller (PIDCCC) is replaced with the neural network cross‐coupled controller (NNCCC) to maintain biaxial servo system synchronization motion. In addition, neural network PID position velocity and velocity controllers provide the necessary control actions to maintain synchronization while following a variable command trajectory. This scheme provides strong robustness with respect to uncertain dynamics and nonlinearities. The simulation results reveal that the proposed control structure adapts to a wide range of operating conditions and provides promising results under parameter variations and load changes.     

Neural network solution for finite-horizon H ∞ state feedback control of nonlinear systems

In this article, neural networks are used to approximately solve the finite-horizon optimal H _∞ state feedback control problem. The method is based on solving a related Hamilton–Jacobi–Isaacs equation of the corresponding finite-horizon zero-sum game. The neural network approximates the corresponding game value function on a certain domain of the state-space and results in a control computed as the output of a neural network. It is shown that the neural network approximation converges uniformly to the game-value function and the resulting controller provides closed-loop stability and bounded L ₂ gain. The result is a nearly exact H _∞ feedback controller with time-varying coefficients that is solved a priori offline. The results of this article are applied to the rotational/translational actuator benchmark nonlinear control problem.     

Design of decentralized robust controller for voltage regulation and stabilization of multimachine power systems

This paper presents a decentralized state-feedback controller design based on robust control theory to ensure system stability and voltage regulation in multimachine power systems. The power system is decomposed in n subsystems each represented by a state-space model with bounded parameter uncertainties and unknown input disturbances of class L _∞ which model couplings with the generators of the others subsystems. The proposed controller designed according to a Riccati-based approach is robust with respect to uncertain network parameters and counteracts the effects of the disturbances. A stability analysis in presence of L _∞ disturbances is also given. The control law is straightforward and cost effective because it is function of constant gains and of local measurable machine variables. Numerical simulations give evidence of the achievements in terms of system transient stability as well as voltage regulation, also in comparison with another design technique.     

A self-evolving functional-linked wavelet neural network for control applications

The structure of a neural network is determined by time-consuming trial-and-error tuning procedure in advance for the reason that it is difficult to consider the balance between the neuron number and the desired performance. To attack this problem, a self-evolving functional-linked wavelet neural network (SFWNN) is proposed. Without the need for preliminary knowledge, a self-evolving approach demonstrates that the properties of generating and pruning the hidden neurons automatically. Then, an adaptive self-evolving functional-linked wavelet neural control (ASFWNC) system which is composed of a neural controller and a supervisory compensator is proposed. The neural controller uses a SFWNN to online estimate an ideal controller and the supervisory compensator is designed to eliminate the effect of the approximation error introduced by the neural controller upon the system stability in the Lyapunov sense. To investigate the capabilities of the proposed ASFWNC approach, it is applied to a chaotic system and a DC motor. The simulation and experimental results show that favorable control performance can be achieved by the proposed ASFWNC scheme.     

Finite-time H∞ control of periodic piecewise linear systems

In this paper, the finite-time stability, stabilisation, L₂-gain and H_∞ control problems for a class of continuous-time periodic piecewise linear systems are addressed. By employing a time-varying control scheme in which the time interval of each subsystem constitutes a number of basic time segments, the finite-time controllers can be developed with periodically time-varying control gains. Based on a piecewise time-varying Lyapunov-like function, a sufficient condition of finite-time stability and the relevant time-varying controller are proposed. Considering the finite-time boundedness of the closed-loop periodic system, the L₂-gain criterion with continuous time-varying Lyapunov-like matrix function is studied. A finite-time H_∞ controller is proposed based on the L₂-gain analysis. Finally, numerical simulations are presented to illustrate the effectiveness of the proposed criteria.     

Intelligent position tracking control for LCM drive using stable online self-constructing recurrent neural network controller with bound architecture

In this paper, an intelligent position tracking control (IPTC) is developed for a linear ceramic motor (LCM) drive system. The IPTC system is comprised of a neural controller and a robust controller. The neural controller utilizes a self-constructing recurrent neural network (SCRNN) to mimic an ideal computation controller, and the robust controller is designed to achieve L₂ tracking performance with a desired attenuation level. If the approximation performance of SCRNN is insufficient, SCRNN can create new hidden neurons to increase the learning ability. If the hidden neuron of SCRNN is insignificant, it should be removed to reduce the computation load; otherwise, if the hidden neuron of SCRNN is significant, it should be retained. Moreover, the adaptive laws of controller parameters are derived in the sense of Lyapunov, so system stability can be guaranteed. Finally, the experimental results of the LCM drive system show a perfect tracking response can be achieved using the self-constructing mechanism and the on-line learning algorithm.     

Robust Tracking Control For A Wheeled Mobile Manipulator With Dual Arms Using Hybrid Sliding‐Mode Neural Network

In this paper, a robust tracking controller is proposed for the trajectory tracking problem of a dual‐arm wheeled mobile manipulator subject to some modeling uncertainties and external disturbances. Based on backstepping techniques, the design procedure is divided into two levels. In the kinematic level, the auxiliary velocity commands for each subsystem are first presented. A sliding‐mode equivalent controller, composed of neural network control, robust scheme and proportional control, is constructed in the dynamic level to deal with the dynamic effect. To deal with inadequate modeling and parameter uncertainties, the neural network controller is used to mimic the sliding‐mode equivalent control law; the robust controller is designed to compensate for the approximation error and to incorporate the system dynamics into the sliding manifold. The proportional controller is added to improve the system's transient performance, which may be degraded by the neural network's random initialization. All the parameter adjustment rules for the proposed controller are derived from the Lyapunov stability theory and e‐modification such that uniform ultimate boundedness (UUB) can be assured. A comparative simulation study with different controllers is included to illustrate the effectiveness of the proposed method.     

L 2-gain analysis for networked predictive control systems based on switching method

This article deals with the problem of L ₂-gain for a class of networked control systems with time-varying network delay in both forward and feedback channels. An improved network predictive control scheme is proposed to compensate the effects of network delay and data dropout using a switching control strategy. Based on the average dwell time method, the restrictions that the original delay-compensation strategy imposes on the subsystems are relaxed and a sufficient condition for weighted L ₂-gain is developed for a class of switching signal. An example illustrates the effectiveness of the proposed method.     

Noncertainty equivalent adaptive control of robot manipulators without velocity measurements

This paper presents a noncertainty equivalent adaptive motion control scheme for robot manipulators in the absence of link velocity measurements. A new output feedback adaptation algorithm, based on the attractive manifold design approach, is developed. A proportional-integral adaptation is selected for the adaptive parameter estimator to strengthen the passivity of the system. In order to relieve velocity measurements, an observer is designed to estimate the velocities. The controller guarantees semiglobal asymptotic motion tracking and velocity estimation, as well as L_∞ and L₂ bounded parameter estimation error. The effectiveness of the proposed controller is verified by simulations for a two-link robot manipulator and a four-bar linkage. The results are further compared with the earlier certainty-equivalent adaptive partial and full state feedback controller to highlight potential closed-loop performance improvements.     

Passivity-based robust feedback control for non-linear systems with input dynamical uncertainty

This paper addresses the robust feedback control problem for a class of non-linear systems with uncertain input dynamics. The main objective is to develop a passivity-based systematic design approach for this kind of uncertain system. First, a passivity condition is presented for a non-linear system in feedback interconnection form, and then it is shown that with the help of this condition, a state feedback control law can be designed to render the uncertain system passive. Moreover, the extensions of the passivation controller are investigated in two cases where the uncertainty allows unknown control direction and there exists the external disturbance, respectively. It will be shown that an adaptive controller with a Nussbaum-type function can be incorporated into the passivation controller to deal with the unknown control direction and the L ₂-gain performance can be achieved by gain re-assignment of the passivation controller. Finally, a numerical example is given to demonstrate the proposed approach.     

基于神经网络Smith预估器的预测控制

针对非线性系统时滞问题,给出了一种新型的单神经元Smith预测控制算法.神经网络的预测控制器由不完全微分的单神经元自适应PID控制器和神经网络的Smith预估器组成.预估器对输出进行多步预测,控制器超前动作以消除时滞对系统的影响.不完全微分的单神经元自适应PID控制器通过改进的Hebb学习规则实现其权值调节,通过权系数的在线调整实现自适应控制.仿真实验证明了该方法具有较快的响应速度和较好的响应性能.     

The C-loss function for pattern classification

This paper presents a new loss function for neural network classification, inspired by the recently proposed similarity measure called Correntropy. We show that this function essentially behaves like the conventional square loss for samples that are well within the decision boundary and have small errors, and L₀ or counting norm for samples that are outliers or are difficult to classify. Depending on the value of the kernel size parameter, the proposed loss function moves smoothly from convex to non-convex and becomes a close approximation to the misclassification loss (ideal 0–1 loss). We show that the discriminant function obtained by optimizing the proposed loss function in the neighborhood of the ideal 0–1 loss function to train a neural network is immune to overfitting, more robust to outliers, and has consistent and better generalization performance as compared to other commonly used loss functions, even after prolonged training. The results also show that it is a close competitor to the SVM. Since the proposed method is compatible with simple gradient based online learning, it is a practical way of improving the performance of neural network classifiers.     

Neural network-based robust finite-time control for robotic manipulators considering actuator dynamics

A novel neural network-based robust finite-time control strategy is proposed for the trajectory tracking of robotic manipulators with structured and unstructured uncertainties, in which the actuator dynamics is fully considered. The controller, which possesses finite-time convergence and strong robustness, consists of two parts, namely a neural network for approximating the nonlinear uncertainty function and a modified variable structure term for eliminating the approximate error and guaranteeing the finite-time convergence. According to the analysis based on the Lyapunov theory and the relative finite-time stability theory, the neural network is asymptotically convergent and the controlled robotic system is finite time stable. The proposed controller is then verified on a two-link robotic manipulator by simulations and experiments, with satisfactory control performance being obtained even in the presence of various uncertainties and external disturbances.     

船舶航向非线性系统鲁棒跟踪控制

对船舶航向非线性系统, 提出了一种基于神经网络方法的鲁棒跟踪控制器. 系统由船舶运动非线性响应模型和舵机伺服系统串联构成, 其中运动响应模型考虑了建模误差和外界干扰力等非匹配不确定性. 对建模误差和期望舵角的一阶导数项应用在线二层神经网络予以辨识和补偿, 不确定性干扰项处理应用L2增益设计. 采用Lyapunov函数递推法, 得到包括神经网络权值算法在内的跟踪控制器. 跟踪误差和神经网络权值误差的一致终值有界性保证了系统的鲁棒稳定性, 合理的控制器参数选择保证了控制精度. 仿真结果验证了控制器的有效性.