Two‐stage auxiliary model gradient‐based iterative algorithm for the input nonlinear controlled autoregressive system with variable‐gain nonlinearity

This article focuses on the parameter estimation problem of the input nonlinear system where an input variable‐gain nonlinear block is followed by a linear controlled autoregressive subsystem. The variable‐gain nonlinearity is described analytical by using an appropriate switching function. According to the gradient search technique and the auxiliary model identification idea, an auxiliary model‐based stochastic gradient algorithm with a forgetting factor is presented. For the sake of improving the parameter estimation accuracy, an auxiliary model gradient‐based iterative algorithm is proposed by utilizing the iterative identification theory. To further optimize the performance of the algorithm, we decompose the identification model of the system into two submodels and derive a two‐stage auxiliary model gradient‐based iterative (2S‐AM‐GI) algorithm by using the hierarchical identification principle. The simulation results confirm the effectiveness of the proposed algorithms and show that the 2S‐AM‐GI algorithm has higher identification efficiency compared with the other two algorithms.     

Three‐stage multi‐innovation parameter estimation for an exponential autoregressive time‐series model with moving average noise by using the data filtering technique

This paper studies the data filtering‐based identification algorithms for an exponential autoregressive time‐series model with moving average noise. By means of the data filtering technique and the hierarchical identification principle, the identification model is transformed into three sub‐identification (Sub‐ID) models, and a filtering‐based three‐stage extended stochastic gradient algorithm is derived for identifying these Sub‐ID models. In order to improve the parameter estimation accuracy, a filtering‐based three‐stage multi‐innovation extended stochastic gradient (F‐3S‐MIESG) algorithm is developed by using the multi‐innovation identification theory. The simulation results indicate that the proposed F‐3S‐MIESG algorithm can work well.     

Performance analysis of multi-innovation gradient type identification methods

It is well-known that the stochastic gradient (SG) identification algorithm has poor convergence rate. In order to improve the convergence rate, we extend the SG algorithm from the viewpoint of innovation modification and present multi-innovation gradient type identification algorithms, including a multi-innovation stochastic gradient (MISG) algorithm and a multi-innovation forgetting gradient (MIFG) algorithm. Because the multi-innovation gradient type algorithms use not only the current data but also the past data at each iteration, parameter estimation accuracy can be improved. Finally, the performance analysis and simulation results show that the proposed MISG and MIFG algorithms have faster convergence rates and better tracking performance than their corresponding SG algorithms.     

Performance Analysis of The Auxiliary‐Model‐Based Multi‐Innovation Stochastic Newton Recursive Algorithm for Dual‐Rate Systems

The stochastic Newton recursive algorithm is studied for dual‐rate system identification. Owing to a lack of intersample measurements, the single‐rate model cannot be identified directly. The auxiliary model technique is adopted to provide the intersample estimations to guarantee the recursion process continues. Intersample estimations have a great influence on the convergence of parameter estimations, and one‐step innovation may lead to a large fluctuation or even divergence during the recursion. In the meantime, the sample covariance matrix may appear singular. The recursive process would cease for these reasons. In order to guarantee the recursion process and to also improve estimation accuracy, multi‐innovation is utilized for correcting the parameter estimations. Combining the auxiliary model and multi‐innovation theory, the auxiliary‐model‐based multi‐innovation stochastic Newton recursive algorithm is proposed for time‐invariant dual‐rate systems. The consistency of this algorithm is analyzed in detail. The final simulations confirm the effectiveness of the proposed algorithm.     

Online identification of nonlinear multivariable processes using self‐generating RBF neural networks

This paper addresses the problem of online model identification for multivariable processes with nonlinear and time‐varying dynamic characteristics. For this purpose, two online multivariable identification approaches with self‐organizing neural network model structures will be presented. The two adaptive radial basis function (RBF) neural networks are called as the growing and pruning radial basis function (GAP‐RBF) and minimal resource allocation network (MRAN). The resulting identification algorithms start without a predefined model structure and the dynamic model is generated autonomously using the sequential input‐output data pairs in real‐time applications. The extended Kalman filter (EKF) learning algorithm has been extended for both of the adaptive RBF‐based neural network approaches to estimate the free parameters of the identified multivariable model. The unscented Kalman filter (UKF) has been proposed as an alternative learning algorithm to enhance the accuracy and robustness of nonlinear multivariable processes in both the GAP‐RBF and MRAN based approaches. In addition, this paper intends to study comparatively the general applicability of the particle filter (PF)‐based approaches for the case of non‐Gaussian noisy environments. For this purpose, the Unscented Particle Filter (UPF) is employed to be used as alternative to the EKF and UKF for online parameter estimation of self‐generating RBF neural networks. The performance of the proposed online identification approaches is evaluated on a highly nonlinear time‐varying multivariable non‐isothermal continuous stirred tank reactor (CSTR) benchmark problem. Simulation results demonstrate the good performances of all identification approaches, especially the GAP‐RBF approach incorporated with the UKF and UPF learning algorithms. Copyright © 2010 John Wiley and Sons Asia Pte Ltd and Chinese Automatic Control Society     

A variable‐structure gradient RBF network with its application to predictive ship motion control

The prediction of dynamic behavior of the nonlinear time‐varying process plays an important role in predictive control applications. Although neural network algorithms have been intensively researched in modeling and controlling nonlinear systems in recent years, most of them mainly focused on the static dynamics. In this paper, a variable‐structure gradient radial basis function (RBF) network is implemented for nonlinear real‐time model predictive control, which is achieved by the proposed gradient orthogonal model selection (GOMS) algorithm. By learning the gradient message of real‐time updated data in a sling window, the structure and the connecting parameters of the network can be adaptively adjusted to adapt to the time‐varying dynamics. The proposed algorithm is evaluated with Mackey‐Glass chaotic time series prediction. Moreover, the variable structure network achieved by GOMS algorithm is applied as a multi‐step predictor in a ship course‐tracking control study, results demonstrate the applicability and effectiveness of the proposed GOMS algorithm and the variable‐RBF‐network based predictive control strategy. Copyright © 2011 John Wiley and Sons Asia Pte Ltd and Chinese Automatic Control Society     

反馈非线性系统随机梯度辨识算法及其收敛性

反馈非线性受控自回归系统是由前向通道的受控自回归模型和反馈通道的静态非线性构成, 这类系统经过参数化后得到双线性参数辨识模型. 本文通过对辨识模型中双线性参数乘积项进行分解, 基于梯度搜索原理, 提 出了反馈非线性系统的随机梯度辨识算法. 为了改善随机梯度算法的收敛速度, 引入遗忘因子, 文章给出了遗忘因子随机梯度算法, 利用随机过程理论, 建立了随机梯度算法的参数估计收敛定理, 证明了算法的收敛性. 最后, 通过数值仿真验证了算法的有效性.     

Recursive parameter estimation methods and convergence analysis for a special class of nonlinear systems

This paper is concerned with the joint estimation of states and parameters of a special class of nonlinear systems, ie, bilinear systems. The key is to investigate new estimation methods for interactive state and parameter estimation of the considered system based on the interactive estimation theory. Because the system states are unknown, a bilinear state observer is established based on the Kalman filtering principle. Then, the unavailable states are updated by the state observer outputs recursively. Once the state estimates are obtained, the bilinear state observer–based hierarchical stochastic gradient algorithm is developed by using the gradient search. For the purpose of improving the convergence rate and the parameter estimation accuracy, a bilinear state observer–based hierarchical multi‐innovation stochastic gradient algorithm is proposed by expanding a scalar innovation to an innovation vector. The convergence analysis indicates that the parameter estimates can converge to their true values. The numerical example illustrates the effectiveness of the proposed algorithms.     

非均匀采样数据系统的一种辨识方法

阐述了非均匀采样方案,推导了非均匀多率采样系统的状态空间模型,进一步获得了对应的传递函数模型.为解决辨识模型信息向量中存在未知变量的问题,使用辅助模型技术,用辅助模型的输出代替系统的未知变量,进而提出了非均匀采样数据系统的辅助模型随机梯度辨识算法.为了提高算法收敛速度和改善参数估计精度,在算法中引入遗忘因子,给出了相应的辅助模型带遗忘因子随机梯度算法.仿真结果表明,引入遗忘因子后,算法的收敛速度加快,参数估计精度提高.     

Parameter estimation for block‐oriented nonlinear systems using the key term separation

This article considers the parameter estimation problems of block‐oriented nonlinear systems. By using the key term separation, the system output is represented as a linear combination of unknown parameters. We give a key term separation auxiliary model gradient‐based iterative (KT‐AM‐GI) identification algorithm and propose a key term separation auxiliary model three‐stage gradient‐based iterative (KT‐AM‐3S‐GI) identification algorithm by using the hierarchical identification principle. Meanwhile, the multiinnovation theory is used to derived the key term separation auxiliary model three‐stage multiinnovation gradient‐based iterative (KT‐AM‐3S‐MIGI) algorithm. The analysis shows that compared with the KT‐AM‐GI algorithm, the KT‐AM‐3S‐GI algorithm can improve the parameter estimation accuracy and reduce the computational burden. In addition, the KT‐AM‐3S‐MIGI can give more accurate parameter estimates than the KT‐AM‐3S‐GI algorithm and can track time‐varying parameters based on the dynamical window data. This work provides a reference for improving the identification performance of multiinput nonlinear output‐error systems or multivariable nonlinear systems. The simulation results confirm the effectiveness of the proposed algorithm.     

On linear‐parameter‐varying (LPV) slip‐controller design for two‐wheeled vehicles

This paper describes the application of linear‐parameter‐varying (LPV) control design techniques to the problem of slip control for two‐wheeled vehicles. A nonlinear multi‐body motorcycle simulator is employed to derive a control‐oriented dynamic model. It is shown that, in order to devise a robust controller with good performance, it is necessary to take into account the dependence of the model on the velocity and on the wheel slip. This dependence is modeled via an LPV system constructed from Jacobian linearizations at different velocities and slip values. The control problem is formulated as a model‐matching control problem within the LPV framework; a specific modification of the LPV control synthesis algorithm is proposed to alleviate controller interpolation problems. Linear and nonlinear simulations indicate that the synthesized controller achieves the required robustness and performance. Copyright © 2008 John Wiley & Sons, Ltd.     

An Efficient on‐Line Parameter Identification Algorithm for Nonlinear Servomechanisms with an Algebraic Technique for State Estimation

This paper presents a methodology for on‐line closed‐loop identification of a class of nonlinear servomechanisms. First, a system is defined with the same structure as the actual servomechanism, but using time‐varying estimated parameters. No coupling between the actual and the estimation systems is present. Position, velocity and acceleration errors, defined as the difference of the actual respective signals and the signals coming from the estimation system, are required in the identification method. Then, a recursive algorithm for on‐line identification of the system parameters is derived from a cost function depending on a linear combination of all the estimation errors. Velocity and acceleration estimates, required in the proposed parameter identification algorithm, are obtained by using an algebraic methodology. The identification algorithm is compared by means of real‐time experiments with an on‐line least squares algorithm with forgetting factor and an off‐line least squares algorithm with data preprocessing. Experimental results show that the proposed approach has a performance comparable to that obtained with the off‐line least squares algorithm, but with the advantage of avoiding any preprocessing.     

A recursive hierarchical parametric estimation algorithm for nonlinear systems described by Wiener‐Hammerstein models

In this paper, a recursive hierarchical parametric estimation (RHPE) algorithm is proposed for stochastic nonlinear systems which can be described by Wiener‐Hammerstein (W‐H) mathematical models. The formulation of parameters estimation problem is based on the prediction error approach and the gradient techniques. The convergence analysis of the developed RHPE algorithm is derived using stochastic gradient‐based theory. Wiener‐Hammerstein hydraulic process is treated to prove the efficiency of the proposed approach.     

Estimation of Multi‐Order Spectra for Nonlinear Closed‐Loop Systems

In this paper, a multi‐order spectra estimation method is proposed for a nonlinear closed‐loop system based on the Volterra series. Owing to the correlation between the noise and the input, the estimation accuracy is poor when the nonlinear spectra of the plant is obtained using the traditional estimation method. In order to overcome this problem, a two‐step scheme is used to estimate the multi‐order spectrum of a nonlinear system operating in closed‐loop. Firstly, the generalized frequency response functions (GFRFs) from the reference signal to the input of the plant are estimated, and they are used to simulate the noise‐free input spectra of the plant. Secondly, the GFRFs of the controlled plant are estimated using the noise‐free input spectra and the output spectra. Because the GFRFs are multidimensional functions, the required amount of calculation for the estimation is very large. To reduce computational complexity, a simplified GFRF model is adopted to estimate the multi‐order nonlinear spectrum of the plant. In this model, the GFRF is transformed to a one‐dimensional function. Two simulation experiments are provided to illustrate the proposed approach.     

Modeling and Control Approach to Coupled Tanks Liquid Level System Based on Function‐Type Weight RBF‐ARX Model

A multi‐input multi‐output (MIMO) FWRBF‐ARX model, which adopts radial basis function (RBF) neural networks with function‐type weights (FWRBF) to approximate the coefficients of the state‐dependent AutoRegressive model with eXogenous input variables (SD‐ARX), is utilized for describing the dynamics of a coupled tanks liquid system. Based on local linearization information of the MIMO FWRBF‐ARX model, a predictive control strategy is proposed. In the algorithm, the control actions of the model predictive control (MPC) are calculated based on the local linearization of the MIMO FWRBF‐ARX model at current working point. Real‐time control experiments are carried out on the coupled tanks liquid system. The detailed comparative experiments demonstrate the feasibility and effectiveness of the proposed modeling and model‐based control strategy for the coupled tanks plant.     

Parameter identification of wiener models by multi-innovation algorithms

An output nonlinear Wiener system is rewritten as a standard least squares form by reconstructing the input-output items of its difference equation. Multi-innovation based stochastic gradient (MISG) algorithm and its derivate algorithms are introduced to formulate identification methods of Wiener models. In order to increase the convergence performance of stochastic gradient (SG) algorithm, the scalar innovation in SG algorithm is expanded to an innovation vector which contains more information about input-output data. Furthermore, a proper forgetting factor for SG algorithm is introduced to get a faster convergence rates. The comparisons of convergence performance and estimation errors of proposed algorithms are illustrated by two numerical simulation examples.     

Extended stochastic gradient identification algorithms for Hammerstein–Wiener ARMAX systems

An extended stochastic gradient algorithm is developed to estimate the parameters of Hammerstein–Wiener ARMAX models. The basic idea is to replace the unmeasurable noise terms in the information vector of the pseudo-linear regression identification model with the corresponding noise estimates which are computed by the obtained parameter estimates. The obtained parameter estimates of the identification model include the product terms of the parameters of the original systems. Two methods of separating the parameter estimates of the original parameters from the product terms are discussed: the average method and the singular value decomposition method. To improve the identification accuracy, an extended stochastic gradient algorithm with a forgetting factor is presented. The simulation results indicate that the parameter estimation errors become small by introducing the forgetting factor.     

An E‐HOIM Based Data‐Driven Adaptive TILC of Nonlinear Discrete‐Time Systems for Non‐Repetitive Terminal Point Tracking

This work proposes a new adaptive terminal iterative learning control approach based on the extended concept of high‐order internal model, or E‐HOIM‐ATILC, for a nonlinear non‐affine discrete‐time system. The objective is to make the system state or output at the endpoint of each operation track a desired target value. The target value varies from one iteration to another. Before proceeding to the data‐driven design of the proposed approach, an iterative dynamical linearization is performed for the unknown nonlinear systems by using the gradient of the nonlinear system with regard to the control input as the iteration‐and‐time‐varying parameter vector of the equivalent linear I/O data model. By virtue of the basic idea of the internal model, the inverse of the parameter vector is approximated by a high‐order internal model. The proposed E‐HOIM‐ATILC does not use measurements of any intermediate points except for the control input and terminal output at the endpoint. Moreover, it is data‐driven and needs merely the terminal I/O measurements. By incorporating additional control knowledge from the known portion of the high order internal model into the learning control law, the control performance of the proposed E‐HOIM‐ATILC is improved. The convergence is shown by rigorous mathematical proof. Simulations through both a batch reactor and a coupled tank system demonstrate the effectiveness of the proposed method.     

Event‐Triggered Consensus of Nonlinear Multi‐Agent Systems with Unknown External Disturbance

This paper addresses the consensus problem of nonlinear multi‐agent systems with unknown external disturbance. First, a distributed observer is proposed to estimate the state and unknown disturbance of each agent simultaneously. Then, a novel event‐triggered control scheme based on the agent state estimation and disturbance estimation is proposed. Unlike the existing strategies, our event‐triggered conditions depend on agent state estimation and disturbance estimation, which are more effective and practical. Under this observer and control strategy, some sufficient conditions are derived to ensure the consensus of the multi‐agent system with unknown external disturbance. Moreover, the Zeno‐behavior of triggering time sequences is also excluded. Finally, a simulation example is given to verify the theoretical analysis.     

Application of temporal difference learning rules in short‐term traffic flow prediction

Studying dynamic behaviours of a transportation system requires the use of the system mathematical models as well as prediction of traffic flow in the system. Therefore, traffic flow prediction plays an important role in today's intelligent transportation systems. This article introduces a new approach to short‐term daily traffic flow prediction based on artificial neural networks. Among the family of neural networks, multi‐layer perceptron (MLP), radial basis function (RBF) neural network and wavenets have been selected as the three best candidates for performing traffic flow prediction. Moreover, back‐propagation (BP) has been adapted as the most efficient learning scheme in all the cases. It is shown that the coefficients produced by temporal signals improve the performance of the BP learning (BPL) algorithm. Temporal signals provide researchers with a new model of temporal difference BP learning algorithm (TDBPL). The capability and performance of TDBPL algorithm are examined by means of simulation in order to prove that the wavelet theory, with its multi‐resolution ability in comparison to RBF neural networks, is a suitable algorithm in traffic flow forecasting. It is also concluded that despite MLP applications, RBF neural networks do not provide negative forecasts. In addition, the local minimum problems are inevitable in MLP algorithms, while RBF neural networks and wavenet networks do not encounter them.