Fuli Wang

A simple nonlinear controller with diagonal recurrent neural network

A simple control law analogous to the linear generalized minimum variance (GMV) control is presented for the general unknown nonlinear dynamic processes. With this control law, the iterative search of the control input, which is often encountered in the nonlinear control, can be eliminated, resulting in an e

Phase-based Statistical Modeling,Online Monitoring and Quality Prediction for Batch Processes

cient computation for real-time implementation. The implementation of this control law requires two key quantities to be calculated: the input}output sensitivity function and the quasi-one-step-ahead predictive output. The selection of a diagonal recurrent neural network (DRNN) as the process identier allows a direct estimation of these two quantities, resulting in the proposed control law to be implemented in a straightforward manner. Both simulation and experiment are given to demonstrate the e!ectiveness of the proposed control algorithm. ( 1999 Elsevier Science Ltd. All rights reserved.

Robust adaptive control of bilinear plants with high-order perturbation uncertainties

The paper first presents a comprehensive description of some hot problems in statistical modeling,online monitoring,and quality prediction for batch process based on multivariate statistical techniques,including the development of various solutions and their advantages and disadvantages.Then,phase-based statistical analysis strategies are addressed with a focus on the multiplicity of operation phase and phase transition behaviors.This part analyzes the phase-based process characteristics and their effects on product quality,discusses the inherent basis,and reveals their significance.Finally,from the viewpoint of solving practical problems,the existing problems are explored and their prospective development is discussed.Phase-based statistical analysis for batch processes is important in both theory meaning and application,which will benefit further process monitoring,fault diagnosis and quality prediction.

Predictive control for processes with input dynamic nonlinearity

The robust adaptive control issue is considered for a class of bilinear plants with uncertainties of high-order unmodelled dynamics and bounded disturbances. The generalized minimum variance control strategy is first employed to give a basic optimal control law, followed by modification of introducing the modelling error estimate to the control law. Modified least-squares scheme with a relative dead zone is developed to work with the control law to form a novel robust adaptive control algorithm, without much priori knowledge of the plant. The resulting closed loop system is proven theoretically to be robustly stable to high-order unmodelled dynamics and bounded disturbances.

Optimal iterative learning control with uncertain initializations and disturbances

This paper is concerned with the modeling and control of processes with input dynamic nonlinearity. Rather than modeling the overall process with a nonlinear model, it is proposed to represent the process by a composite model of a linear model (LM) and a feedforward neural network (FNN). The LM is to capture the dominant linear dynamics, while the FNN is to approximate the remaining nonlinear dynamics. The controller, in correspondence, consists of two sub-controllers: a linear predictive controller (LPC) designed based on the LM, and an iterative inversion controller (IIC) designed based on the FNN. These two sub-controllers work together in a cascade fashion that the LPC computes the desired reference input to the IIC via an analytic predictive control algorithm and the IIC then determines the process manipulated variable. Since the neural network is used to model the nonlinear dynamics only, not the overall process, a relatively small sized network is required, thus reducing computational requirement. The combination of linear and nonlinear controls results in a simple and effective controller for a class of nonlinear processes, as illustrated by the simulations in this paper.

PCA-Based Modeling and On-line Monitoring Strategy for Uneven-Length Batch Processes

Abstract This paper presents a design of robust iterative learning controller. Sufficient and necessary condition to ensure robust BlBO (bounded-input boundedoutput) stability is derived for the optimal iterative learning controllers when tracking arbitrary bounded output. A practical scheme of selecting the weighting matrices is proposed for the process with uncertain initial resetting and disturbances, to ensure system performance improvement from batch to batch. An application to the injection molding control is given to demonstrate the effectiveness of the analysis.