Performance Assessment For A Class Of Nonlinear Systems

As investigated by the foriegn companies, until now about 60% controllers existing inindustrial process have performance problems. Control performance of industrial processwould in?uence the product quality, or even the safty of the runnig process. Thus, with thedevelopment of the automation for the industrial process, academic and industrial circleshave recognized the importance of maintenance of industrial control loops. Control perfor-mance assessment has been an significant direction of process control. Since the year 1989when a method based on minimum variance using time series technique for control perfor-mance assessment was proposed by Harris, performance assessment for linear systems hasdevoloped a lot. As for linear systems, performance assessment for PID and predictive con-trollers is becoming mature. Some products for control performance assessement have beenapplied successfully to the actual industrial process. However, nonlinearity exists commonlyin various industrial process, and performance assessment for nonlinear systems is still inthe stage of research.This paper focuses on performance assessment for a class of nonlinear system which iswidely used in industrial process, and make the following progress:1). the effect when using traditional control performance assessment method for nonlinearsystems is discussed2). a piece-wise linear method for fitting a class of nonlinear system is analyzed3). control performance assessment for nonlinear system using piece-wise linear model isproposedThe result extends the traditional control performance assessment. The result revealsthat such method is qualified to performance assessment for a class of nonlinear system.