A rule-based cement kiln control system using neural networks

Abstract

The application of knowledge based systems in process control is now growing apparently, especially in the control of complicated systems. A rule based cement kiln control system using neural networks is presented. A description method of the control rules for cement kiln is first introduced. Then, an implementation scheme making use of object oriented programming is discussed. Based on this method, the hardware and software architecture of the cement kiln control system are presented. The system presents a combination of traditional PID control and knowledge based supervisory control, and has been successfully used for the control of several cement kilns.