Optimal for ball mill pulverizing system and its applications

A ball mill pulverizing system is a typical strongly coupled multi-input multi-output (MIMO) plant, which is also nonlinear and widely used in coal-fired power plants. A conventional PID control strategy is not effective for its regulation, so it becomes a difficult problem in thermo-power automation. In view of the key problems in the automatic control of a ball mill system, the authors exploited an optimal expert fuzzy control system based on a net configuration-the Net-Fuzzy control system. The control system has the ability to solve the following technical problems satisfactorily. (1) It solves the measurement of coal load in the ball mill tube using a noise sensor as well as neural networks based soft-sensing, in which, a signal processing technique, multi-function approximation, a multi-media technique, and a fast Fourier transform (FFT) are used. This provides the basic guarantee for control strategy studies. (2) A steady state optimization strategy is presented based on a simple optimization method in order to boost the system´s efficiency. (3) Net configuration software is developed for the system controlling database maintenance; system control and configuration are managed separately, so the more flexibility obtained, the easier it is to be mastered by control and operation engineers. Practicability and advanced quality have been verified by its application to many ball mill-pulverizing systems in several power plants.