Decoupling Of The Multi-temperature Measurement And Control System Analysis

In recent years, multiple temperature zones measurement and control system is used in industrial production more and more widely. This control object is a multi-inputs and multi-outputs system, which has features of strong coupling, non-linear, variable parameter. Because of the serious coupling among variables, perfect control result can hardly be gotten by single loop PID controller. This paper introduces two decoupling control algorithm for practical object—multiple temperature zones electric furnace.Firstly, the present status of decoupling controlling is studied, the stability and couple characteristic of system is analyzed based on the present model; feedback compensation decoupling control consisting of two classical structures of decoupling system is proposed and the decoupling controller is designed, integral separation PID is used in subsystem after decoupling. The simulation results show that this algorithm was better than the general PID control. Then its nonlinear and hysteretic nature was analyzed essentially, fuzzy neural networks decoupling control which combines fuzzy logic with NN is put forward to realize multiple temperature zones decoupling. This method doesn't depend on accurate mathematic model and implements self-learning of controller parameters.The simulation results show that the algorithms discussed in the paper are effective method to couple of the multiple temperature zones measurement and control system and has the better ability of disturbance rejection.