| Along with the continuous development of our nation’s road, car, and transportation industry, tire manufacturing industry get more and more attention. Internal mixer is the major equipment of mixing process in tire industry. Its control quality could affect the quality of tire product directly. Moreover, temperature control is the key segment in the mixer control system, so it is necessary to research the temperature control of the mixer.The thesis takes mixing production line in Tire Corporation as the background, and expounds the composition of mixing production line and the working principle of mixer. In the paper it is analyzed for mixer temperature control system in detail, and the structure diagram of the control system is illustrated. Based on analyzing of the mixer temperature control process, temperature object itself has an inertia, delay and nonlinear features. At the same time, the factors that influence the temperature during the rubber mixing are complex. To deal with the delay feature of the control object, combined with internal model control structure is simple; adjust course definitude, strong anti-interference, using internal model control structure designed control system. In order to solve the problem of precise math equation can not describe the temperature object; BP neural network is used to identify the internal model and its inverse. A neural network internal model control system is designed. In order to reach the aim of an ideal control effect, it should overcome the error between the object and the internal model, so the filter is introduced into the feedback loop.At last, the author uses the Matlab to simulate the control system. A comparison between the temperature control system designed in this paper and traditional PID control system is done and analyzed the filter parameters on system performance; as well as the step response of control system. The simulation results indicate that the neural network internal model structure has better control effects in the mixer control system which has nonlinear, great inertial and pure lag characters. |