Application Research Of Model Free Adaptive Predictive Control In Grinding Process

Grinding process is the intermediate process of mineral processing technology. It liberates valuable minerals from worthless gangue for next working procedure through the method of physical grind. As a result of process itself complex characteristics for instance multivariable, nonlinearity and coupling, only manual operation is difficult to achieve the desired working condition. The operation index of grinding directly affects the product quality of next process even the final index of mineral processing technology. The grinding process automatic control is one of the present research contents in grinding field and has great significance in social production development.In view of the control objective and difficulty in grinding process, the whole control process was divided into several single loop control. Model free adaptive control based on data has robustness and few parameters to adjust on-line, is very suitable for industrial process control. Based on model free adaptive control, this paper put forward model free adaptive predictive control combining with predictive control. They are used in design of single loop controller and main controller respectively, and the simulation results in MATLAB environment reach the expected requirements.Finally, model free adaptive predictive control strategy is verified experimentally relying on comprehensive experiment platform of automation in process industry. In this paper, the whole project of grinding process automatic control system is introduced in detail about how to establish mechanism model, how to simulate characteristics of actuator and detector, and how to operate Win CC monitor and optimal control. Reserved experimental results of comprehensive experiment platform show satisfying control effect, which prove the effectiveness and practicality of the control tactic based on loop control.