Temperature Control Of Reactor Based On Hysteresis Loop Biased Relay Feedbark

Batch reactor is a typical industrial equipment.Due to its own characteristics and the complexity of chemical reactions,the process of batch reactor has many characteristics,such as large time delay,time-varying and nonlinear,which aggravate the difficulty of temperature control of batch reactor.Conventional control strategies have been difficult to meet the accurate control of temperature of reactor.Therefore,it is necessary to seek higher quality control strategies.The main work and innovation of this paper are as follows:1.In this paper,the mathematical model of the reactor is established based on the mechanism analysis method,and the second order plus dead time model(SOPDT)of the reactor is identified by the hysteresis loop biased relay feedback technique.A large number of relay feedback tests are used to classify the response shapes of SOPDT models with different parameters and the criterion of model structure selection is established.Thus,the identification accuracy is improved.2.A predictive PI controller composed of prediction module and PI module is used for the established large time delay model.In addition,considering the problems of coordination between robustness,tracking performance,and controller output,a set of predictive PI controller parameter self-tuning strategy based on improved particle swarm optimization algorithm is proposed.Finally,based on the first-order Pade approximation and the second-order and unsymmetrical Pade approximation,the internal relationship between the predictive PI and PID is analyzed.Additionally,a novel PID parameter tuning method is applied to the SOPDT model for the first time,so that the PID controller has the similar performance as the predictive PI controller.3.In order to improve the active anti-disturbance ability of the predictive PI control algorithm,a new two-degree-of-freedom predictive PI algorithm with active anti-disturbance ability is proposed.The structure and robustness of the algorithm are analyzed,and the algorithm is applied to the common first order plus dead time model(FOPDT)and SOPDT model.4.Based on the OPTO22 platform,the engineering of temperature control system of reactor is completed,including the realization of control strategy,communication connection between Simulink and OPTO22,development of monitoring interface,etc.The results show that hysteresis loop biased relay feedback identification can better fit higher-order process,and has more accurate identification accuracy and stronger anti-disturbance ability,compared with standard relay feedback identification.In addition,the predictive PI controller with self-tuning parameters has better adaptability,better robustness,smaller overshoot and faster tracking performance,and can provide stable,accurate and rapid control of the temperature of the reactor.The predictive PI controller with active anti-disturbance structure can effectively reduce the adverse effects of internal and external disturbance,where the robustness and anti-disturbance ability of the system are obviously improved.In conclusion,the predictive PI temperature control strategy based on hysteresis loop biased relay feedback has good practical value.