Research On Generalized Predictive Control Algorithm Of Continuous Stirred Tank Reactor

Continuous Stirred Tank Reactor(CSTR)is one of the most important equipment in chemical industry.It is widely used in the production of dye products,pharmaceutical production,additive preparation and other industries.The chemical reaction in the reactor is accompanied by a large amount of endothermic or exothermic.The sudden rise or sudden drop of temperature in the reactor directly leads to the difficulty in controlling the quality of products.Therefore,the research on the temperature control of the reactor has always been one of the key research contents in this field.CSTR systems often show strong nonlinear,large time delay and strong coupling characteristics,which make it difficult for conventional PID control methods to meet the control requirements of CSTR.In this thesis,Generalized Predictive Control(GPC)algorithm is used to Control CSTR temperature Control system.For a typical CSTR temperature control system,the GPC algorithm is applied to the it,and MATLAB software is used to compare the simulation experiment with the conventional PID control.The experimental results show that the dynamic performance of the system controlled by GPC algorithm is better.In the rolling optimization process,GPC algorithm uses gradient optimization method to obtain the predictive control input sequence,which has the disadvantage of slow optimization process and low search accuracy.Aiming at the above problems,Particle Swarm Optimization(PSO)algorithm is introduced into the rolling Optimization of GPC algorithm to form a hybrid Optimization PSO-GPC algorithm,which improved the speed of optimization and the accuracy of solution.In order to solve the problem that the standard PSO algorithm is prone to fall into local optimization,this thesis proposes an Improved Particle Swarm Optimization GPC(IPSO-GPC)algorithm based on the PSO-GPC algorithm.Through the test function,it is proved that IPSO algorithm can effectively avoid the above problems,it also has better global search ability,and the control performance and stability of the system are improved.Finally,the CSTR temperature control system is simulated by MATLAB software,and the system response curve is compared and analyzed.The results show that IPSO-GPC algorithm shows better control performance,especially in the case of system model mismatch,it has better anti-interference ability and robustness.The comprehensive control ability is better than other algorithms.