The Research And Implementation Of Time-delayed Object Predictive Control Algorithm Based On The Physical Platform

The thermodynamic objects in industrial production process often exist the characteristics of large delay,great inertia,strong nonlinearity and uncertain model parameter,which makes the traditional PID control is difficult to achieve satisfactory control effect.And this is very general during the electric power generation.Dynamic matrix predictive control algorithm is easy to implement,it has the advantages of quick response,strong robustness,strong anti-interference,low requirements for the accuracy of model and great comprehensive control quality.Therefore,this article uses dynamic matrix predictive control algorithm to design the controller for thermodynamic objects.In order to make the prediction control algorithm can be implemented in the limited hardware resources equipment,this paper adopts the aggregation strategy algorithm to simplify dynamic matrix predictive control algorithm,and put forward new prediction model,dynamic matrix control,unconstrained control law.In the last,make simulation in MATLAB.For the output constraint of actuator in actual thermodynamic objects control process,this article further considers the output constraint of dynamic matrix predictive control algorithm.According to the dynamic matrix predictive control algorithm which dimension is reduced by aggregation strategy,this paper proposes the output constraint formula of prediction controller and uses the golden section method and improved particle swarm optimization algorithm respectively to solve optimal control output in the constraint interval.In order to make the algorithm research results are more valuable to the practical application,this article uses the typical and actual thermodynamic objects to establish a set of predictive control physical experiment platform.Meanwhile,the constrained predictive control algorithm is implemented and checked in this platform.Finally,this paper combines the project of boiler denitration algorithm optimization and automatic control system commissioning in a power plant.According to the actual situation,this article designs a set of constrained feedforward predictive control algorithm and uses the Siemens S7-200 PLC to realize algorithm functions.The on-site operation shows that the constrained feedforward predictive control algorithm can effectively overcome the large delay,nonlinear in the denitration process,and achieve the satisfactory control results.