Application Research Of Improved Stair-like Predictive Control Algorithm In Thermal System

The main steam temperature object of thermal power plant has the characteristics of large lag,large inertia,time-varying and nonlinearity,which makes it difficult to achieve satisfactory control results with traditional control strategies such as cascade PID control.Predictive control has low requirements for model and strong robustness.It has been paid more and more attention in thermal automatic control system.In this paper,a stair-like dynamic matrix control algorithm is constructed by introducing the stair-like strategy,which avoids the matrix inversion process of the basic dynamic matrix control,does not need to consider the existence of the inverse matrix,and improves the real-time performance of the operation.After obtaining the new control law,the closed-loop control system is equivalent to the internal model control structure,and the stability of the controller is analyzed.On this basis,a series of improvements are proposed for the stair-like dynamic matrix control algorithm itself.The measurable but uncontrollable disturbance is suppressed by introducing feedforward compensation,the controller parameters are improved for objects with pure delay,and the shift matrix of the feedback correction link is improved for the truncation error.The law of controller parameter selection is obtained through simulation,and the excellent control performance of the proposed algorithm and the effectiveness of the improved strategy are verifiedThe traditional control strategy use a single controller to achieve the set value tracking and anti-disturbance requirements,which leads to a contradiction between the rapidity and anti-disturbance performance of the controller,and the controller parameters must be considered in compromise.Therefore,by introducing the equivalent input disturbance estimator,the unmeasurable disturbance compensation and feedback control are separated,the independent design of disturbance compensation and feedback controller is realized,which constitutes a two-degree-of-freedom control system.Further stability analysis of the basic equivalent input disturbance estimator is carried out,and it is found that the state observer gain and the disturbance estimator transfer function are coupled,which makes it impossible to arbitrarily design the relevant parameters of the disturbance compensation link.The parameter decoupling is achieved by adding a stable zero point,and the gain of the state observer and the disturbance estimator can also be designed separately,For the nonlinear main steam temperature object caused by the large change of dynamic characteristics in the global working condition range,the local model network is introduced to form a multi-model control strategy,combined with the improved equivalent input disturbance estimator,and finally obtain the multi-model disturbance suppression stair-like dynamic matrix control strategy.Compared with the stair-like dynamic matrix control strategy based on disturbance observer,the effectiveness of the proposed algorithm is verified by global working condition simulation experiments.This paper introduces the optimization control platform developed by the author.In the software of the optimization control platform,the C language is used to develop the relevant modules to carry out the control experiment of simulated objects.Experiments show that the proposed algorithm can improve the global set point tracking,robustness and anti-disturbance ability,and can meet the control requirements of the main steam temperature object.