Research On Steam Temperature Control System Based On Improved Generalized Predictive Control Algorithm

The temperature of superheated steam is an important performance index of thermal power generator set.In actual operation,the temperature of superheated steam should be kept within a certain range,neither too high nor too low.As a typ ical self-balancing object of thermal engineering,the superheated steam temperature object has the characteristics of large inertia,large delay,strong coupling,nonlinear and time-varying,and the traditional cascade PID controller can not meet the cont rol requirements of superheated steam temperature.Aiming at these problems,this paper studies the superheated steam temperature control system.(1)Can be adaptive optimization of particle swarm optimization(pso)algorithm based on particle swarm optimization(pso)algorithm,in order to avoid in the process of search into local optimum and low convergence precision of problem,using an adaptive optimal velocity updating formula to control individual and global optimal effect of particle velocity,the early stage of the individual optimal effect is bigger,use cauchy mutation for global optimization,In the later stage,the global optimal influence becomes larger and larger.At this time,Gaussian mutation is used to search for the optimization,so as to achieve the joint optimization effect of global search and local development.(2)Based on the field operation data of the power plant,a single input and single output dynamic mathematical model was established for the controlled objects in the leading area and the inertial area of the superheated steam temperature control system respectively by using the adaptive mutation optimization particle swarm optimization algorithm,and the identified model was used to reconstruct the historical operation data of the power plant.The experimental results show that the identified model can well reflect the dynamic characteristics of the controlled object in the superheated steam temperature control system.(3)The direct relationship between the parameters of the generalized predictive control algorithm and the control performance is analyzed and verified b y MATLAB simulation.A control increment matrix compensation GPC algorithm optimized by crowd search algorithm is proposed.The algorithm compensates the current control increment with other items in the control increment matrix to achieve the purpose of suppressing overshot.At the same time,the self-parameter optimization of GPC is carried out by crowd search algorithm to improve the robustness of the control system.(4)An improved GPC-P cascade superheated steam temperature control system is designed with an improved GPC controller as the main regulator and a proportional controller as the secondary regulator.Start with traditional GPC-P superheated steam temperature control system for the control step response simulation,and then adjust the amount of perturbation experiments,the external disturbances and robustness experiments,simulation experiments show that the modified the GPC controller than traditional GPC controller has better stability and rapidity,and has more ideal ability to resist disturbance and the robustness,Its control quality meets expectations.