Reserch On Supercritical Unit Coordinated Control System Based On Intelligent PID

In recent years, with the development of low-carbon economy, the growing demandfor electricity, for generating set the requirement of increasing the economic performance,the development of large capacity and high parameter, low energy consumption, strongload adaptability of supercritical unit will become a trend. Supercritical unit adoptsunique soda process organization, strong coupling, nonlinear and complex dynamiccharacteristics of dc boiler system.Supercritical unit operation mode of variableparameters, variable characteristics of the control and the dc operation of the specialcharacteristics make them more complicated control system design and debugging.Therefore, the study of supercritical units of intelligent control systems, security, stabilityand economic operation of the unit is extremely important.A lot of control system in power plant adopt PID controller, which has a simplestructure, robustness, adaptability, high reliability, suitable for a precise mathematicalmodel of controlled object. However, supercritical units is difficult to obtain exactmathematical model, the traditional PID controller on-site commissioning method is verycomplex and not well adapted to run variable conditions, the control effect is notparticularly desirable. With the development of the theory of intelligent control researchwill be applied to coordinated control system design is a development direction.In this paper, the intelligent control algorithm in the fuzzy theory, neural networkalgorithm, adaptive control, and decoupling control system for the study and research.In supercritical unit superheated steam temperature model is studied on the basis ofvariable universe fuzzy adaptive PID control algorithm, based on the scaling factor,designed based on fuzzy adaptive PID and variable region of superheated steamtemperature intelligent control system, by using the method of variable domain to adjustthe fuzzy control strategy. The method of fuzzy control flexibility, robustness, simplicityand traditional PID control combined with its simple control, small workload. Simulationresults show that the method of controlling small overshoot, high control accuracy,improve the dynamic and static performance of the system, and achieved satisfactorycontrol effect.Based on the traditional PID controller design methods of learning, on the basis of the neural network theory is more in-depth study, designed the single neuron adaptivePID self-tuning controller gain. The controller used in the existing supercriticalsuperheated steam temperature model, as opposed to the conventional PID controller,which has a relatively good track, robustness, immunity.Study the decoupling control system, the relative gain calculation matrix.Supercritical unit model for coupling analysis. After comparison of the decouplingmethod, multivariate class feedforward decoupling can be accused of model transferfunction matrix into a diagonal matrix, and the diagonal after decoupling the transferfunction of the same, to keep the control characteristic of the main control channel.Thedecoupling method combined with PID, through simulation shows that the methodimplements completely decoupling, the decoupling effect is very good.Finally, the variable universe fuzzy adaptive PID controller, a single neuronadaptive PID controller with feedforward compensation decoupling class systemscombined, respectively, of the main steam pressure supercritical units, temperaturesimulation. Analysis of simulation results, both based intelligent PID control algorithm,tracking, robustness, better immunity than conventional PID stability, improved qualitycontrol multivariable coupling model of coordinated control system.