Intelligent PID Tuning And Its Application In Control System Of Thermal Power Plant

In the process of industrial production, PID control is still the most widely used forits simple structure, strong robustness and easiness to master by operating personnel. Inthermal power plants, more than90%of control systems are using PID controller.Nowadays, because of units’ larger capacity and higher value of parameter, systemsrequire higher reliability and more accuracy. In order to achieve expected control effect,the core of the PID control technology is how to determine the parameters of PIDcontroller. Therefore, it is significant to research PID controller’s parameter tuningmethod in current time.This paper first describes the structure of the PID controller, explicitly calls out theinfluence of each control parameters on the quality of reguliation control loop, and thetroditional PID tuning methods are analyzed. After application of MATLAB SIMULINKsimulation platform for a number of researches, troditional tuning methods make mehave a deeper feeling. This work laid a theoretical foundation for the later research. Thenintroduces several PID parameters self tuning method. With the development ofcomputer technology and the artificial intelligence technology, the advanced PIDparameter tuning methods appear continuously. From the practial point of view, thispaper mainly experts on the expert pattern recognition based self-tuning PID controllerdesign method and relay feedback based self-tuning PID controller design method.Self-tuning PID controller algorithm based on relay feedback, first we learned thebasic theoretical knowledge of relay feedback, detailed derivation of Astrom method andPM method controller tuning formula, design of PID controller based on a given phasemargin. Then analysed the advantages and disadvantages of the two methods bycompared the simulation setting effect. According to the shortcomings found in practicalapplications, I make the improvement to the PM algorithm. Relay self-tuning method issimple to use with less time, it’s an effective way to obtain initial parameters of thecontroller. Expert self-tuning PID control based on pattern recognition algorithm use thestep response for the dynamic characteristics of the controlled object, analysis of therelationship between integration time and oscillation periods, creative by using therelationship between them, a group of experts setting PID rules are gived. By simulationexperiments show that this rule has a good guidance of PID parameter tuning. Using the thermodynamic optimization platform, the two algorithms are applied tothe hearth negative pressure control system. Combining the simulation illustration, theresults of the two methods are analyzed in comparison, to verify the practicalsignificance of the PID parameter tuning method in thermal control system in the thermalpower plant.