Study Of Intelligent Bypass System Based On Particle Swarm Optimization Algorithm

Bypass control system is an important auxiliary equipment in the large unitsrunning.It has the functions of harmonizing startup, reclaiming fluid，reducingspoilage, reducing emission. It is great significance in the large-scale thermal power plantsrunningThe steam turbine bypass control system is in parallel with the steam turbinesteam vacuum system, it is composed of valves, piping, and regulatory agencies. Therole is in the unit start-up phase or under accident conditions that steam produced bythe boiler don’t through the steam turbine, it go into a pipeline or condense. Thebypass system needs to be coordinated with the units control system and it has theinterlocking device. PID control has the dominant status in the bypass control system.The routine PID control has many shortages. PSO is an optimized evolution algorithmwith enlighten colony intelligence. It simulates the birds’ behavior of looking for foodand translates the problem of finding the optimal solutions into the partical iterativesearching process in the specific space, which will be applied in this paper for PIDparameters optimization in bypass control system.This paper firstly introduces the basic principles of PID controller, parametersadjusting methods and performance judgment index of the control system, and thenintroduces the Particle Swarm Algorithm and the common improved methods. Aimingat the shortages of prematurity convergency, slow convergency speed at later stages,and easy to be trapped into local optimum in the Particle Swarm Algorithm, theSimulated Annealing Algorithm is introduced in the Particle Swarm Algorithm and anew optimized Algorithm—NSAPSO, that is, the Simulated Annealing Algorithm isintroduced during the process of optimization in the Particle Swarm Algorithm.Random acceptance criteria combining SA makes up the shortage of easy to betrapped into the local optimum in the classical Particle Swarm Algorithm solvingcomplicated problems, which makes the algorithm have the ability to choose to acceptnon optimal solution and jump out of the local optimum, at the same time, it has theability of efficient searching. At the meantime, extremum disturbance is introduced inthe algorithm to increase the partical diversity, which can improve the algorithmperformance further. Then the improved PSO algrithm is applied into the parameteroptimized adjusting in PID controller. The simulation results prove that the algorithmis efficient. At last, aiming at the control of unit#3in thermal power plant of the oil field, the bypass system is optimized and upgraded. The improved PSO algrithm isapplied into the parameter optimized adjusting in PID controller. The running resultproves that the bypass system based on the improved PSO algrithm can gain the goodadjusting quality and strong robustness.