The Research Of Unit Commitment In Thermal Power Plant Based On An Immune-Partical Swarm Optimization

Immune particle swarm optimization algorithm (IPSO) is a kind of improved particle swarm optimization algorithm based on the theory of immune algorithm. In order to solve some defects of particle swarm optimization (PSO),such as poor diversity, low quality of solution, premature convergence and so on, this paper introduces immune memory cells,module of blocking antibody and inhibiting antibody,and the adaptive crossover and mutation rate to improve particle swarm algorithm. In the new algorithm, immune memory cells can insure the probability of the convergence; the module of blocking antibody and inhibiting antibody can antibody inhibition can be effectively ensure the rationality of particle aggregation degree; adaptive crossover and mutation operator can guarantee the diversity of particles, as well as the ability of the local search. By analyzing the simulation experiment of several typical test functions, we can see that the improved algorithm is superior in the quality of solution and the convergence.On the optimization problem of combination of thermal power unit (UC), this paper adopts the hybrid immune particle swarm optimization algorithm including binary particle swarm optimization (BPSO) algorithm and immune particle swarm algorithm. BPSO is used to solve unit commitment, when immune particle swarm optimization algorithm is contributed to load distribution. At the same time, to the problem of strong divergence and poor local search ability of BPSO, memory retention strategies and heuristics receptor editing strategy are put forward to improve the algorithm. And the result of simulation proves that the hybrid algorithm is effective on the problem of UC. Finally, after considing the Environmental conservation based on the economical load dispatch in thermal plant, we join a model of environmental cost and establish a new mathematical model of the problem.we also turn the multi-objective optimization problem into single objective optimization problem by weighting method and dimensionless way. The simulation shows the rationality and the feasibility of the new model.