Applied Research of a Cooperative Evolution Model in Operation Optimization of the Thermal Power Plant

The on-line optimal control of operation is difficult to achieve with routine optimization methods because of the nonlinear thermal system and the changing frequently conditions in start-up, stop and load change of the large-scale thermal power plant. Based on the research of three typical evolution optimization methods which are evolution strategies, genetic algorithm and evolution programming in evolution optimization theory, the author analyzed the feasibility to solve the parallelism of on-line optimal calculation for thermal system by using evolution optimization methods, the existed problems in on-line optimal application of thermal system by using evolution optimization methods in the world at present. By understanding the relationship between the sub-population scales, the evolutional efficiency of layered cooperative evolution genetic algorithm and the thermal system characteristic, the author proposed the sub-population evolution layer model and the population evolution layered cooperative model. Furthermore, the author addressed the adaptive adjusting algorithm of sub- population scale, presented the adjusting standard and method of sub-population scale. The feasibility and efficiency of this model are verified by simulation experiments in 300MWXPDS simulator at Wuhan University.