基于免疫进化算法的粒子群算法在梯级水库优化调度中的应用

Application of particle swarm optimization and immune evolutionary algorithm to optimal operation of cascade reservoirs

针对高维、复杂的梯级水库优化调度在求解时易出现"维数灾"或陷入局部最优解的问题,本文提出了基于免疫进化算法的粒子群优化算法,该算法充分利用了免疫进化算法的全局搜索特性和粒子群算法的局部搜索能力,克服了粒子群寻优中对初始种群的依赖和易陷入局部最优的不足。通过实例计算表明,应用该算法求解梯级水库优化调度问题,结果可靠、合理,计算效率高,从而为求解高维,复杂的梯级水库优化调度提供了新的思路。

Optimal operation of a complex cascade reservoirs system, mathematically high dimensional, often encounters dimension disaster and is trapped into a local optimum. This paper suggests a method of particle swarm optimization that is solved by the immune evolutionary algorithm. The algorithm combines the abilities of global searching of immune evolution with local searching of particle swarm, and hence avoids dependence on initial population and local trapping. A test case shows this new algorithm superior to progressive optimization.