摘要
水电系统的短期发电量最大问题是一大型、动态、有时滞的非线性约束优化问题。针对标准遗传算法的缺陷,提出一种基于实数编码技术的混沌遗传算法用于求解此问题。该算法根据给定个体概率分布函数构造杂交算子,结合混沌和人工神经网络理论,设计了一种混沌变异算子,使算法能有效维持群体多样性,防止和克服进化中的“早熟”现象,对约束条件采用不需要设置惩罚因子的直接比较罚函数方法加以处理。仿真算例验证了该算法在提高解的精度和加快收敛速度方面都有明显改善。
The short-term maximization generation of hydropower system is a large-scale dynamic nonlinear constrained optimization problem with water delay time. Based on the analysis of shortcoming of standard genetic algorithm, a new method of real-value encoding chaotic genetic algorithm to solve this problem is presented in this paper. A new crossover operator is designed in light of probability distribution function and a chaotic mutation operator combined the chaos dynamic property with artificial neural network theory, which maintains the population diversity to prevent and overcome the premature phenomena in the evolutionary process. Constraint coditions can be dealt with using direct comparison penalty function method without need of penalty coefficient. An example of simulation results shows that it is improved on the solution of precision and increased convergence speed.
出处
《水力发电学报》
EI
CSCD
北大核心
2006年第4期1-5,共5页
Journal of Hydroelectric Engineering
基金
国家自然科学基金项目(50539140
50309013
40572166
50409010)
湖北省自然科学基金项目(2005ABA228)
关键词
水电系统
经济运行
遗传算法
混沌
hydropower system
economic operation
genetic algorithm
chaos
