摘要
从随机优化技术及生物进化机制角度出发 ,设计出一种新颖的自适应进化规划算法。该算法包含了 2个重要部分 :①采用倒指数形式来描述均方差与适合度之间的关系 ;②在寻优过程中 ,变异量自适应发生改变 ,并结合模糊集理论应用于求解具有可伸缩约束的电网多目标模糊优化运行问题 ,求解时对优化模型、遗传操作等方面进行了探讨。试验系统的计算结果表明 ,该算法具有很强的自适应性及通用性 ,在全局收敛特性、约束项的处理、算法复杂度及运算效率等方面显示了一定的优势 ,体现了求解电网多目标优化运行的良好前景。
A new global or near global optimization method,called self adaptive evolutionary programming is proposed in this paper.The new algorithm includes two important aspects.First,a new modal of mutation which primely reflects the principle of organic evolution in nature is presented.Secondly,the mutation operator is self adaptive during the optimization.Furthermore,the new method and fuzzy set theory is used for solving multi objective optimal operation problem with soft constraints.Some corresponding technical problems,such as model optimized,genetic operation et al are investigated.Numerical results show that the new algorithm not only has the strong self adaptability,versatility and the global optimization capability of escaping local optimum,but can reduce the CPU time effectively.
出处
《中国电机工程学报》
EI
CSCD
北大核心
2001年第3期53-57,61,共6页
Proceedings of the CSEE
