动态输电网络规划的组合编码遗传算法

针对动态输电网络规划过程中需要考虑时间决策量的问题，提出了组合编码方式。组合编码方式将多阶段输电网络规划中的时间决策量隐含在编码中，从而使得多阶段的动态输电网络规划问题能够转换成静态规划问题进行求解。该编码方式满足表现型和基因型的1对1映射，及表现型空间与基因型空间距离上的一致性，从而保证了遗传算法的搜索效率。此外，针对动态输电网络规划的特点，对遗传算法的交叉算子、变异方式、适应度函数、惩罚系数等方面进行了改进，进一步改善了遗传算法求解多阶段输电网络规划问题的性能。以19节点系统为例对算法进行了验证，结果表明能够在较短的进化代数内得到问题的最优解。

Combination encoding genetic algorithm for dynamic transmission network planning

A combination encoding is proposed for handling the time decision variable in a dynamic transmission network planning (DTNP). The time decision variable is implied in the combination encoding, which converts the DTNP problem into a static programming problem. The encoding satisfies 1 to 1 mapping between the phenotype and genotype, and meets the consistency of phenotypic space distance and genotypic space distance, which ensure the search efficiency of genetic algorithm. Moreover, according to the characteristics of DTNP problem, some improvement measures are given for crossover operators, mutation mode, fitness function, penalty coefficient so as to enhance the performance of the genetic algorithm to solve DTNP problem. The proposed method is validated using a 19-node system, and results show that the optimal solution can be obtained within reasonable evolutionary generation.