王海军, 董颖华. 基于多策略改进的遗传算法在配电网规划中的应用[J]. 电网与清洁能源, 2021, 37(11): 47-54.
引用本文: 王海军, 董颖华. 基于多策略改进的遗传算法在配电网规划中的应用[J]. 电网与清洁能源, 2021, 37(11): 47-54.
WANG Haijun, DONG Yinghua. Application of an Improved Genetic Algorithm with Multiple-Strategies Improvement in the Distribution Network Planning[J]. Power system and Clean Energy, 2021, 37(11): 47-54.
Citation: WANG Haijun, DONG Yinghua. Application of an Improved Genetic Algorithm with Multiple-Strategies Improvement in the Distribution Network Planning[J]. Power system and Clean Energy, 2021, 37(11): 47-54.

基于多策略改进的遗传算法在配电网规划中的应用

Application of an Improved Genetic Algorithm with Multiple-Strategies Improvement in the Distribution Network Planning

  • 摘要: 针对电力电子化配电网规划复杂的优化问题,提出一种基于多策略改进的多目标遗传算法(简称遗传算法)。将遗传算法与配电网规划进行有效结合,研究了遗传算法在规划方案中的染色体组编码方式;对遗传算法进行具有针对性的多策略改进,涉及种群选择、交叉与变异算子以及自适应遗传算子的改进;通过种群修复提高算法的搜索能力,使染色体的决策变量在满足约束的同时,确保种群多样性启发式地进化为规划问题的最优解。通过Schaffer函数与Griewank函数对基于多策略改进的遗传算法进行性能测试,并对其组成内容、搜索特点与搜索寻优的过程分别进行了分析和讨论。结果表明,基于多策略改进的遗传算法在搜索精度与计算效率方面具有较大优势,对于配电网规划优化具有重要价值。

     

    Abstract: To solve the complex optimization problem of power electronic distribution network planning(PEDNP),a multiple objective genetic algorithm with multiple-strategies improvement is proposed in this paper. First,the combination of genetic algorithm and distribution network planning is carried out, and the chromosome group coding method of genetic algorithm in PEDNP is studied. Second,the multiple-strategies improvement of genetic algorithm is made, including the improvement of the population selection,the crossover and variation operator,and the adaptive genetic operator. Third,the search ability of genetic algorithm is improved by the population repair,so that the decision variables of chromosome can satisfy the constraint requirements while ensure the heuristic evolution of population diversity into the optimal solution of planning problems. Finally, through the Schaffer function and the Griewank function,the performance of the modified genetic algorithm with proposed multi-strategy improvement is tested,and its components, search characteristics and search optimization process are analyzed and discussed respectively.The obtained results show that the modified genetic algorithm with multi-strategy improvement has great advantages in search accuracy and calculation efficiency,and is of important value for the planning and optimization in PEDNP.

     

/

返回文章
返回