摘要
提出了一种参数自适应微分进化算法(PSADE)来解决配电网变电站选址定容问题。根据算法中变异权重因子和交叉因子对进化过程的影响,经分析认为提高算法性能的关键在于平衡大范围搜索和小范围集中寻优之间的矛盾,通过为种群中个体设置单独的变异权重因子和交叉因子,以进化过程中个体适应度函数值的增量为依据对参数进行动态调整,提高算法的全局寻优能力。在变电站选址定容的数学模型中,考虑负荷预测结果误差对规划的影响,采用三角模糊数描述负荷的不确定性,根据投资费用的模糊期望值构造适应度函数。通过对典型的函数优化问题和实际变电站规划问题的寻优结果分析,验证了所提方法具备更好的全局寻优能力,规划方案经济可行。
A parameter self-adaptive differential evolution algorithm（PSADE） for locating and sizing distribution substations is proposed.According to the function of mutation weighting factor and crossover factor on differential evolution（DE）,the key factor to improve the performance is on the equilibrium between exploration of the entire search space and exploitation of the space already visited.The separate parameters for each individuals and the corresponding updating scheme of evolution procedure are designed.On the problem formulation of substation locating and sizing,the uncertainty of load forecasting result is considered.The fitness function is formulated based on fuzzy expected value model of investment costs.From the testing results of two well known benchmark problems and two practical substation planning areas,it is observed that the proposed PSADE is more efficient than DE and PSO,and the planning results are economical and applicable.
出处
《电力系统保护与控制》
EI
CSCD
北大核心
2010年第15期75-80,109,共7页
Power System Protection and Control
关键词
变电站
规划
微分进化
自适应
选址定容
substation
planning
differential evolution
self-adaptive
locating and sizing
