Abstract: A large amount of distributed photovoltaic accessing to smart distribution grid may lead to an imbalance in the three-phase current, and then could damage the security and stability of the distribution system. So this paper constructs a distribution multi-objective optimization model, the objectives of which are current imbalance minimization and energy loss minimization. This model aims at solving phase balance problems of distribution network after large-scale distributed PV into grid. Then, Shuffled Frog Leap Algorithm is optimized by Random Nelder Mead, improving shortcomings of SFLA. Combining it with decision algorithm, this paper proposes improved multi-objective decision-shuffled frog leaping algorithm for the numerical example. IMO-SLFDA can ensure it obtains the optimal solution in an extremely fast search speed and high accuracy. Last, this paper takes the IEEE 123-bus three phase unbalanced test system as the numerical example, and three phase balance of the distribution system is achieved by controlling the variables related operations. Making the comparative analysis of the difference between base case and optimized example, the proposed algorithm is verified, advanced and practical.

Key words: new energy management planning, shuffled frog leaping decision-making algorithm, distributed generation, smart distribution grid

摘要： 分布式光伏的大量接入智能配电网后,可能导致三相电流的失衡,进而破坏配电系统的安全稳定性。对此,本文构建了以电流不平衡和电能损失最小化为目标的含分布式光伏的配电网优化的多目标模型,旨在解决大规模分布式光伏发电并网后配电网相位平衡的问题;然后,用随机单纯形法对混合蛙跳算法进行优化,改进了蛙跳算法求解优化问题时极易陷入局部最优以及计算效率较低的缺点,并和决策算法相结合,提出适用本文算例的改进的多目标混合蛙跳决策算法,确保能以极快的搜索速度和较高的计算精度得到最优解;最后,以IEEE-123节点三相不平衡测试系统为例,通过控制变量的相关操作实现配电系统的三相平衡。对比分析基础案例和优化算例的差异,验证了本文所提算法的先进性和实用性。

关键词: 新能源管理规划, 混合蛙跳决策算法, 智能配电网, 分布式发电