以降低网损为目标函数,将食物链生态进化算法(EEAFC)应用于配电网重构。该算法模拟生态食物链多种群并存的竞争进化方式,提高了个体的多样性和进化的层次性.具有良好的全局和局部搜索能力,避免了已有仿生类算法未成熟收敛等缺陷。根据配电网的特点,提出了基于独立环路的实编码方法,显著降低了变量维数。在对开关的运行方式采用断开、闭合、常闭、联闭4种状态描述的基础上,进一步将联闭状态分为直接联闭和间接联闭2类.并根据它们各自的特点,采用相应的处理方法,确保迭代过程中搜索到的所有解都自动满足网络辐射状及无孤岛的运行要求。对IEEE16节点、IEEE33节点和IEEE69节点3个不同规模测试系统进行分析计算,并与遗传算法及相关文献结果进行比较,结果表明所提算法具有快速的全局寻优能力。
A new distribution network reconfiguration method employing ecology evolutionary algorithm of food chain as optimal means is presented. The algorithm, which simulates the competitive evolutionary approach of diverse stocks of ecological food chain and enhances the diversity of individuals and the levels of evolution, has a good overall and local search capability and overcomes the immature convergence of some other bionic algorithms. According to the characleristics of distribution network, a real number encoding method is proposed for reducing the dimension of decision variable. On the basis of description of 4 operating manner of switches such as open, close, incident close and long-term close, the incident close manner of switches is further divided into direct incident close and indirect incident close, and the different measures are taken to ensure that all the solutions found in iterative process meet the requirements of radial network and non-isolated island. To verify the effectiveness of the proposed method, comparative studies are conducted on IEEE 16 bus, IEEE 33 bus and IEEE 69 bus sample systems, and rather encouraging results are attained.