食物链生态进化算法的改进及其在输电网络规划中的应用  

Improved ecology evolutionary algorithm of food chain and its application in power transmission network planning

作　　者：龚娇龙[1] 王淳[1] 程虹[2] 杨超[2] 吕彬林 彭春华[4] 

机构地区：[1]南昌大学信息工程学院,江西南昌330031 [2]江西省电力科学研究院,江西南昌330096 [3]赣州供电公司,江西赣州341000 [4]华东交通大学电气与电子工程学院,江西南昌330013

出　　处：《电力系统保护与控制》2011年第7期38-43,49,共7页Power System Protection and Control

基　　金：国家自然科学基金(50747025);江西省自然科学基金(2009GZS0016);江西省教育厅项目~~

摘　　要：基于食物链生态进化算法各级种群的规模和品质的不同特点,提出了在低级种群和中级种群中分别采取确定式采样选择和比例选择、高级种群不进行选择操作而采用最优保留策略的改进方法,并将其应用于求解输电网络扩展规划问题。所提出的改进方法进一步均衡了食物链生态进化算法的全局搜索和局部搜索能力,改善了其收敛性能。以具备新建(扩建)线路的走廊作为决策变量,并用实数进行编码,缩短了码串长度。通过对18节点系统的计算,得到了已有文献给出的多个全优和局优方案,并获得了其他文献还未给出的四个比较理想的局优方案。采用所提方法对一个实际系统进行了规划,验证了其适应性。According to the different characteristics of the scale and quality of different population, an improved ecology evolutionary algorithm of food chain（IEEAFC）, which employs proportional model in intermediate population and deterministic sampling in low population, andemploys the optimal reservedst rategyra ther than selective operation in senior population, is proposed and applied to solve the power network expansion planning problem. The IEEAFC further balances the search performance between the overall region and the local region of EEAFC（ecology evolutionary algorithm of food chain）, which improves the convergence of EEAFC. Right of way with new （expanded） lines is taken as decision variable and a real number encoding method is given, which shortens the length Of encoding. The tests are performed on the 18-bus system. As a result, an entire family of optimal solutions, which include four local optimal schemes that have not been reported elsewhere, are gained. Moreover, to verify the feasibility of the proposed method, an actual system is planned.

关 键 词：电力系统 输电网络规划 改进食物链生态进化算法 全局优化 随机搜索 

分 类 号：TM715[电气工程—电力系统及自动化]