摘要
提出一种基于风险度评价和改进蚁群算法的配电网灵活规划方法。将规划年的预测负荷、电价及导线价格等参数的变化由原来的离散状态转变为多场景区间,从而将电网规划中的单一不确定性问题转化为多个确定性问题。在求解不确定性问题时,提出一种基于云模型的改进蚁群算法,通过定性关联规则推理对蚁群算法中信息素参数ρ和信息素强度Q进行定性控制和动态选取,根据算法进化情况自适应更新支路信息素,有效克服了传统蚁群算法易陷入局部最优解及收敛速度慢的问题。应用改进蚁群算法,依次获得各个场景的规划方案,并以风险度最小为标准确定鲁棒性最优的规划方案,实现电网的灵活规划。算例分析验证了所提方法的有效性。
An algorithm of power distribution network flexible planning was proposed based on risk assessment and improved ant colony algorithm. The parameters of object year, such as load, electrovalence, lead price were changed from discrete state to a section of multi-scenario. The uncertainty problem of planning was changed into several confirmed problems. An adaptive ant colony algorithm based on cloud model was presented. The parameters p and Q of ant colony algorithm are qualitatively controlled and dynamically selected by employing the uncertain qualitative association rule inference based on cloud model. The pheromone of branches is adaptive updated according to the process of the algorithm. The adaptive algorithm overcomes the main disadvantage of being inclined to local convergence and slow convergence speed of traditional ant colony algorithm. The adaptive algorithm is applied in confirmed distribution network structure planning. The scheme of each scenario was obtained and the optimum robust planning scheme was made with considering the minimum cost expectation scheme. Then the distribution network flexible planning could be realized. The instance of the paper indicates the method is feasible.
出处
《中国电力》
CSCD
北大核心
2009年第11期52-56,共5页
Electric Power
关键词
配电网规划
风险度评价
云模型
改进蚁群算法
distribution network planning
risk assessment
cloud model
improved ant colony algorithm
