| 摘要: |
| 【目的】针对城镇配电网的特点，研究辐射状配电网多目标经济性优化模型与算法，为农村中压配电网优化决策系统的建设提供技术支持。【方法】利用经济电流密度概念和多目标规划理论，提出计入导线截面策略因子的最小投资费用和年网损费用的多目标函数优化模型，研究并利用实数编码并行优化处理方法、引入退火动态罚因子设计动态罚函数处理约束条件、基于遗传算法的电网路径与潮流的自识别解码方法、遗传算子的自适应调整方法等，并提高遗传算法在辐射状配电网拓扑约束中的自动识别能力，确保可行解的搜索方向和遗传寻优的顺利进行。【结果】通过陕西杨凌示范区新桥变电站10 kV配电网待建区域实例验证表明，随着选取种群数目的增多，综合费用逐渐趋向于1 230万元。与单目标优化方案的投资费用相比，多目标优化方案的建设投资费用约增加72万元，而年网损费用约减少323万元，综合费用约降低250万元，获得了该配电网的最优建设方案。【结论】提出了配电网络多目标经济性优化模型及其算法，实现了规划线路长度和导线截面策略因子并行优化，提高了解的搜索精度。 |
| 关键词: 辐射状配电网 多目标优化 遗传算法 经济性 |
| DOI： |
| 分类号: |
| 基金项目:国家自然科学基金项目(50879072)；西北农业科技大学拔尖人才专项基金项目(BJRC-2009-01)；西安市科技计划创新支撑计划项目(CXY1016(5))；西北农业科技大学教学研究项目(JY1102057) |
|
| Optimal method for rural radial power distribution network |
|
|
| Abstract: |
| 【Objective】Aiming at the characteristics of town distribution network,optimal multi-objective economic model and algorithm of radial power grid has been researched to provide a technical support for the decision system of the midvoltage distribution network optimal planning in rural area.【Method】Concerning the strategy factor of wire section,the multi-objetive function of minimum investment costs and annual network loss costs in optimal model were put forward by taking economic current density and theories in multiple objective planning.The processing method of real code parallel optimization,method introducing annealing penalty factor to the design of dynamic penalty function to handle the restrictions,self-recognition decoding method based on genetic algorithm of power grid path and flow and self-adapting adjustment methods of genetic operators were studied and used,to ensure that the genetic algorithm can improve the ability to deploy self-adapting in restrain radial power distribution network topology,the search direction of feasible solution and that the genetic optimization can be carried out smoothly.【Result】When the method is used for 10 kV electricity distribution network for planning area in Xinqiao substation of Yangling,the comprehensive cost tends to be 12.3 million yuan along with the increase of the selected population.Compared with the single objective optimization scheme,the multi-objective optimization scheme needs about an increase of 0.72 million yuan investment and the annual network loss costs about an decrease of 3.23 million yuan,so its comprehensive cost is reduced about 2.50 million and therefore realizes the optimal scheme.【Conclusion】Results show that the proposed mathematical model and algorithm of multiple objective economic optimization in power distribution network can realize the parallel optimization in power lines and the strategy factor of wire section and improve the searching precision of solution. |
| Key words: radial power distribution network multiple objective optimization genetic algorithm economization |
