基于负荷趋势判断的无功优化精细化控制方法及求解策略  

Refined Controlling Method and Solving Strategy of Reactive Power Optimization Based on Bus Load Trend Estimation

作　　者：陈光宇 张仰飞[1] 郝思鹏[1] 边二曼 李亚平 邓勇 CHEN Guangyu1, ZHANG Yangfei1, HAO Sipeng1, BIAN Erman2, LI Yaping3, DENG Yong4(1. School of Electric Power Engineering, Nanjing Institute of Technology, Nanjing 211167, Jiangsu Province, China; 2. Department of Development Planning, State Grid Heilongjiang Electric Power Co., Ltd., Harbin 150090, Heilongjiang Province, China; 3. China Electric Power Research Institute （Nanjing）, Nanjing 210003, Jiangsu Province, China; 4. State Grid Fujian Power-Dispatch ＆ Control Center, Fuzhou 350003, Fujian Province, China)

机构地区：[1]南京工程学院电力工程学院,江苏省南京市211167 [2]国网黑龙江省电力有限公司发展策划部,黑龙江省哈尔滨市150090 [3]中国电力科学研究院有限公司(南京),江苏省南京市210003 [4]国网福建电力调度控制中心,福建省福州市350003

出　　处：《电网技术》2018年第4期1259-1265,共7页Power System Technology

基　　金：国家自然科学基金项目(51407165);江苏省配电网智能技术与装备协同创新中心开放基金资助项目(XTCX201713)~~

摘　　要：针对短期内母线负荷波动导致无功优化控制效果不理想的问题,提出一种无功优化精细化控制方法用于缓解因控制方案滞后而导致的控制不平滑现象,并采用多目标无功优化松弛模型来改善预测控制中可能出现的电压越限和收敛性问题。为了获得最优控制方案,给出一种基于动态搜索策略的多目标混沌差分进化算法,该算法能根据种群中可行解的比例动态调整进化过程中最优解的搜索策略,提高多目标模型Pareto最优前沿的搜索能力和求解效率。IEEE 30标准节点数据仿真结果表明,多目标优化算法在最优解集,外部解收敛性,以及解集的均匀性等方面都好于经典多目标算法;真实电网数据计算表明,精细化控制方法相比传统方法能进一步减小电压偏差和网损,并提高模型的收敛性。In order to solve unsatisfactory control effect of reactive power optimization with bus load fluctuation in short term, a refined control method of reactive power optimization is put forward to relieve unsmooth phenomenon in control due to left-behind control plan. A relaxation model of multi-objective reactive power optimization is proposed to improve situations of voltage overlimit and model convergence problems, possibly occurring in control prediction. At last, a multi-objective chaotic difference evolution algorithm based on dynamic search strategy is proposed. This algorithm is able to dynamically allocate search strategies pursuant to proportion of feasible solutions in the group to enhance efficiency of multi-objective models. Simulation results of IEEE 30 bus system indicate that the multi-objective optimization algorithm is better than classic algorithm in terms of optimal solutions, convergence of outer solution and uniformity of solution sets. Real power grid data calculation shows that, comparing to traditional method, the refined control method is able to reduce voltage deviation, and improve convergence of the model.

关 键 词：无功优化 母线负荷预测 精细化控制 多目标优化 松弛模型 

分 类 号：TM721[电气工程—电力系统及自动化]