摘要
风电-光伏机组的大量接入对传统的无功优化模型提出了新的挑战。提出了计及风光出力相关性的配电网多目标无功优化模型,采用粒子群优化神经网络(particle swarm algorithm-BP neural network,PSO-BP)依据过去天气预报和风电出力的历史数据训练得到风电预测出力曲线,利用综合场景概率法生成光伏出力曲线。用斯皮尔曼相关系数将风光之间的出力相关性量化,再考虑风机和光伏机组共同参与无功优化。采用多目标粒子群算法(multi-obj ective particle swarm optimization,MOPSO)求解模型,以改进的IEEE 33节点配电网系统作为仿真样本,求得兼顾网损和电压偏差的Pareto最优解集,从中选择最优方案。算例结果验证了风光出力相关性对无功优化的影响,以及分布式电源接入配电网能有效降低网损和提高节点电压。在实际运行中,各地区的风光出力均满足一定的自然规律,可以以斯皮尔曼相关系数大小为参考依据,实现分布式电源(distributed generation,DG)和静止无功补偿装置(static var compensator,SVC)的协同优化运行,为配电网的安全经济运行保驾护航。
Large-scale grid integration of windphotovoltaic units poses new challenges to traditional reactive power optimization models. In this paper, a multi-objective reactive power optimization model for distribution network considering the correlation between the outputs of wind power and photovoltaics is proposed. particle swarm algorithm-BP neural network(PSO-BP) is used to train the wind power output forecast curve based on the historical data of past weather forecast and wind power output. Integrated scene probability method is used to generate PV output curve. Spearman correlation coefficient is used to quantify the correlation between the outputs of wind power and photovoltaics, and then participation of wind turbine and photovoltaic unit in reactive power optimization is considered. Multi-objective particle swarm optimization(MOPSO) algorithm is used to solve the model. Improved IEEE 33-node distribution network system is used as simulation sample to obtain Pareto optimal solution set considering both network loss and voltage deviation, and the optimal solution is selected. The results of the example verify the influence of the correlation between the outputs of wind power and photovoltaics on reactive power optimization, and the distributed power supply distribution network can effectively reduce network loss and increase node voltage. In actual operation, the wind power and photovoltaic outputs of each region satisfies a certain natural law. Based on Spearman correlation coefficient, the collaborative optimization operation of DG and SVC can be realized to ensure safe and economic operation of the distribution network.
作者
刘梦依
邱晓燕
张志荣
赵长枢
赵有林
张楷
LIU Mengyi;QIU Xiaoyan;ZHANG Zhirong;ZHAO Changshu;ZHAO Youlin;ZHANG Kai(Key Laboratory of Intelligent Electric Power Grid in Sichuan Province(Sichuan University),Chengdu 610065,Sichuan Province,China)
出处
《电网技术》
EI
CSCD
北大核心
2020年第5期1892-1899,共8页
Power System Technology
基金
四川省科技厅重点研发项目(2017FZ0103)。
关键词
分布式电源
PSO-BP神经网络
相关性
多目标无功优化
distributed generation
PSO-BP neural network
correlation
multi-objective reactive power optimization
