基于深度学习的配电网智能检修模型及其求解方法

Intelligent Maintenance Model of Distribution Network Based on Deep Learning and Its Solution Method

随着风、光等可再生能源发电大规模地并入配电网,配电网规模日渐庞大复杂,使传统检修方案无法满足配电网经济性、可靠性二者共赢的要求。对此,在传统配电网检修方案的基础上,建立了配电网检修的经济性、可靠性模型;采用博弈论中的帕累托纳什均衡原理建立了配电网经济性、可靠性均衡优化模型。利用人工智能中的深度学习方法对此模型进行求解。通过IEEE RBTS BUS5仿真算例表明,与传统方法相比,该模型能够优化检修时间、负荷转移通路、网损等方面,提高配电网供电的可靠性和检修的经济性。

With the large-scale integration of wind,light,storage and other renewable energy power generation into the distribution network,the scale of the distribution network is becoming increasingly large and complex,which makes the traditional maintenance scheme unable to meet the requirements of economic and reliability win-win of distribution network.In this regard,on the basis of traditional distribution network maintenance scheme,the economic and reliabi-lity model of distribution network maintenance is established;the Pareto Nash equilibrium principle in game theory is used to establish the optimization model of distribution network economy and reliability.The model is solved based on the deep learning method in artificial intelligence.The simulation results of IEEE RBTS Bus5 show that,compared with the traditional methods,the model can optimize the maintenance time,load transfer path,minimum network loss,and improve the power supply reliability and maintenance economy of distribution network.