Research On Distribution Network Reconfiguration Considering Correlation Between Wind Speeds And Load Demands

Distribution network reconfiguration is a significant part of distribution automationand it is an important way to improve economy and security of distribution network.The penetration of wind power generation in distribution network, as an important formof generation, is gradually increasing with the rapid development of distributedgeneration in recent years. Therefore, developing wind power generation is aninevitable trend of power system in the future.There are some correlations between wind speeds and load demands due to theinfluence of some climate factors. Wind speeds and load demands both change overtime and they are not independent random variables. Therefore, it is of greatsignificance to use joint distribution to describe the correlation between wind speedsand load demands and establish a more accurate mathematical model in the distributionnetwork reconfiguration.In order to overcome the shortcomings of the traditional correlation models, thispaper proposes a method to establish the correlation model between wind speeds andload demands based on Copula theory. The marginal distribution functions of windspeeds and load demands from historical samples are obtained. The parameters ofalternative Copula functions are estimated by the maximum likelihood estimationmethod. According to the evaluation criteria of shortest distance based on Copula theory,an optimal Copula function is selected from these alternative functions to describe thecorrelation structure between wind speeds and load demands. The wind data atSaskatchewan of Canada and time series load demands curve of the IEEE-RTS are usedas samples. The case study shows that this Copula function can simulate wind speedsand load demands accurately and it can well solute the correlation problem betweenthem.According to the influence that distributed generation has brought to thedistribution network power flow and the node forms in the power flow caculation, arandom model of power flow with wind power generation is established based on theempirical distribution function. According to the simulation method of Copula fucuton,a certain amount of wind speeds and load demands sequence are generated by thecorrelation model. An algorithm of stochastic power flow considering the correlationbetween wind speeds and load demands is proposed based on Monte Carlo simulation. The algorithm is tested in the IEEE33and PG&E69systems. The case study shows thatthis algorithm can effectively consider the influence that the correlation between windspeeds and load demands has brought to stochastic power flow in the distributionnetwork.The minimal expectation of distribution network active power loss is used as theobjective function to establish a mathematical model in the distribution networkreconfiguration with wind power generation, considering both randomness andcorrelation of wind speeds and load demands. An improved genetic algorithm is usedfor avoiding the non-feasible solutions in the genetic manipulation. According to thestochastic power flow considering the correlation between wind speeds and loaddemands based on Monte Carlo simulation, an algotithm of distribution networkreconfiguration considering the correlation between wind speeds and load demands isproposed. The algorithm is tested in the IEEE33and PG&E69systems. The case studyshows that this method can effectively decrease the active power loss and cancomprehensively considers the influence that the correlation between wind speeds andload demands has brought to the distribution network reconfiguration.