Flexible Medium and Long-term Load Curve Prediction Method for Distribution Network Based on Multivariable Grey Genetic Model
The traditional flexible medium to long term load curve prediction method for distribution networks lacks a reasonable and effective allocation of load prediction weights,resulting in low prediction accuracy.Therefore,a multivariable grey genetic model based flexible medium and long-term load curve prediction method for distribution networks is proposed.Train the pro-cessed distribution network load data using the back-propagation through time(BPTT)algorithm to obtain the distribution net-work load characteristics.On the basis of considering the characteristics of load data,a multivariate grey genetic model is used to allocate weights for load curve prediction.According to the predicted weights,cluster processing parameters such as load growth,output the flexible medium and long-term load prediction curve of the distribution network,and thus curvaturally pre-dict the flexible medium and long-term load of the distribution network.The experimental results show that the designed meth-od has a smaller prediction error and higher accuracy for the medium to long-term flexible load of the power grid,meeting the demand for curved prediction of the flexible medium and long-term load of the distribution network.
multivariable grey genetic modeldistribution networkflexible medium and long-term loadcurve predictionBPTT algorithm
万方数据
维普
