基于长短时记忆网络-纵横交叉算法的含高比例新能源电力市场日前电价预测

精准的日前电价预测能够协助电力市场参与者做出合理的决策.随着高比例新能源接入电力系统,日前电价的预测难度不断加大.为了提升含高比例新能源电力市场日前电价的预测精度,提出了一种基于奇异谱分析(singular spectrum analysis,SSA)和纵横交叉算法(crisscross optimization,CS...

Day-ahead Electricity Price Forecasting of Electricity Market With High Proportion of New Energy Based on LSTM-CSO Model

Accurate day-ahead price forecasting may help the power market participants make reasonable decisions.With a high proportion of new energy access to the power system,it is increasingly difficult to predict the day-ahead electricity prices.In order to improve the forecasting accuracy of day-ahead prices in the electricity market with high proportion of new energy,this paper proposes a day ahead price forecasting model based on the singular spectrum analysis(SSA)and the crisscross algorithm(CSO)to optimize the long-term and short-term memory network(LSTM).Firstly,the original data is decomposed into the trend series,the periodic series and the residual series by using the SSA.Secondly,the multi-step prediction model of LSTM is established for each of the sub-sequences.Aiming at the problem that the parameters of the full connected output layer of LSTM are apt to fall into local optimum,the secondary LSTM training strategy is proposed.After training the LSTM,the CSO algorithm is used to fine tune the weight coefficient and the bias between the full connected layers.Finally,all the forecast series are superimposed to get the final electricity price forecast value.The experimental results show that the proposed method can effectively improve the forecasting accuracy of day-ahead price.