基于改进Bagging算法与模糊MP-LSTM融合的短期负荷预测模型*
Short-term Load Forecasting Model Based on Improved Bagging Algorithm and Fuzzy MP-LSTM Fusion
-
摘要: 为提高负荷预测精度,提出一种基于改进Bagging算法与模糊最小窥视孔长短期记忆(Min peephole long short-term memory,MP-LSTM)融合的短期负荷预测模型。MP-LSTM模型相较于传统长短期记忆网络(Long short-term memory,LSTM)模型舍弃了输入门和输出门,只保留遗忘门,模型包括一个sigmoid网络层和一个tanh网络层,减少了模型参数,优化了模型结构。通过将温度进行模糊化处理,减小温度波动对负荷的影响。采用改进Bagging算法对MP-LSTM模型集成处理来提高模型预测的精度。以某地区的实际负荷数据进行算例仿真,并与传统LSTM神经网络预测法、MP-LSTM神经网络法和模糊MP-LSTM神经网络法进行对比,仿真结果表明文中所提模型具有较好的预测精度。Abstract: In order to improve the accuracy of load forecasting, a short-term load forecasting model based on the fusion of improved Bagging algorithm and fuzzy minimum peephole long short-term memory (MP-LSTM) is proposed. Compared with the traditional long short-term memory (LSTM) model, the MP-LSTM model discards input gates and output gates, and only retains forgetting gates. The model includes a sigmoid network layer and a tanh network layer, which reduces model parameters and optimizes the model structure. By fuzzing the temperature, the influence of temperature fluctuations on the load is reduced the improved Bagging algorithm is used to integrate the MP-LSTM model to improve the accuracy of model prediction. The actual load data of a certain area is used for simulation, and compared with the traditional LSTM neural network prediction method, MP-LSTM neural network method and fuzzy MP-LSTM neural network method, and simulation results show that the proposed model has better predictions accuracy.