基于CNN-LSTM的短期电力负荷预测研究作者:刘月峰 杨宇慧来源:《科技创新与应用》2020年第01期 新型广告媒介 摘 ;要:针对短期电力负荷预测中电力负荷影响因素提取不准确以及长期依赖信息丢失的问题,提出一种结合一维卷积神经网络CNN与长短期记忆网络LSTM的短期电力负荷预测模型,将卷积神经网络的速度和轻量与长短期记忆网络的顺序敏感性的优势结合起来,考虑历史电力负荷、时间日期、温度对电力负荷的影响,实例结果表明,与其他模型相比,预测误差更小,较好地提高了预测性能。排毒柜
关键词:短期电力负荷预测;卷积神经网络;长短期记忆网络互助系统
木薯干 中图分类号:TM715 ; ; ; ;文献标志码:A ; ; ; ; 文章编号:2095-2945(2020)01-0084-02摆线齿轮
Abstract: In order to solve the problems of inaccurate extraction of power load influencing factors and long-term dependence on information loss in short-term power load prediction, a short-term power load prediction model combining a convolutional neural network (CNN) and a long-term and short-term memory network (LSTM) is proposed. Speed and light weight are combined with the order sensitivity of long-term and short-term memory networks, considering the effects of historical power load, time and date, and temperature on the power load. The results of the examples show that compared with other models, the prediction error is smaller and the prediction performance is better.
剖分式油封