基于经验模态分解-粒子群优化-长短期记忆的车间电力能耗预测方法  

Prediction method of power consumption in workshop based on EMD-PSO-LSTM

作　　者：孙宁可 王艳 纪志成 Sun Ningke;Wang Yan;Ji Zhicheng(Engineering Research Center of Internet of Things Technology Applications Ministry of Education,Jiangnan University,Wuxi 214122,China)

机构地区：[1]江南大学物联网技术应用教育部工程研究中心,江苏无锡214122

出　　处：《南京理工大学学报》2023年第2期238-244,253,共8页Journal of Nanjing University of Science and Technology

基　　金：国家自然科学基金(61973138);国家重点研发计划(2018YFB1701903)。

摘　　要：针对传统能耗预测方法不能充分提取数据特征并利用神经网络的学习、预测能力,提出了一种基于经验模态分解-粒子群优化-长短期记忆(EMD-PSO-LSTM)的电力能耗预测模型。该模型首先采用经验模态分解算法将时间序列数据分解为多个本征模函数分量和趋势分量,然后对每个分量分别建立长短期记忆网络进行预测。各个长短期记忆网络的参数则由粒子群算法分别进行最优化求解,最后将所有分量的预测结果进行叠加得到最终的能耗预测结果。将预测结果与实际能耗数据进行对比分析来验证所提模型的预测性能。试验结果表明,该方法能够对电力能耗数据进行合理预测,预测精度较高。In view of the fact that the traditional energy consumption prediction methods can not fully extract data features and make use of the learning and prediction ability of neural network,a power consumption prediction model based on EMD-PSO-LSTM(Empirical mode decomposition-particle swarm optimization-long-short term memory)is proposed.Firstly,the empirical mode decomposition algorithm is used to decompose the time series data into several eigenmode function components and trend components,and the long short-term memory network is established for each component to predict.The parameters of each longshot-term memory network are optimized by particle swarm optimization algorithm.Finally,the prediction results of all components are superimposed to obtain the final energy consumption prediction result.The prediction results are compared with the actual energy consumption data to verify the prediction performance of the proposed model.Experimental results show that this method can reasonably predict the power consumption data,and the prediction accuracy is high.

关 键 词：能耗预测 经验模态分解 粒子群优化 参数寻优 长短期记忆 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]