基于CEEMD-LSTM的短期天然气负荷预测模型

1.西南石油大学石油与天然气工程学院;2.长庆油田分公司油气营销事业部;3.长庆油田长北作业分公司

天然气负荷预测;互补集合经验模态分解;长短期记忆神经网络;超参数优化

Short-term natural gas load forecasting model based on CEEMD-LSTM
YAN Ke1,PENG Xingyu1,LIU Xiaokun2,ZHANG Kun3,ZHANG Yuchun3,MU Weiwei3,LI Fusheng3

1.Petroleum Engineering School, Southwest Petroleum University; 2.Oil and Gas Marketing Division of PetroChina Changqing Oilfield Company; 3.Changbei Operation Branch of PetroChina Changqing Oilfield Company

natural gas load forecasting, complementary ensemble empirical mode decomposition (CEEMD), long short-term memory (LSTM) network, hyperparameter optimization

DOI: 10.6047/j.issn.1000-8241.2024.03.012

备注

【目的】管网公司为保障稳定供气,需对下游用户的短期天然气负荷进行预测,但传统负荷预测方法存在拟合效果差、预测精度低等问题。为了提高短期天然气负荷预测的精度,在此提出一种基于互补集合经验模态分解(Complementary Ensemble Empirical Mode Decomposition, CEEMD)与长短期记忆神经(Long Short-Term Memory, LSTM)网络的短期天然气负荷组合预测模型。【方法】为充分挖掘负荷序列的内部隐藏特征信息,避免不同分量特征及额外噪声的相互干扰,通过CEEMD分解将原始负荷序列分解为有限个本征模态函数(Intrinsic Mode Function, IMF)分量,随后将不同IMF分量输入LSTM模型进行多步迭代预测,并基于贝叶斯算法对LSTM模型的超参数进行优化以提升学习效果和预测精度,最后将各分量的预测结果叠加重构得到最终预测结果。【结果】对比不同组合模型预测结果,CEEMD-LSTM预测值与真实值误差更小,预测精度更高,贝叶斯调参进一步提升了预测精度,但耗费时间更长。【结论】相比其他预测模型,CEEMD-LSTM组合预测模型能够有效提取负荷序列的时序信息并消除非线性因素的影响,在抑制模态混叠的同时减小了重构误差,提升了预测精度,该方法可为天然气管网的调度管理和运行优化提供参考。(图6表5,参[20]
[Objective] Pipeline network companies need to forecast the short-term natural gas loads of downstream users to ensure a stable gas supply. However, traditional load forecasting methods have various problems such as poor fitting effect and low forecasting accuracy. A combined short-term natural gas load forecasting model based on the Complementary Ensemble of Empirical Mode Decomposition (CEEMD) and the Long Short-Term Memory (LSTM) network is proposed to improve the accuracy of short-term natural gas load forecasting. [Methods] To fully explore the internal hidden feature information of the load series and to avoid the mutual interference between different component features and additional noise, the original load series was decomposed into a finite number of Intrinsic Mode Function (IMF) components through CEEMD. Different IMF components were input into the LSTM model for iterative multi-step forecasting, the hyperparameters of the LSTM model were optimized based on the Bayesian algorithm to improve the learning effect and the forecasting accuracy, and the forecasting results of individual components were superimposed and reconstructed to obtain the final forecasting results. [Results] The comparison of forecasting results obtained by different combined models indicates that the errors in the values forecast by CEEMD-LSTM based on the real values are smaller, proving that the forecasting accuracy is higher. The Bayesian hyperparameter tuning further improves the forecasting accuracy but takes longer forecasting time. [Conclusion] Compared with other forecasting models, the combined CEEMD-LSTM forecasting model can effectively extract the time order information of load series, eliminate the influence of nonlinear factors, and reduce reconstruction errors while suppressing modal superposition, thus improving forecasting accuracy. This method can provide a reference for the scheduling management and operation optimization of natural gas pipeline networks. (6 Figures, 5 Tables, 20 References)
·