College of Management, Xi’an University of Science and Technology

为了更精准地预测未来能源结构调整方向及成效,选用ISSA-LSTM组合预测模型对中国2023—2030年可再生能源的电力需求进行预测。首先,利用Circle混沌映射改进麻雀搜索算法(SSA)以提高搜索能力以及种群多样性;然后引入长短期记忆神经网络(LSTM)以有效捕捉可再生能源电力需求随机波动性和时序性;最后,通过ISSA-LSTM模型预测长期可再生能源的电力需求,验证测试集数据,并与其他传统模型进行对比。结果表明:ISSA-LSTM模型预测结果能够满足对可再生能源电力需求预测的精度要求;在未来2023—2030年可再生能源电力需求稳定,波动幅度不大,可达到全国用电量的1/3;利用Circle混沌映射改进策略能有效提升SSA寻优能力。与PSO算法相比,SSA算法寻找LSTM超参数最优解的能力更优,ISSA-LSTM模型预测可再生能源电力需求精度更高。

More accurate prediction of renewable energy electricity demand is an important indicator for predicting the direction and effectiveness of future energy structure adjustments. The ISSA-LSTM com- bined forecasting model was used to predict the electricity demand for renewable energy in China from 2023 to 2030. Firstly, the Sparrow Search Algorithm( SSA) was improved using Circle Chaos Mapping to enhance search capability and population diversity. Then the long Short Term Memory Neural Network ( LSTM) was introduced to effectively capture the random fluctuations and temporal characteristics of renewable energy electricity demand. Finally, the ISSA-LSTM model was used to predict long-term re- newable energy electricity demand, to validate the test set data, and to compare it with other traditional models. The results show that the ISSA-LSTM model can meet the accuracy requirements for predicting renewable energy electricity demand; In the future from 2023 to 2030, the demand for renewable energy electricity will remain stable and not fluctuate significantly, reaching one-third of the national electricity consumption; The use of Circle chaotic mapping to improve the strategy can effectively enhance the op- timization ability of SSA. Compared with PSO algorithm, SSA algorithm has better ability to find the op- timal solution of LSTM hyperparameters, and the ISSA-LSTM model has higher accuracy in predicting renewable energy electricity demand.

国家自然科学基金项目(71704140);西安科技大学哲学社会科学繁荣项目(2018SY03);陕西省科技厅软科学研究计划项目(2019KRM016)

闫晓霞,刘娴.基于ISSA-LSTM模型的可再生能源电力需求预测[J].西安科技大学学报,2024,44():604-614.

3YAN Xiaoxia, LIU Xian. Renewable energy electricity demand prediction based on ISSA-LSTM model[ J] . Journal of Xi’an Universi-ty of Science and Technology, 2024, 44() : 604-614.3