基于CEEMDAN-IGWO-DELM的现货市场电价混合预测算法  

The Mixed Forecasting Algorithm for Spot Market Electricity Prices Based on CEEMDAN-IGWO-DELM

作　　者：恩格贝 李玉璐 张岩 EN Gebei;LI Yulu;ZHANG Yan(School of Economic and Management,North China Electric Power University,Beijing 102206,China)

机构地区：[1]华北电力大学经济与管理学院,北京102206

出　　处：《河北电力技术》2024年第4期1-9,共9页Hebei Electric Power

摘　　要：准确预测现货市场电价对于保护电力市场参与者的利益具有重要意义。目前,大量的可再生能源参与现货市场交易,使得预测现货市场电价变得极具挑战性。为此,本文提出了一种基于分解-优化-集成的混合电价预测模型。首先,使用完全集合经验模态分解方法将原始电价时间序列进行分解;然后,采用Cat混沌映射策略、停滞检测策略和高斯Levy扰动策略克服灰狼优化算法陷入局部最优问题,提高了种群多样性;其次,利用改进的灰狼优化算法对深度极限状态机隐层参数进行优化并构造现货电价预测模型,最后,通过仿真实验对所提方法进行分析和验证。结果表明所提出的方法有效地改善了预测模型的精度。Accurately predicting spot market electricity prices is of significant importance for safeguarding the interests of participants in the electricity market.Currently,a large amount of renewable energy participates in spot market trading,making it extremely challenging to predict spot market electricity prices.To address this,this paper proposes a mixed electricity price prediction model based on decomposition-optimization-integration.Firstly,the original electricity price time series is decomposed using the complete ensemble empirical mode decomposition method.Then,a combination of Cat chaotic mapping strategy,stagnation detection strategy,and Gaussian-Levy perturbation strategy is employed to overcome the problem of the grey wolf optimization algorithm falling into local optima,thereby enhancing population diversity.Next,the improved grey wolf optimization algorithm is used to optimize the hidden layer parameters of the deep extreme learning machine and construct the spot electricity price prediction model.Finally,the proposed method is analyzed and validated by simulation experiments.The results indicate that the proposed approach effectively enhances the accuracy of the prediction model.

关 键 词：现货市场电价预测 深度极限学习机 灰狼优化算法 高斯-Levy变异策略 

分 类 号：TM73[电气工程—电力系统及自动化]