北华大学学报（自然科学版）2024，Vol.25Issue(4)：545-551,7.DOI:10.11713/j.issn.1009-4822.2024.04.022
基于蒙特卡罗法优化GRU神经网络的热电联产负荷预测
Optimization of GRU Neural Network for Cogeneration Load Prediction Based on Monte Carlo Algorithm
摘要
Abstract
Load forecasting is the premise of improving energy utilization and optimal scheduling of cogeneration system,and the prediction accuracy greatly affects the reliability of system operation and the cost of thermal power plant.With the widespread use of cogeneration units,it is difficult for a single load forecast to accurately reflect the coupling characteristics between cogeneration loads,which has a certain impact on the accurate operation of cogeneration systems.An optimized GRU neural network model is proposed,in which two gated recurrent unit neural networks are connected in parallel,processed through the corresponding dropout layer and fully connected layer,and finally the results are combined.At the same time,the Monte Carlo algorithm is combined to correct the error and improve the accuracy of the prediction model.The feasibility of the proposed prediction model is verified by using actual thermal power plant data.The results show that the proposed model can fully learn the coupling characteristics between thermoelectric loads and improve the accuracy of load prediction.关键词
负荷预测/热电联产/热电耦合/GRU神经网络/蒙特卡罗算法Key words
load prediction/cogeneration/cogeneration coupling/GRU neural network/Monte Carlo algorithm分类
信息技术与安全科学引用本文复制引用
张奇,王禄,邢吉生..基于蒙特卡罗法优化GRU神经网络的热电联产负荷预测[J].北华大学学报（自然科学版）,2024,25(4):545-551,7.基金项目
吉林省科技发展计划重点研发项目(20210203109SF). （20210203109SF）
