摘要
热负荷是智慧供热系统中的重要参数,能够准确地预测负荷需求,并进行精准调控是智慧热网中实现能源节约和可持续运行的重要举措。该文依据新疆某换热公司2023年度供热数据,选取关联度高的气象、参数特征量作为输入,构建基于PSO-SVM算法的区域供热模型,同时为验证该模型的预测精度,将所提出的PSO-SVM算法模型与基于SVM模型、极限学习机ELM模型、BP神经网络模型的预测方法开展对比实验,并采用3种评价函数验证。研究表明,所构建的PSO-SVM算法模型预测精度显著提升。
Heat load is an important parameter in intelligent heating system.Accurate prediction of load demand and accurate regulation are important measures to achieve energy conservation and sustainable operation in intelligent heat supply network.Based on the heat supply data of a heat exchange company in Xinjiang in 2023,this paper selects highly correlated meteorological and parameter characteristics as inputs to build a district heating model based on the PSO-SVM algorithm.Meanwhile,in order to verify the prediction accuracy of the model,The proposed PSO-SVM algorithm is compared with ELM model and BP neural network model prediction method in three evaluation functions.The results show that the prediction accuracy of the constructed PSO-SVM algorithm model is significantly improved.
出处
《科技创新与应用》
2025年第8期73-76,共4页
Technology Innovation and Application
基金
第二批自治区产学合作协同育人项目(2023210042,2023210046)
中央引导地方科技发展专项资金项目资助(ZYYD2023B19)。
作者简介
第一作者:韩英杰(1993-),男,工学硕士,助教。研究方向为机械设计与自动化控制。
