摘要
提出的负荷预测模型是一种基于时间序列并带有外部输入的自回归模型(ARX模型),其中的系数通过递推式的最小二乘法回归估计。研究的负荷预测方法考虑分析了影响热负荷的内外因素,这样可以使预测效果较好。研究的递推最小二乘回归方法对异常值具有鲁棒性,适用于快速自适应系数估计。研究还从对某住宅进行集中供热管理的供热公司获取供热运行数据进行模型验证。结果表明,基于外部输入的自回归模型具有良好的预测精度,为供热系统调控中重要的前期工作提供了依据,由此能够对供热的调节方式进行指引,提升用户体验感。
The load prediction model proposed herein is a self-regression model(ARX model)based on a time series and has an external input,wherein the coefficient is estimated by the minimum multiplier regression of the push.The load prediction method of this study considers the internal and external factors affecting the heating load,so that the predictive effect can be better.The recursive least squares regression method of this study has robustness for the exception value and is suitable for rapid adaptive coefficient estimation.This study also got the model verification from the heating company of a centralized heating company for a residential heating.The results show that the self-return model based on the external input has good prediction accuracy,providing a basis for important pre-work in the control of the heating system,whereby the adjustment method of heating can be guided to enhance the user experience.
作者
李明柱
陈灏馨
LI Mingzhu;CHEN Haoxin(Jilin Jianzhu University Municipal and Environment Engineering College)
出处
《上海节能》
2022年第5期554-560,共7页
Shanghai Energy Saving
基金
吉林省教育厅“十三五”科学技术项目“适应性人体热舒适架构下公共建筑空调系统自适应节能控制算法与动态热响应机理研究”(JJKH20200282KJ)。
关键词
热负荷预测
ARX模型
参数辨识
递推最小二乘法
Heating Load Prediction
ARX Model
Parameter Identification
Recursive Least Squares
