基于支持向量机及粒子群算法的空调负荷预测方法研究及应用

Research and application on forecasting methods of air conditioning load based on both the support vector machine and particle swarm optimization

以空调负荷的预测方法为研究对象,通过预测建筑物的空调负荷,及时有效地调控系统,达到降低建筑能耗的目的。基于最小二乘支持向量机(LS-SVM)提出预测方法,并采用粒子群算法(PSO)实现优化计算,围绕济南及成都地区不同类型建筑物的空调负荷进行预测,确定影响空调负荷的各个因素并在计算过程中不断调整以简化预测过程。采用Matlab语言进行编程计算和分析,通过不断调整输入因素及部分PSO参数以降低预测误差。基于支持向量机及粒子群算法的空调负荷预测模型性能较好,预测精度高。研究结果可为暖通空调系统的负荷预测提供参考,揭示空调负荷的变化特点以实现节能调控的目的。

Taking the prediction method of air conditioning load as the research object,the system can be timely and effective regulated to achieve the purpose of reducing building energy consumption by predicting the air conditioning load of the building.The prediction method is proposed based on the least squares support vector machine(LS-SVM),and the particle swarm optimization(PSO)is employed to realize the optimization calculation.Aiming at the air conditioning load of different types of buildings in Jinan and Chengdu,the factors affecting the air conditioning load are determined and constantly adjusted during the calculation to simplify the prediction process.The Matlab language is used to conduct both programming calculation and analysis,and both the input factors and the some PSO parameters are continuously adjusted to reduce the prediction error.The air conditioning load's prediction model based on both the support vector machine and the particle swarm algorithm has good performance and high prediction accuracy.The results can provide a reference for the load prediction of HVAC system,and reveal the changing characteristics of air conditioning load to realize the purpose of energy saving regulation.