基于FCM-LSSVM-GSA空调负荷预测  

A prediction for air-conditioning load base on FCM-LSSVM-GSA model

作　　者：赵超 郑守锦 

机构地区：[1]福州大学石油化工学院,福建福州350108

出　　处：《计算机与应用化学》2017年第10期787-795,共9页Computers and Applied Chemistry

基　　金：国家自然科学基金资助项目(6080402;61374133);高校博士点专项科研基金(20133314120004)~~

摘　　要：准确、快速的空调负荷预测是实现空调系统经济运行的基础。为提高空调负荷预测模型的精度以及稳定性,本文提出了一种基于FCM-LSSVM-GSA空调负荷预测方法。根据数据的相似统计分布特征,利用模糊C均值聚类算法(FCM)将历史负荷数据划分成多个簇类,以降低样本数据中相关噪声对建模精度的影响;并根据每个簇类的训练和测试数据集建立相应的最小二乘支持向量机预测模型(LSSVM);通过引入万有引力搜索算法(GSA)优化LSSVM的模型参数,以避免人为选择的盲目性,从而提高模型的预测精度。基于DeST平台模拟数据,将FCM-LSSVM-GSA模型运用于南方某办公大楼的逐时空调负荷预测。通过对比均方根误差(RMSE)和平均绝对百分误差(MAPE),结果表明该模型的预测精度明显优于传统LSSVM模型和简单FCM-LSSVM模型。Air conditioning load forecast is the fundamental of the efficient operation of air conditioning system. In order to improve the accuracy of building air conditioning load prediction, a new predictive model based on FCM and LSSVM was proposed to forecast air conditioning load. According to the similarity statistics, FCM method was employed to divide the historical samples into several clusters to reduce the correlated noise in load data. Then, the LSSVM model of the identified cluster was constructed with the training samples. Furthermore, GSA algorithm was used to select the optimal parameters of LSSVM model to improve its generalization ability and prediction accuracy. Based on the simulated data from the DeST platform, the FCM-LSSVM was applied to predict the hourly air-conditioning load of an office building in South China. The results showed that the prediction efficiency of the FCM-LSSVM-GSM model are better than those of the LSSVM model and FCM-LSSVM model in terms of the root mean square error （RMSE） and the mean absolute percentage error （MAPE）. It provides good basis for optimal control of air conditioning system.

关 键 词：空调负荷预测 最小二乘支持向量机 模糊C均值聚类算法 万有引力搜索算法 

分 类 号：TU831[建筑科学—供热、供燃气、通风及空调工程]