基于RF-GA-SVM的医院集中供暖系统一次侧流量预测模型研究
Research on primary side flow prediction model of hospital central heating systems based on RF-GA-SVM
摘要:
医院集中供暖系统一次侧流量受多种不确定因素影响。为了降低输入空间维度、节约运算成本、提高预测精确度,提出了一种基于随机森林(RF)特征重要性评估-遗传算法(GA)优化支持向量机(SVM)参数算法的预测模型。首先利用RF算法对特征变量实施重要性评估,利用交叉验证法对特征变量进行过滤,构建供暖系统影响因素指标体系,其次利用遗传算法优化支持向量机参数建立回归预测模型(RF-GA-SVM),最后结合某医院集中供暖系统数据进行了实例分析并与RF预测模型、GA-SVM预测模型进行了对比。预测误差分析表明,本文提出的预测模型(RF-GA-SVM)降低了输入空间维度,避免了局部最优,提高了预测精确度。
Abstract:
The primary side flow of the hospital central heating systems is affected by many uncertain factors. In order to reduce the dimension of input space, save the operation cost and improve the prediction accuracy, this paper proposes a prediction model based on the random forest (RF) feature importance evaluation-genetic algorithm (GA) optimization support vector machine (SVM) parameter algorithm. Firstly, the RF algorithm is used to evaluate the importance of characteristic variables, and the cross validation method is used to filter the characteristic variables to build an index system of influencing factors of the heating system. Secondly, the genetic algorithm is used to optimize the parameters of the support vector machine to establish a regression prediction model (RF-GA-SVM). Finally, an example is analysed based on the data of a hospital’s central heating system and compared with the RF prediction model and GA-SVM prediction model. The prediction error analysis shows that the prediction model (RF-GA-SVM) proposed in this paper reduces the dimension of input space, avoids local optimization and improves the prediction accuracy.
Keywords:hospital;buildingenergyconsumption;centralheating;primarysideflow;randomforest(RF);geneticalgorithm(GA);supportvectormachine(SVM)