基于PSO-SVR航站楼CO2浓度时间序列预测

Time Series Prediction of CO2 Concentration in Terminal Based on PSO-SVR

针对航站楼CO2浓度时间序列预测中存在历史样本少和高度非线性的特点,该文运用基于支持向量回归SVR的预测建模方法,并利用粒子群优化PSO算法对模型进行参数寻优,对基于机场大气环境数据采集系统采集的航站楼CO2浓度数据建立PSO-SVR预测模型,并从均方误差和平方相关系数2个指标与基于网格搜索法、遗传算法优化的SVR预测模型以及BP神经网络预测模型进行对比。试验结果表明,PSO-SVR模型预测精度高、泛化能力强,适合于航站楼CO2浓度时间序列的预测。

Due to the lack of historical data and high nonlinearity,time series prediction for the CO2 concentration in the terminal has become a hard issue. In this paper,a prediction method based on support vector regression（SVR） is proposed and the particle swarm optimization（PSO） approach is adopted to optimize the prediction model parameters. The PSO-SVR prediction model is established through the CO2 concentration data collected by the airport atmospheric environment data acquisition system. The mean square error and the square correlation coefficient is used to compare the performance of the proposed approach with that of the approach optimized by the grid search method ,the method optimized by the genetic algorithm and the BP neural network algorithm. Experimental results have illustrated that the prediction model established by the PSO-SVR method has better performance in generalization and is more ap- propriate for the time series prediction of the CO2 concentration in the terminal.