摘要
空气中PM2.5浓度问题越来越受到各界的关注。根据PM2.5浓度数据的特征,首先选择ARIMA预测模型进行浓度预测;考虑到BP神经网络易陷入局部最小,而遗传算法具有全局搜索的能力,给出了遗传算法优化的BP神经网络预测模型;为了进一步提高预测精度,引入IOWGA算子,将ARIMA预测模型与遗传算法优化的BP神经网络预测模型相组合,给出了基于IOWGA算子的组合预测模型;最后经过实例仿真分析验证了模型的可行性和有效性,为PM2.5浓度预测提供基础资料。
The problem of PM2.5 concentration in air is receiving more and more attention. First,according to the characteristics of PM2.5 concentration data,the ARIMA prediction model was used to forecast the PM2.5 concentration. Then,taking into account of the BP neural network easy to fall into the local minimum whereas the genetic algorithm has the capability of global search,the BP neural network model optimized by genetic algorithm was established. In order to further improve the prediction accuracy,through introducing IOWGA operator the ARIMA prediction model was combined with the BP neural network optimized by the genetic algorithm to form an IOWGA operator based prediction model. Finally,the feasibility and effectiveness of the combined model were verified by simulations of a practical case. The use of the prediction model provides basic references for the prediction of PM2.5 concentration.
作者
王镱嬴
刘洪
WANG Yiying;LIU Hong(School of Science,University of Science and Technology Liaoning,Anshan 114051,China)
出处
《辽宁科技大学学报》
CAS
2018年第1期75-80,共6页
Journal of University of Science and Technology Liaoning
