Abstract:

To achieve accurate prediction of PM_2.5 concentrations during heavy haze pollution events, the model of deep feature fusion (long short-term memory and multivariate linear regression, LSTM-MLR) was proposed in this paper to predict PM_2.5 concentrations in 3h, 6h, 12h and 24h of Xi’an and to explain the relationship between the main meteorological factors precursors and haze concentration. In the proposed model, a series of long short-term memory (LSTM) with different hyper-parameter were used to extract the deep features of PM_2.5 precursors and meteorological time series, and a full connect layer was used to adjust the dimensions of LSTM outputs. Then, a feature fusion model with the structure of multivariable linear regression (MLR) was used to obtain the predicted PM_2.5 concentrations. Data from January 2015 to March 2020 in Xi’an were used to determine parameters of the proposed model, and model performances for PM_2.5 concentration prediction in 3h, 6h, 12h and 24h were also evaluated. Results showed that LSTM-MLR, which was based on deep feature fusion of time series, can predict heavy haze pollution samples correctly with the accuracy of 94.12%, 85.29%, 77.57% and 51.10% in 3h, 6h, 12h and 24h, respectively. The true positive rate (TPR) of the proposed model outperforms random forest (RF), support vector regression (SVR), MLR, LSTM_PM_2.5 (PM_2.5 used only), multivariable LSTM (M_LSTM) and long short-term memory-random forest (LSTM-RF) which had the same input as LSTM-MLR. The fusion coefficients showed that the influence of current PM_2.5 concentrations on that in the future decreased from 80.89% (t + 3) to 16.34% (t + 24), and the influence of precursor concentration increased from 5.23% (t + 3) to 29.43% (t + 24), which illustrated the importance of time for taking emergency measures to reduce the peak value and decrease the pollution duration.

Key words: heavy haze pollution event, LSTM-MLR model, prediction, multiscale, neural networks

摘要：

为准确预测雾霾重污染演变过程，本文提出深度时间序列特征融合模型（long short-term memory and multivariate linear regression，LSTM-MLR）并对西安市PM_2.5浓度进行了临近预测。该模型利用不同超参数长短期记忆网络（long short-term memory，LSTM）提取PM_2.5前体和气象因素时间序列中的深度特征；采用多元线性回归（multivariate linear regression，MLR）形式融合LSTM单元输出的深度时间序列特征，最终输出PM_2.5浓度预测值。为评估模型性能，采用西安市2015年1月至2020年3月采暖季数据进行建模并计算未来3h、6h、12h、24h的PM_2.5浓度预测精度。结果表明：LSTM-MLR模型对雾霾严重污染样本的准确预测率分别为94.12%、85.29%、77.57%和51.10%，显著高于随机森林（random forest，RF）、支持向量回归（support vector regression，SVR）、MLR、单变量LSTM（LSTM_PM_2.5）、多变量LSTM（M_LSTM）和LSTM-RF（long short-term memory and random forest）；融合系数显示当前PM_2.5浓度对未来PM_2.5浓度的影响随预测步长的增加从80.89%（t+3）急剧降低至16.34%（t+24），前体浓度影响力从5.23%（t+3）上升至29.43%（t+24），说明提前控制前体物排放强度对雾霾重污染事件消峰降速效果具有显著影响。

关键词: 雾霾重污染, LSTM-MLR模型, 预测, 多尺度, 神经网络