School of Instrument Science and Engineering, Southeast University, Nanjing, 210096, China.
Nanjing South New Town Development and Construction Group Co., Ltd, Nanjing, 210096, China.
Environ Sci Pollut Res Int. 2022 Oct;29(46):70323-70339. doi: 10.1007/s11356-022-20744-7. Epub 2022 May 19.
PM concentration is an important indicator to measure the concentration of air pollutants, and it is of important social significance and application value to realize accurate prediction of PM concentration. To further improve the accuracy of PM concentration prediction, this paper proposes a hybrid machine learning model (WD-SA-LSTM-BP model) based on simulated annealing (SA) optimization and wavelet decomposition. Firstly, the wavelet decomposition algorithm was used to realize the multiscale decomposition and single-branch reconstruction of PM concentrations to weaken the prediction error caused by time series data. Secondly, the SA optimization method was used to optimize the super-parameters of each machine learning model under each reconstructed component, so as to solve the problem that it is difficult to determine the parameters of machine learning model. Thirdly, the optimized machine learning model was used to predict the PM concentration, and the error value was calculated from the actual measured value. Then, the optimized machine learning model was used to predict the error value. Finally, the predicted error value was added to the predicted PM concentration to obtain the final predicted PM concentration. The study is experimentally validated based on daily PM data collected from Nanjing air quality monitoring stations. The experimental results showed that the RMSE and MAE values of the constructed WD-SA-LSTM-BP model were 5.26 and 3.72, respectively, which were superior to those of the WD-LSTM and LSTM models, indicating that the hybrid machine learning model based on SA optimization and wavelet decomposition could better predict the PM concentration.
PM 浓度是衡量空气污染物浓度的一个重要指标，实现 PM 浓度的准确预测具有重要的社会意义和应用价值。为了进一步提高 PM 浓度预测的准确性，本文提出了一种基于模拟退火（SA）优化和小波分解的混合机器学习模型（WD-SA-LSTM-BP 模型）。首先，采用小波分解算法对 PM 浓度进行多尺度分解和单支重构，以削弱时间序列数据引起的预测误差。其次，采用 SA 优化方法对每个重构分量下的各个机器学习模型的超参数进行优化，从而解决机器学习模型参数难以确定的问题。然后，利用优化后的机器学习模型对 PM 浓度进行预测，并计算实际测量值与预测值之间的误差值。接着，利用优化后的机器学习模型对误差值进行预测。最后，将预测的误差值添加到预测的 PM 浓度中，得到最终的预测 PM 浓度。本研究基于从南京空气质量监测站收集的每日 PM 数据进行实验验证。实验结果表明，所构建的 WD-SA-LSTM-BP 模型的 RMSE 和 MAE 值分别为 5.26 和 3.72，优于 WD-LSTM 和 LSTM 模型，表明基于 SA 优化和小波分解的混合机器学习模型可以更好地预测 PM 浓度。
