基于自适应时序分解的空气污染物浓度预测

Prediction of Air Pollutant Concentration Based on Adaptive Time Series Decomposition

中文关键词:  空气污染物浓度预测  自适应完整集成经验模态分解  排列熵  门控循环单元  神经网络

英文关键词:air pollutant concentration prediction  CEEMDAN  PE  GRU  neural networks

中文摘要:

      为准确、有效地预测空气污染物浓度，建立了基于自适应完整集成经验模态分解（CEEMDAN）和排列熵（PE）的门控循环单元（GRU）空气污染物预测模型。首先利用CEEMDAN算法对非线性信号的自适应分解能力将原始序列分解为一组不同频率、复杂度的固有模态函数（IMF）和一个残差分量（REC），其次根据PE算法将复杂度相近的IMF分量和REC一起进行重新组合，最后将重组后的子序列分别使用GRU模型进行预测，并将子序列预测结果相加得到最终预测结果。实验结果表明，基于CEEMDAN-PE-GRU模型预测的误差明显低于其他模型，验证了该模型对空气污染物浓度预测的有效性。

英文摘要:

      In order to accurately and effectively predict the concentration of air pollutants,a gated recirculation unit (GRU) air pollutant prediction model based complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) and permutation entropy (PE) is established.Firstly,the CEEMDAN decomposition algorithm is used to decompose the nonlinear signal adaptively,and the original sequence is decomposed into a set of intrinsic mode functions (IMF) of different frequencies and complexity and a residual difference component (REC),and then according to the PE algorithm,the IMF components and residual components with similar complexity are recombined.Finally,the reorganized sub-sequences are respectively predicted by using the GRU model,and the sub-sequence prediction results are added to obtain the final prediction result.The experimental results show that the prediction accuracy of the CEEMDAN-PE-GRU model is significantly higher than other models,which verifies the effectiveness of the model for predicting the concentration of air pollutants.