| 基于注意力机制优化的LSTM河流溶解氧预测模型研究 |
| |
| 引用本文： | 周泉,胡轩铭,王东昆,张武才,陈中颖,王金鹏,汪澎洋,任秀文.基于注意力机制优化的LSTM河流溶解氧预测模型研究[J].环境科学研究,2023,36(6):1135-1146. |
| |
| 作者姓名： | 周泉 胡轩铭 王东昆 张武才 陈中颖 王金鹏 汪澎洋 任秀文 |
| |
| 作者单位： | 1.生态环境部华南环境科学研究所，国家环境保护水环境模拟与污染控制重点实验室，广东 广州 510530 |
| |
| 基金项目： | 国家自然科学基金项目(No.42207097)；中央级公益性科研院所基本科研业务专项(No.PM-zx703-202104-074)；广东省自然科学基金项目(No.2019A1515012123) |
| |
| 摘 要： | 溶解氧(DO)是水体中的重要水质指标，构建数据驱动模型，实现对溶解氧的准确预测，将为水环境管理提供科学有效的技术手段. 考虑到溶解氧序列数据非线性强、非平稳性突出的特点，提出一种基于双阶段注意力权重优化机制的长短时记忆网络(long short-term memory, LSTM)的河流溶解氧预测模型(DAIW-LSTM模型)，该模型的编码器包含双阶段权重优化的空间注意力机制，而解码器包含双阶段权重优化的时间注意力机制. 将该模型应用于流溪河流域白云李溪坝、流溪河山庄、从化街口等水质监测站溶解氧日均值预测的研究，开展了该模型与DA-LSTM、LSTM、Bi-LSTM等基线模型的预测效果对比分析，探讨了特征权重优化机制及上游站点水质数据输入对模型预测性能的影响. 结果表明:①通过与基线模型的预测效果对比，验证了DAIW-LSTM模型的精准性，其对白云李溪坝站溶解氧预测的对称平均绝对百分比误差(SMAPE)、平均绝对误差(MAE)、均方误差(MSE)分别为0.075、0.611、0.712，在所有模型中最优. ②对于新的注意力权重优化机制，第二阶段会对第一阶段的初步权重进行优化修正；针对pH、电导率、水温、气温等影响溶解氧预测的重要特征，DAIW-LSTM模型会自适应调整其权重在时间序列上的分布，从而提高该模型的预测精度. ③加入上游水质特征的输入影响，通过9个组合试验对比可知，DAIW-LSTM模型仍然为表现最佳的模型，该系列组合试验也证明上游站点及其特征变量选取的重要性. 研究显示，注意力权重优化机制的引入使得该模型相较其他基线模型展现出更好的适用性和精准性，可为地表水水质预测研究提供新思路.
|
| 关 键 词： | 注意力机制 时间序列预测 溶解氧预测 LSTM模型 |
| 收稿时间： | 2022-09-17 |
Prediction of Dissolved Oxygen in Rivers Based on LSTM Model with Improved Attention Mechanism |
| |
| Affiliation: | 1.South China Institute of Environmental Sciences, Ministry of Ecology and Environment, State Environmental Protection Key Laboratory of Water Environmental Simulation and Pollution Control, Guangzhou 510530, China2.State Key Laboratory of Internet of Things for Smart City, University of Macau, Macau S.A.R., 999078, China |
| |
| Abstract: | Dissolved oxygen (DO) is a key index of the aquatic environment. A data-driven model for accurately predicting DO will provide scientific and effective technical methods for water environment management. Considering the strong nonlinearity and non-stationarity of river DO time series, a novel river DO concentrations prediction model based on a LSTM method with improved weights dual-stage attention mechanism (DAIW-LSTM Model) was proposed. The model uses spatial attention in the encoder and temporal attention in the decoder, and both encoder and decoder contain a new mechanism of weight optimization in two stages. The model was used to predict the daily average DO at Baiyunlixiba monitoring station, Liuxiheshanzhuang monitoring station and Conghuajiekou monitoring station in the Liuxihe River Basin. A comparative analysis among different baseline models (DA-LSTM, LSTM and Bi-LSTM) was carried out, and the effects of feature weight optimization mechanism and the upstream feature variables input were discussed. The results showed that: (1) Comparing with other baseline models, the accuracy of the proposed DAIW-LSTM model was verified. The SMAPE, MAE and MSE predicted by the DAIW-LSTM model at Baiyunlixiba station were 0.075, 0.611 and 0.712 respectively, which were the best of all models. (2) The second stage could optimize and correct the initial weights of the first stage in the proposed attention weight optimization mechanism. Since the important features such as pH, conductivity, water temperature, and air temperature, were adaptively adjusted in the time series, the prediction accuracy of the proposed DAIW-LSTM model could be improved. (3) Further 9 combination tests with the input of upstream characteristics showed that the performance of the proposed DAIW-LSTM model was still best, it also proved that the importance of upstream stations and feature variables selection. The research shows that the attention weight optimization mechanism makes the model exhibit better applicability and accuracy than other baseline models, which can provide new ideas for surface water quality prediction. |
| |
| Keywords: | |
|
| 点击此处可从《环境科学研究》浏览原始摘要信息 |
|
点击此处可从《环境科学研究》下载全文 |
|
