Scholarlink

<p>Water quality monitoring and management play an important role in protecting river ecology and river governance. To meet the demand for water quality prediction and capture the temporal and nonlinear characteristics of water quality data, an optimized bidirectional long short-term memory network model is developed, which has three sub models: short interval, medium interval, and long interval, which correspond to the analysis of different time series data. Then, a dual-stage attention mechanism is applied to optimize key feature extraction from the data. The study validated the proposed model through collected river water quality monitoring data. The Root Mean Squared error (RMSE) on the testing set was 0.0921, and the Mean Squared Error (MSE) was 0.0094. In a low ammonia nitrogen concentration environment, the model had an mean absolute error of 0.0689 and a MSE of 0.0095. In high ammonia nitrogen concentration environments, the mean absolute error was 0.1939, and the MSE was 0.0628. Therefore, the designed method helps to capture potential long-term and short-term dependencies, thereby making the model more stable in predicting water quality changes.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Analysis of River Management Method Based on Improved Bidirectional Long Short-Term Memory Network for Water Quality Prediction

  • 期刊论文

  • 发表日期：

    2025年3月8日

查看全文
  • Zhongya Wang,
  • Shuang Liu

摘要

Water quality monitoring and management play an important role in protecting river ecology and river governance. To meet the demand for water quality prediction and capture the temporal and nonlinear characteristics of water quality data, an optimized bidirectional long short-term memory network model is developed, which has three sub models: short interval, medium interval, and long interval, which correspond to the analysis of different time series data. Then, a dual-stage attention mechanism is applied to optimize key feature extraction from the data. The study validated the proposed model through collected river water quality monitoring data. The Root Mean Squared error (RMSE) on the testing set was 0.0921, and the Mean Squared Error (MSE) was 0.0094. In a low ammonia nitrogen concentration environment, the model had an mean absolute error of 0.0689 and a MSE of 0.0095. In high ammonia nitrogen concentration environments, the mean absolute error was 0.1939, and the MSE was 0.0628. Therefore, the designed method helps to capture potential long-term and short-term dependencies, thereby making the model more stable in predicting water quality changes.