A hybrid water quality prediction model based on variational mode decomposition and bidirectional gated recursive...

Area of Science:

• Environmental Science
• Data Science
• Hydrology

Background:

• Accurate water quality prediction is crucial for effective river ecological management and pollution prevention.
• Nonlinearity and instability in water quality data present significant challenges for traditional prediction methods.

Purpose of the Study:

• To develop a hybrid model for enhanced water quality prediction accuracy.
• To address the challenges posed by nonlinear and unstable water quality data.

Main Methods:

• A hybrid model combining Sparrow Search Algorithm-optimized Variational Mode Decomposition (SSA-VMD) and Bidirectional Gated Recurrent Unit (BiGRU) was proposed.
• The Sparrow Search Algorithm (SSA) optimized VMD parameters using fuzzy entropy (FE) as the fitness function.
• SSA-VMD decomposed water quality data into components, which were then predicted individually using BiGRU.

Main Results:

• The proposed SSA-VMD-BiGRU model demonstrated superior prediction accuracy and stability compared to existing methods like EMD- and CEEMDAN-based models.
• Prediction accuracy for dissolved oxygen (DO) reached 97.8%, and for potential of hydrogen (pH) reached 96.1% in validation tests.
• The model was validated using real-world DO and pH data from Qiandao Lake, China.

Conclusions:

• The developed hybrid model offers a robust solution for predicting water quality parameters like DO and pH.
• This approach provides valuable technical support for river water quality protection and pollution prevention strategies.
• The findings highlight the potential of SSA-VMD-BiGRU for complex environmental data analysis and forecasting.