Water quality prediction (WQP) plays an essential role in water quality management for aquaculture to make aquaculture production profitable and sustainable. In this work, we propose hybrid deep learning (DL) models, convolutional neural network (CNN) with the long short-term memory (LSTM) and gated recurrent unit (GRU) for aquaculture WQP. CNN can effectively fetch the aquaculture water quality characteristics, whereas GRU and LSTM can learn long-term dependencies in the time series data. We conduct experiments using the two different water quality datasets and present an extensive study on the impact of hyperparameters on the performance of the proposed hybrid DL models. Furthermore, the performance of hybrid CNN-LSTM and CNN-GRU models are compared with different baseline LSTM, GRU and CNN DL models and also with attention-based LSTM and attention-based GRU DL models. The results show that the hybrid CNN-LSTM outperformed all other models in terms of prediction accuracy and computation time.
Efficient monitoring of water quality parameters is crucial for the successful management of aquaculture systems by ensuring the health and growth of aquatic life. Traditional sensor-based systems are often cost-prohibitive, limiting their adoption, particularly in resource-constrained settings. In this research, we introduce a cost-effective method leveraging virtual sensing to estimate dissolved oxygen (DO) levels using solubility and interdependency equations. Furthermore, we propose deep learning-enhanced virtual sensing (DLEVS) employing deep recurrent neural networks, including long short-term memory, gated recurrent unit (GRU), their bidirectional variants, and attention mechanisms, to accurately estimate DO. With extensive hyperparameter tuning, we have identified that the attention-based GRU (A-GRU) model is optimal, achieving significant reductions in RMSE from 0.3802 to 0.1534 for ADAK dataset and from 0.7505 to 0.0848 for MAC dataset in comparison with conventional virtual sensing methods. Our findings demonstrate that the DLEVS mechanism delivers precise DO estimation with substantially reduced costs, presenting an appealing solution for aquaculture water quality monitoring.