Quantum differential evolution–optimized deep learning for water quality prediction in smart cities

Smart environments play a crucial role in achieving sustainable development goals, particularly those related to environmental protection and public health. In recent years, environmental degradation caused by air and water pollution, chemical contaminants, and noise has significantly affected human health, contributing to the growing prevalence of chronic diseases. Among these concerns, water pollution has become a major environmental challenge, emphasizing the need for reliable and accurate water quality monitoring and prediction systems. To address this issue, this study proposes a novel Quantum Differential Evolution with Deep Learning–based Water Quality Prediction framework (QDEDL-WQI) designed to support intelligent environmental management in smart cities. The proposed framework enables accurate estimation of the Water Quality Index (WQI) and classification of water quality levels through a multi-stage process that includes data preprocessing, outlier detection, prediction, and classification. In the prediction stage, an Adaptive Neuro-Fuzzy Inference System (ANFIS) model is employed to forecast WQI values, while water quality classification is carried out using an Attention-Based Bidirectional Gated Recurrent Unit (ABiGRU) model. To further enhance the performance of the classification model, Quantum Differential Evolution (QDE) is applied to optimize the hyperparameters of the ABiGRU network. Experimental results demonstrate that the proposed QDEDL-WQI framework outperforms several recent approaches, highlighting its effectiveness for accurate water quality prediction and intelligent environmental monitoring in smart city environments.