Enhanced water quality index prediction using a dual-key convolutional transformer autoencoder network integrating with poplar optimization

Water quality, which represents the biological, chemical, and physical properties of water and its appropriateness for different applications including drinking, recreation, and agriculture, is an essential component of environmental health. Severe health hazards, environmental damage, and economic challenges can result from poor water quality. Water quality has a direct impact on ecosystem stability, biodiversity, and human well-being, making it a critical environmental health indicator. For efficient administration of water supplies, early pollution identification, and the assurance of clean water for a variety of purposes, accurate prediction of the Water Quality Index (WQI) and the Classification of Water Quality (WQC) is crucial. The accuracy and flexibility needed for intricate water systems are frequently lacking from traditional methods. In order to predict WQI and WQC with remarkable accuracy, this study presents a novel method, the Dual-Key Convolutional Transformer Autoencoder Network (DKCTAN). The proposed method uses the Indian water quality Dataset. To improve data integrity and lower noise, the dataset is subjected to a rigorous preprocessing process that includes decimal scaling, new adjusted min-max scaling, and statistical column normalization. This is followed by zero-correlation linear analysis. The novel method DKCTAN is used for WQI prediction and WQC. The Poplar Optimization Algorithm (POA) is utilized to optimize the model, yielding an incredible 99 % accuracy rate with a remarkably low error rate. In addition to providing a reliable means of monitoring and evaluating water quality in real time, this approach offers a scalable framework for more extensive environmental management applications, opening the way to more sustainable practices and better decision-making.