Internet of things with nanomaterials-based predictive model for wastewater treatment using stacked sparse denoising auto-encoder

Abstract

Wastewater is a serious concern for the environment. There is a substantial amount of toxins that are discharged continuously from several pharmacological companies that lead to serious damage to public health and the ecosystem. Present wastewater treatment technologies include primary, tertiary, and secondary treatments that remove numerous contaminants; but pollutants in the nanoscale range were hard to remove with these steps. Some of these include inorganic and organic pollutants, pathogens, pharmaceuticals, and pollutants of developing concern. The utility of nanoparticles was a promising solution to this issue. Nanoparticles have exclusive properties permitting them to potentially eliminate residual pollutants but being eco-friendly and inexpensive. This study develops a new Archimedes optimization algorithm (AOA) with Stacked Sparse Denoising Auto-Encoder (SSDAE) model, named AOA-SSDAE for wastewater management in the IoT environment. The presented AOA-SSDAE technique aims to predict wastewater treatment depending on the influent indicators. In the presented AOA-SSDAE technique, the IoT devices are initially employed for the data collection process and then data normalization is performed to transform the collected data into a uniform format. For the predictive process, the SSDAE model is employed in this paper. To improve the SSDAE model's prediction capability, the AOA-based hyperparameter tuning process is involved.