Machine Learning-Optimized Microplastic Adsorption Kinetics in Marine Environments with Edge Computing

Microplastics in marine environments exhibit complex adsorption behavior towards persistent organic pollutants, posing an ecological threat. In this paper, a framework based on edge computing is proposed to predict the adsorption dynamics of microplastics, and a population balance equation model based on machine learning optimization is established. This model can dynamically adjust the nonlinear coefficient, and distributed edge nodes realize the $G^{\prime}/G$ expansion method, which is used to solve the traveling wave equation in real time. The Internet of Things integrates sensor data for continuous model updating. The experimental results show that compared with traditional methods, the adsorption isotherm fitting performance is superior, and edge based processing will effectively reduce computational energy consumption, providing theoretical insights into the influence of microplastic surface characteristics on co pollutant behavior.