Comprehensive water quality evaluation based on kernel extreme learning machine optimized with the sparrow search algorithm in Luoyang River Basin, China

Water quality evaluation is crucial to water environmental quality management. Due to the low efficiency and rationality of the traditional automatic monitoring in water quality evaluation, a comprehensive water quality evaluation model based on kernel extreme learning machine (KELM) was proposed to improve the performance of the model in Luoyang River Basin, China. Besides, a novel metaheuristic optimization algorithm, sparrow search algorithm (SSA), was implemented to compute the optimal parameter values for the KELM model. Extreme learning machine (ELM), KELM, support vector regression (SVR), and backpropagation neural network (BPNN) were considered as the benchmark models to validate the proposed hybrid model. Results showed that the water quality evaluation model based on KELM optimized with the SSA (SSA-KELM) outperformed other models. The proposed hybrid model can successfully overcome the nonstationarity, randomness, and nonlinearity of the water quality parameters data with a simple structure, fast learning speed, and good generalization performance, which is worthy of promotion and application. The research results can objectively and accurately determine the status of basin water quality and provide a scientific basis for basin water environment protection and management planning.

Automated summary

A water quality evaluation model based on KELM and optimized with SSA in the Luoyang River Basin outperformed other benchmark models, showing fast learning speed and good generalization performance. The hybrid model can successfully overcome nonstationarity, randomness, and nonlinearity in water quality data and provide valuable insights for water environment protection and management planning in the basin.