Improved Moth-Swarm Algorithm to predict transient storage model parameters in natural streams

Abstract

Transient storage model (TSM) is the most popular model for simulating solutes transport in natural streams. Accurate estimate of TSM parameters is essential in many hydraulic and environmental problems. In this study, an improved version of high-level Moth-Swarm Algorithm (IMSA) was used to predict the TSM parameters. First, the performance of the improved model was successfully assessed through several benchmark functions. Next, a series of 58 measured hydraulic and geometric datasets was used to validate the model. The data were divided into two series randomly, 38 datasets were selected for derivation and the remaining 20 datasets were used to verification. Then the results of IMSA were compared with other algorithms proposed by previous researchers. Two statistical indices of root mean square error (RMSE) and coefficient of correlation (CC) were employed to evaluate the performance of the model. The results showed that despite the high complexity and uncertainty associated with the dispersion processes, the IMSA algorithm could accurately predict the TSM parameters.