Evaluation of the performance of bat algorithm in optimization of nonlinear Muskingum model parameters for flood routing

In this study, bat algorithm is used as an algorithm based on velocity and location of bats to optimize the parameters of Muskingum's nonlinear model for flood routing. The case study of Wilson flood as well as a historical flood from Lighvan area were selected for flood routing and calculating the parameters of Muskingum's model, with the aim of examining the efficiency of this algorithm. The sum of squares of deviations and the sum of the absolute values of deviations between routed and observational flows were considered as the objective functions. According to the results obtained from the Wilson flood routing using the bat algorithm, the values of these objective functions are equal to 35.14 and 22.76 m3 per second, respectively. The results of routing of Lighvan flood by using bat algorithm also indicated that the sum of squared deviations, the sum of absolute values of deviations, and the difference between observed and routed peak flows are equal to 7.24, 6.23 and 0 m3/s, respectively. In the present study, the performance of the bat algorithm was compared with evolutionary algorithms such as genetic, particle swarm, and harmony algorithms.