Closure to'Parameter Estimation of the Nonlinear Muskingum Flood-Routing Model Using a Hybrid Harmony Search Algorithm'

摘要

Thanks to the penalty approach proposed in the original paper, negative discharges in the routing procedure were prevented and a global optimum solution for NLMM was obtained without depending on the optimization algorithm. Moreover, thanks to the hybrid use of HS, which has a high search capability, and the BFGS algorithm, which has a high capability of local searching, optimal values of NLMM parameters within a wide range were obtained in each run. In addition, the HS algorithm was made parameter-free and it became possible for it to find quicker and more accurate solutions compared with the original HS. By using the HS-BFGS algorithm with the Wilson data, the optimal parameters obtained for 4P-NLMM and 5P-NLMM are (K = 0.523710, x = 0.297660, m = 1.861927, a = 1.003951) and (K = 0.653465, x = 0.058198, m = 2.608352, a =1.00135, b = 0.651931), respectively. The output hydrograph for the Wilson data was obtained through the use of mentioned optimal parameters, and it is given in Table 9.