A Shuffled Frog Leaping Algorithm based on the Improved Simplex Method

Abstract

The shuffled frog leaping algorithm is easily sunk into local optimum and not enough accurate when optimizing the problem of some higher dimensional functions. In view of these shortcomings, a shuffled frog leaping algorithm based on the improved simplex method was presented. To begin with, the search ability of simplex method was improved by generating randomly reflection coefficient, expansion coefficient and compression coefficient within a given range, and adding escape behavior, greedy selection strategy and small-probability dimension mutation operator in the edge reduction operation. Then, the local update strategies of shuffled frog leaping algorithm and modified simplex method on a small scale were executed according to a certain probability in the times of mixed iteration, which combined the global search of shuffled frog leaping algorithm and the local search ability of modified simplex method, so as to have a better balance between global search and local search ability of the put forward algorithm, eventually enhanced its optimization performance. At last, the results of simulation experiments showed that the shuffled frog leaping algorithm based on the modified simplex method had higher optimization performance.