An interactive wandering Wolf Pack algorithm for solving High-dimensional complex functions

High-dimensional complex function optimization is a significant problem in engineering applications. Wolf pack algorithm (WPA) has a good performance in the optimization of high-dimensional complex functions, however in solving high-dimensional, multi-peak complex optimization problems, there are still some disadvantages, such as low precision and ease to fall into local optimum. Thus, this paper proposes an interactive wandering wolf pack algorithm (IWWPA). IWWPA uses an interactive wandering strategy based on differential evolution algorithm to enhance the global exploration ability of scout wolf; adopts adaptive striding step length, centripetal siege strategy and optimizes the termination condition of calling behavior, which improves the efficiency of the algorithm; in the late stage of the iteration, the Gaussian-Cauchy combined mutation operator is introduced to avoid the algorithm from falling into the local optimum and "premature". In the paper, the convergence of the algorithm is analyzed by using Markov process, and then IWWPA and 6-population intelligent algorithm are used to test 14 benchmark functions and 4 variable dimension test functions in 500 and 1000 dimensions. The simulation results show that the improved algorithm has better accuracy and speed performance in solving high-dimensional complex functions.