An Improved Island Algorithm and Its Application in Model Optimization of Micro Soft Robot

Abstract

Aiming at the shortcomings of the original island algorithm (IA), which has a slow convergence speed and is prone to local optimality, the island algorithm with the characteristics of Levy flight (LevyIA) was proposed by introducing the Levy flight strategy, which replaced the position update method in the original algorithm and made use of the occasional long jump of Levy flight strategy to jump out of the local optimal solution. The simulation test of the improved algorithm is carried out with 6 test functions, and the experimental results show that the improved algorithm LevyIA can effectively solve the problems of slow convergence speed and local optimization of island algorithm. For the micro-soft robot model with multi-mode movement, IA and LevyIA algorithms were used to optimize the size of the robot and the appropriate magnetic field intensity needed to drive the robot to deform and move. Finally, the experimental data of swimming speed of the robot obtained by simulation shows that, among the three optimization results obtained by LevyIA algorithm and IA algorithm, LevyIA algorithm can make the robot swim faster when moving forward with minor perturbations.