Bi-objective optimization for improving the locomotion performance of the vibration-driven robot

The average velocity and the backward velocity are two critical factors to evaluate the locomotion performance of a vibration-driven robot. These two factors are often antagonistic when the ground friction is isotropic and, consequently, lead to a bi-objective optimization. The goal of this paper is to give a practicable bi-objective optimization scheme for vibration-driven locomotion robots. To this end, a generalized dynamic model of the multi-module vibration-driven locomotion robot is built by assuming that the ground friction is isotropic Coulomb type and the driving force is ideal. The non-dominated sorting genetic algorithm II (NSGA-II) is then employed to construct the optimization scheme and then obtain the Pareto front. A numerical example with a two-module vibration-driven system is provided to demonstrate the necessity and effectiveness of applying the bi-objective optimization method. A robot prototype is further designed, and a series of experimental tests are carried out. By choosing the motors’ rotary speed and phase shift as optimization parameters, the average and backward velocities are extracted to check against the Pareto front. The consistency between the theoretical predictions and the experimental results shows the practicability of the proposed bi-objective optimization scheme. Therefore, the proposed optimization scheme will provide comprehensive guidance for system design and optimization in micro-locomotion robots.