Multi-objective adaptive trajectory optimization for industrial robot based on acceleration continuity constraint

Abstract

In noncontact machining, such as welding and spraying, running efficiency and smoothness have been a bottleneck problem in trajectory optimisation of industrial robots. When the dynamic and mechanical properties of robots are fully utilised, significant impact are often produced. Thus, reducing the process impact of the robot and achieving a balance between the efficiency and smoothness of the operation process are complex problems that need to be solved. The trajectory optimisation problem is modelled based on the excellent properties of the convex-optimization (CO) problem. CO parameters were introduced to solve the inconsistency between the constrained and planned spaces. To obtain the weight factor of the multi-objective optimisation problem, a multi-objective adaptive optimisation method was proposed to select the optimal parameters. Aiming at the large process impact of the optimisation problem, which considers running time as the main optimisation objective, an acceleration continuity constraint method was proposed based on the smoothing path. Finally, the objective optimisation problem was transformed into a standardised second-order cone programming problem to complete the solution. The experimental results show that the proposed method can improve the robot operation efficiency and reduce the process impact by approximately 26% and 22%, respectively, from three aspects: optimum time, balance between efficiency and impact, and acceleration continuity.