MULTI-OBJECTIVE OPTIMAL DESIGN OF GRAVITY COMPENSATORS USING PARETO FRONT WITH GENETIC ALGORITHM

This paper presents a multi-objective optimal design method for gravity compensators. The proposed optimization aims to minimize the ratio of the compensated torque to the uncompensated torque and the joint reaction forces of the gravity compensators. Pareto front of multiple fitness functions using a genetic algorithm is adopted to solve the optimization problem. This work takes a spring four-bar mechanism as a case study. The optimization problem of the gravity compensator is established with consideration of the spring constraints, which facilitate the practical design. This work also provides a numerical example to demonstrate the effectiveness of the proposed method. It is shown that the technique can reduce the joint reaction forces by an average of 35.6% as compared to the original single-objective optimal design method. The gravity compensation performance of the mechanism is effective with a torque reduction ratio of 0.05. Moreover, a prototype of an 0.2-kg gravity compensator was built, and an experimental test was conducted. It was found that the measured motor torque of the mechanism was reduced by up to 93% within a range of 3 pi/4.