Closed-loop time-optimal path planning using a multi-objective diversity control oriented genetic algorithm

This paper presents the use of a multi-objective diversity control oriented genetic algorithm (MODCGA) for solving a closed-loop time-optimal path planning problem. The MODCGA is a result of the integration between two types of genetic algorithm: a multi-objective genetic algorithm (MOGA) and a diversity control oriented genetic algorithm (DCGA). The MODCGA is benchmarked against the MOGA and a random search in the path planning problem which is treated as a multiobjective optimisation problem. In this case, the planning problem is represented by a position control task which is given to a 3-dof revolute joint robot. From the optimisation viewpoint, the decision variables consist of the magnitude of torque limits for each joint and the initial and final positions of a fixed length path at which the robot end-effector has to track. The corresponding search objectives are thus expressed in terms of the position tracking error and trajectory time. Two chromosome coding schemes are explored in this investigation: gray and integer-based coding schemes. The simulation results suggest that the integer-based coding scheme is more suitable at representing the decision variables. In addition, the use of diversity control in conjunction with the integer-based coding scheme can further improve the search results.