Multi-objective motion planning of space flexible manipulator system

Abstract

The multi-objective motion planning of a space flexible manipulator system is investigated. The motion planning model we put forward in this paper takes account of the elastic vibration, disturbance to the attitude of spacecraft which carries the manipulator, operation time and energy consumption. The physical programming method is used to convert the multi-objective optimization problem to a single-objective one. A hybrid optimization approach incorporating Gauss pseudospectral method with direct shooting method is then proposed to transform the single-objective problem into non-linear parameter optimization problem. Both Genetic Algorithm(GA) and Sequential Quadratic Programming(SQP) are employed to seek the solution for the parameter optimization problem, which is treated as an approximate optimal solution for the motion planning problem. Finally, numerical simulation results indicate that the method suggested is effective and stable.