Real-time adaptive and trajectory-optimized manipulator motion planning

While there has been a large body of literature addressing offline path planning for manipulators, there is relatively less study on real-time motion planning that occurs as a manipulator moves in an environment with unknown obstacles or unknown changes. This paper introduces a unified and general motion planning approach based on evolutionary computation that is suitable for both offline and real-lime adaptive motion planning for manipulators under various optimization criteria and manipulator constraints in environments with obstacles or changes not known a priori. The implementation and testing results demonstrate the effectiveness and efficiency of the approach.