Kinematic and dynamic motion space analysis and robot motion planning

With the development of the nonlinear feedback robot controller, task space robot motion planning becomes a more immediate approach as well as a necessary step to create robots that are more intelligent. This is because the nonlinear feedback robot motion controller can understand commands describing desired robot task space trajectories and is "intelligent" in this sense. In this paper, we attempt to develop a robot motion planner that is suitable for the more intelligent robot motion planning. The objective of the robot motion planner is such that it can generate robot motion trajectories, based on the robot operator's higher level commands, that satisfy all the robot kinematic and dynamic constraints as well as constraints of the robot working environment and is immediately executable by the robot controller.;For this purpose, we proposed a robot motion planner structure. This robot motion planner is suitable for time-based robot motion planning, while accommodating event-based robot motion planning and sensor based robot control. The components of the proposed robot motion planner are analyzed in detail. It is proposed that Petri Net theory be used to model the generation of simple subtasks from a given task. We proposed to investigate the kinematic and dynamic motion spaces of robot arms as measures of the robots' kinematic and dynamic capabilities. For Multiple robot arm cooperation, we proposed algorithms that can generate the kinematic and dynamic motion spaces of the robot arms. These algorithms are illustrated by examples and computer simulations. Furthermore, an example of the implementation of the proposed robot motion planner is given to show step-by-step operation of such a robot motion planner.