An Improved Online Trajectory Planner With Stability-Guaranteed Critical Test Curve Algorithm for Generalized Parametric Constraints

Challenge issues, such as the simultaneous achievements of high accuracy and productivity in the motion control of mechatronic systems, are widely investigated by the trajectory planning approach. Particularly, a kind of online trajectory planners are developed by Guarino et al. showing the superiority on flexibility and time efficiency in the practical implementation. In this framework, the assigned joint constraints are first transformed into the corresponding bounds on the parametric velocity and acceleration, respectively. However, these existing planners have to assume that the upper and lower bounds of the parametric acceleration are strictly limited to be positive and negative, respectively, which may not be always true in some practical situations and thus will lead to an incorrect trajectory planning solution. To address this restriction, an improved online trajectory planner is proposed in this paper, which consists of a bound estimator and an improved nonlinear filter. The bound estimator transforms the kinematic constraints from the joint space to the parameter space, whereas the online scaling of the rough referenced parametric trajectory is fulfilled by the improved nonlinear filter. Specifically, the limitations on the parametric acceleration with a positive upper bound and a negative lower bound are removed now. And the global convergence of the improved nonlinear filter is presented theoretically. Additionally, an efficient online critical test curve approach is developed to cope with potential stability issues, which are usually dealt with by employing an offline tuning method in the previous online trajectory planning. Comparative planning results show the drawbacks of the previous online trajectory planner in some situations and also confirm the significant improvements of our proposed online trajectory planner. Several contour tracking experiments are conducted and validate that the proposed trajectory planner can achieve a better contour tracking performance.