Efficient computation of smooth minimum time trajectory for CNC machining

Qiang Zhang,Shu-Rong Li

DOI: https://doi.org/10.1007/s00170-013-4790-7

IF: 3.563

2013-01-01

The International Journal of Advanced Manufacturing Technology

Abstract:Previous computation approaches of smooth trajectory for computer numerical control machining either cannot fully utilize the maximum ability of machines, fail to minimum machine time, or cannot be computed efficiently because of the nonlinearity formulation of problem. In this paper, an efficient computation approach is proposed that has been designed as solving a convex optimization problem to generate smooth minimum time trajectory. Since smoothness of trajectory can be adjusted by confining the value of jerk (rate of acceleration change), the proposed approach uses a pseudo-jerk to approximate the real jerk and to realize smoothing trajectory. The difference between the real jerk and pseudo-jerk is that the proposed pseudo-jerk can be linearity formulation in the problem. By using pseudo-jerk constraints, smooth minimum time trajectory planning problem is formulated as a convex optimization problem. Since of convexity of the optimization problem, the solution process can be efficient by using ordinary gradient-based optimization techniques and the optimal numerical solution is uniqueness. The effectiveness of the approach is demonstrated in machining simulations of two plane patterns.