Linear programming and windowing based feedrate optimization for spline toolpaths

Optimality and computational efficiency are competing objectives in the design of feedrate optimization algorithms for spline toolpaths. This paper presents a new approach, in which the use of Linear Programming (based on the 'pseudo-jerk' concept) is combined with a novel windowing algorithm to handle long toolpaths. The windowing uses the Principle of Optimality to select suitable connection points between adjacent portions of the feed profile. The result is an algorithm with linear computational complexity, and the ability to process long toolpaths in a parallel manner. Simulations and experimental results confirm negligible compromise in machining cycle time and dynamic accuracy.