B-Spline-Based Curve Fitting to Cam Pitch Curve Using Reinforcement Learning

Directly applying the B-spline interpolation function to process plate cams in a computer numerical control (CNC) system may produce verbose tool-path codes and unsmooth trajectories. This paper is devoted addressing problem of fitting for cam pitch curves. Considering that curve needs meet motion law follower approximate curve, we use radial error quantify effects curve. The thus boils down solving difficult global optimization find numbers positions points or data such cumulative between original minimized, this attempted with double deep Q-network (DDQN) reinforcement learning (RL) algorithm traceability. Specifically, RL environment, actions set current states are designed facilitate search points, along design reward initialization neural network. experimental results show when angle division value fixed, proposed can maximize number accurately place these right positions, minimum average errors. Our work establishes theoretical foundation studying spline using method.