A combination of off-line feedrate optimization and online adaptive force control is used to maintain a reference peak force during end milling for safe, accurate, and efficient machining. Feedrate optimization algorithms use geometry and force models to calculate feedrates for each tool move, based on a reference peak force. The adaptive controller adjusts the feedrate during machining to maintain the reference peak force. It is the combination of these methods that yields accurate force control, unobtainable with either method by itself. Adaptive control alone is inadequate to handle significant transient cut conditions because of the slow system response time. Optimization algorithms are subject to modeling errors that can lead to significant force errors when cutting. Design parameters for the adaptive controllers are selected using an experimentally validated machining process model. The adaptive controllers are implemented on an open architecture controlled (OAC) 3-axis NC milling machine, and evaluated using three experimental test cases: a sine cut, a prismatic cut, and a corner cut. Feedrates for each cut are first optimized off-line, then used in actual machining with and without controller action. Experimental results demonstrate the ability of the integrated system to effectively regulate peak forces for cutting conditions commonly encountered in end milling operations. In particular, a variable geometry sine cut that initiates chatter shows the advantages of the combined system.
Stable adaptive control of manipulators with improved transients via supervision of the free-design parameters and sampling period
