Fuzzy supervisory control of end milling process

Abstract

The necessity of raising the efficiency of the milling process in a vertical (end) milling machine determines the implementation of a supervisory control system, which should manipulate the cutting process variables, in order to keep constant a high level of the cutting force. Modeling and control of this process using conventional techniques are quite difficult due to its complexity and nonlinearity. Therefore, a fuzzy control approach is undertaken. The algorithm used for the resulting MIMO controller has classic structure, partitions, and rule base, with IF…THEN rules. Also, the classic “sup-min” compositional operator and center of gravity for the defuzzification strategy are employed. A precalculated look-up table is the actual kernel of the implemented control algorithm. The system hierarchical structure has two levels: the lower one includes the CNC loops, and the higher one the fuzzy controller of the cutting force, acting over the cutting variables set points, and based on a PC. The necessary high- and low-level software as well as the complementary hardware were created. The results are evaluated as very satisfactory, especially concerning the increase of the metal removal rate and the suitable transient response, also compared with linear controllers.