An Experimental Study on Fuzzy Controller Robustness Augmented by Fuzzy Supervisor for Cutting Force Control of End-Milling

In this paper, the robustness of controlling the cutting force of an end-milling process using a supervisory fuzzy controller has been experimentally investigated. In the proposed controller an ordinary fuzzy controller is augmented by implementing a fuzzy supervisor. The ordinary fuzzy controller is the highly-used cutting force fuzzy controller due to its advantages. However, because of participating only one cutting parameter in its structure, i.e. mostly feed-rate, the ordinary fuzzy controller is suitable only when other parameters such as depth of cut, spindle speed, etc. have a small amount of variations. Since in practice this assumption is not valid, an ordinary fuzzy controller needs to be augmented. This augmentation has had suggestions in the literature. However, these suggestions never have been on the basis of the fuzzy logic; while fuzzy being is the major advantage of ordinary fuzzy controller for controlling cutting force. Due to this deficit, in this paper, a supervisory fuzzy controller has been added to the system. The designed fuzzy supervisor inspects the dynamic behavior of the cutting force, estimates a pre-defined sensitivity parameter, and cancels the output fluctuations arising from this parameter’s variation. The numerous experiments conducted in different cutting situations proved that this supervisory structure increases the robustness and applicability of the fuzzy controller with respect to an ordinary fuzzy controller. These experiments are conducted on two different computerized numerical control machines to confirm the efficiency of the presented method.