On a supervisory system for cutting processes (Q1066799)

Described are the basic ideas necessary in constructing an intelligent hierarchical supervisory system for cutting (machining) processes with a human boss at the top. First, definitions of the cutting condition space and the cutting state space are given as Euclidean spaces. Then fuzziness is introduced for the evaluation of the cutting state. A good cutting state is defined as a fuzzy set with the degree of membership of each point in the cutting state space. More essential vagueness lies in the ambiguity in understanding the correspondence between the cutting condition and the cutting state. This correspondence is described as a mapping, and the cause of such vagueness is analysed based on the following hypothesis. A man can only grasp a local region surrounding a specific point in the cutting state space, and hence linearization is made around that particular point (a local grasp). A global grasp can be achieved as the patched whole of such small, subjective, and linearized neighborhoods by taking tangents.