Automated intelligent manufacturing system for surface finish control in CNC milling using support vector machines

Understanding the variations in surface finish due to machining is a non-trivial task and cannot be very easily estimated even for a given set of machining parameters and operating conditions due to the complexity of interactions involved. In this work, an attempt has been made to propose an automated intelligent manufacturing system for the estimation and control of surface finish using support vector machines (SVM). SVM is very effective in mapping multi-dimensional parametric problems wherein standard analytical approaches become very complicated to handle. An intelligent surface finish control support system is built to provide assistance to the operator in a priori estimation of surface finish for a given selected set of feed rate, spindle speed and depth of cut. The estimated output can be compared by the operator with the required surface finish specification and if not satisfactory, alternate operating conditions can be defined. If found satisfactory, the operator can directly use these parameters and obtain the desired finish. Such an intelligent system will be a useful support to assist the machine operator in selecting optimum operating conditions to ensure the desired surface finish. The work carried out herein indicates that a lot of scope exists in the application of artificial intelligence techniques in mapping physical phenomena especially in the area of manufacturing wherein the inter-relationships are very complex and hence help build intelligent manufacturing support systems.