Manufacturing Technology Today

Shibendu Shekhar Roy ¹

Affiliations
1 Central Mechanical Engineering Research Institute, M.G. Avenue, Durgapur-713209, West Bengal, IN

Abstract

This paper illustrates the application of combined neural networks and fuzzy inference system for modeling tool life and cutting force in turning operation for set of given cutting parameters, namely cutting speed, feed and depth of cut. The proposed methodology uses a hybrid-learning algorithm i.e., combination of the backpropagation gradient descent method and least squares method, to identify premise and consequent parameters of the first-order Sugeno-fuzzy inference system. The results obtained from proposed method are compared with the experimental results. The comparison indicates that the proposed method can produce efficient knowledge base of fuzzy inference system for modeling the tool life and cutting force in turning operation.

Full Text

     

Shibendu Shekhar Roy ¹

Affiliations
1 Central Mechanical Engineering Research Institute, M.G. Avenue, Durgapur-713209, IN

Abstract

Surface finish is an important parameter in manufacturing engineering. It is a characteristic that could influence the performance of machined parts and the production costs. This paper proposes a method using adaptive neuro-fuzzy inference system to establish the relation between machining parameters and surface finish in end milling and consequently, can effectively predict surface finish of workpiece using input cutting parameters namely, spindle speed, feed rate and depth of cut. The results obtained with the adaptive neuro-fuzzy inference system are compared with the experimental results. The comparison indicates that the proposed method can produce optimized knowledge base of fuzzy system for predicting surface finish in end milling operation.

Full Text

     

Shibendu Shekhar Roy ¹

Affiliations
1 Central Mechanical Engineering Research Institute, Durgapur-713209, W.B., IN

Abstract

The dynamics of a robotic manipulator is of vital significance to its real-time control schemes. The equations related to dynamics of a robotic manipulator are highly non-linear and solving these equations, using the principle of hard computing, is an expensive task computationally. In the present work, an expert system based on fuzzy logic, has been developed to solve the equations of robot dynamics, on-line. The optimal knowledge base (KB) of fuzzy logic-based expert system Is obtained through a GA-based training. Once the expert system Is developed. It can be used on-line, to determine the values of joint torques of the manipulator, for a set of Inputs, namely the joint angles. The effectiveness of the proposed approach is tested by solving dynamics of two degree of freedom manipulator.

Full Text