In this work, a hybrid model of Taguchi technique is presented for manufacturing process optimization in which critical quality parameter selected for study is surface roughness of automobile components. Proposed hybrid approach includes Taguchi technique, particle swarm optimization and Artificial Neural Network models for optimization. In this work, multi-objective optimization problem is degraded into single-objective optimization for manufacturing process. The main contribution of this work is to improve conventional optimization technique by improving optimal solution searching in the given global search space. Here we have introduced Neural network based prediction model which requires training data and testing data for predicting the surface roughness. For training, 9 specimens of the input data are considered and testing includes 18 specimens. During the optimization process, Taguchi technique is implemented first resulting in optimized output response in terms of surface roughness. For further improvement Particle Swarm Optimization is included which helps to find best fit solution for single-objective problem.
