Automated CNC Programming by the Restricted Boltzmann Machine Algorithm

Abstract

The computer numerical control (CNC) machine is efficiently used for the mass production of jobs with high accuracy and precision. The CNC machines perform the required machining operation according to the machining program developed by its user. In this paper, a machine learning algorithm namely restricted Boltzmann machines algorithm (RBM) and deep belief network (DBN) is used for the automatic development of machining codes for machining operation on a CNC machine. The DBN is known as unsupervised, layered greedy pre-training. The algorithm captures the information for the required machining operation to be performed and thereafter generate different options of machining program automatically on the basis of the machine intelligence. The MATLAB platform is used to implement the algorithm so as to determine the position and other parameters of machining operations and generate the machining codes automatically. It is observed that the RBM can be successfully used for the automatic development of CNC machining programs for real-time machining of jobs on the CNC machining centers. The automatically developed machining codes are tested on CNC simulator called CNC TRAIN.