Extreme Learning Machine and Particle Swarm Optimization in optimizing CNC turning operation

International Conference on Innovative Technology, Engineering and Sciences 2018

工业技术;自然科学

Janahiraman, Tiagrajah V.^1 ; Ahmad, Nooraziah^1 ; Nordin, Farah Hani^1

Centre for Signal Processing and Control Systems, Dept. of Electronics and Communication Engineering, College of Engineering, Universiti Tenaga Nasional, Jalan IKRAM-UNITEN, Selangor Kajang

43000, Malaysia^1

关键词: Artificial intelligent techniques;    Extreme learning machine;    In-process parameters;    Modelling techniques;    Optimal cutting parameters;    Optimization techniques;    Performance functions;    Traditional techniques;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/342/1/012086/pdf
DOI  :  10.1088/1757-899X/342/1/012086

来源: IOP

【 摘 要 】

The CNC machine is controlled by manipulating cutting parameters that could directly influence the process performance. Many optimization methods has been applied to obtain the optimal cutting parameters for the desired performance function. Nonetheless, the industry still uses the traditional technique to obtain those values. Lack of knowledge on optimization techniques is the main reason for this issue to be prolonged. Therefore, the simple yet easy to implement, Optimal Cutting Parameters Selection System is introduced to help the manufacturer to easily understand and determine the best optimal parameters for their turning operation. This new system consists of two stages which are modelling and optimization. In modelling of input-output and in-process parameters, the hybrid of Extreme Learning Machine and Particle Swarm Optimization is applied. This modelling technique tend to converge faster than other artificial intelligent technique and give accurate result. For the optimization stage, again the Particle Swarm Optimization is used to get the optimal cutting parameters based on the performance function preferred by the manufacturer. Overall, the system can reduce the gap between academic world and the industry by introducing a simple yet easy to implement optimization technique. This novel optimization technique can give accurate result besides being the fastest technique.