Research on Optimum Processing Parameter of EDM Based on Support Vector Machine

Abstract:

The processing mechanism of electrical discharge machining (EDM) is complex and there are many factors affecting it, therefore the process parameter is very important for processing quality. This paper analyses the relationship between electric parameter and processing quality, then uses support vector machine (SVM) to predict the optimum electric parameter. The simulation result shows that the highest prediction accuracy is 96.10%, the lowest is 89.20%, average accuracy is 94.28%, indicating that the algorithm stability and generalization ability are outstanding. Further verified by experiment, the highest prediction accuracy can amount to 92.65%, the lowest is 81.5%, average accuracy is 89.38%, and electric parameter optimized by SVM can guarantee the expected processing effect better. The exploration in EDM intelligent machining will be convenient for operators to determine the most effective machining conditions.

• Periodical: Applied Mechanics and Materials (Volume 775)
• Pages: 229-233
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.775.229
• Online since: July 2015
• Authors: Hao Jiang, Yuan Lin, Shuo Wang, Xiu Wu Sui
• Keywords: Electric Parameter, Electrical Discharge Machining, Optimization, Support Vector Machine

References

[1] Wansheng Zhao. Advanced EDM technology [M]. National Defense Industry Press, (2003).

[2] Jianhua Hu, Wei Wang, Qihua Xu. A new method [J] for determining electric spark machiningconditions based on the fuzzy network. mechanical science and technology, 2008, 27 (18): 1097-1100.

[3] Haoquan Yang, Ping Zhou. Electrical discharge machining effect prediction and the system research of intelligent selection of processing parameters [J]. science technology and engineering, 2010, 10 (16): 3968-3971.

[4] Liufa Wu. WEDM process parameters intelligent research [J] mold manufacturing, 2007, (4): 70-72.

[5] G. Krishna Mohana Rao et al. Development of hybrid model and optimizationof surface roughness in electric discharge machining using artificial neural networks and genetic algorithm[J]. Journal of Materials Processing Technology, 2009, (209): 1512~1520.

DOI: 10.1016/j.jmatprotec.2008.04.003

[6] Vladimir N. Vapnik (USA). writed by Jianhua Xu, translated by Xuegong Zhang. Statistical learning theory [M]. Electronic Industry Press, (2009).

[7] Naiyang Deng, Yingjie Tian. A new method in data mining. Support vector machine [M]Science Press, (2004).

[8] Fangfang Li, Yingkai Zhao. Matlab toolbox and its application based least squares support vector machine [J]. Computer Applications, 2006, 26: 358-360.

[9] Tong Zhang, Fei Yin, Nizong Zan. Based on support vector machine infant mortality prediction model [J]. China Health Statistics, 2004, 21 (2): 78-80.

[10] Wansheng Zhao, Jinchun Liu. EDM technology workers training materials [M]. Harbin Institute of Technology Press, (2000).