Study on the Optimum Processing Parameter of EDM Based on the SVM
-
摘要: 分析了电火花加工中电参数与加工质量之间的关系,并运用支持向量机(SVM)对电火花加工中电参数进行了优化研究。仿真结果表明:电参数预测精度最高可达96.10%,最低89.20%,平均94.28%,说明SVM算法稳定性及泛化能力优秀。进一步经实验验证,预测精度最高达92.65%,最低81.50%,平均89.38%,同样较高。说明该方法所确定的最优电参数能够很好地保证预期的加工质量,从而可以方便操作者对加工条件的确定。Abstract: The relationship between the electric parameter and the processing quality is analyzed,and the optimum electric parameter is predicted by using the support vector machine(SVM). The simulation result shows that the highest prediction accuracy is 96. 1%,the lowest is 89. 2%,the average accuracy is 94. 28%,which indicate the algorithm stability and generalization ability are outstanding,after the experiment,the highest prediction accuracy can reach to 92. 65%,the lowest is 81. 5%,the average accuracy is 89. 38%,the electric parameter optimized based on the SVM can guarantee the expected processing effect better. The exploration in EDM intelligent machining will be convenient for the operators to determine the most effective machining condition.
-
Key words:
- artificial intelligence /
- computer simulation /
- efficiency /
- electric discharge machining
-
[1] 赵万生. 先进电火花加工技术[M]. 北京:国防工业出版社,2003Zhao W S. Advanced electric spark machining technology[M]. Beijing:National Defense Industry Press,2003 (in Chinese) [2] 胡建华,汪炜,徐启华. 一种基于模糊网络确定电火花加工条件的新方法[J]. 机械科学与技术,2008,27(18):1097-1100Hu J H,Wang W,Xu Q H. A new method fordetermination of the optimal machining conditionsin electrical discharge machining using fuzzy neuralnetwork [J]. Mechanical Science and Technology forAerospace Engineering,2008,27(18):1097-1100 (in Chinese) [3] 杨浩泉,周平. 电火花加工效果预测及工艺参数智能选择系统的研究[J]. 科学技术与工程,2010,10(16):3968-3971Yang H Q,Zhou P. Intelligent system for qualityprediction and parameters optimization of EDM[J].Mechanical Science and Engineering,2010,10 (16):3968-3971 (in Chinese) [4] 吴流发. 电火花线切割加工工艺参数智能化研究[J]. 模具制造,2007,(4):70-72Wu L F. Study of the parameters optimization in WEDM [J].Mould Manufacture,2007,(4):70-72 (in Chinese) [5] G. Krishna Mohana Rao et al. Development of hybridmodel and optimization of surface roughness in electricdischarge machining using artificial neural networks andgenetic algorithm[J]. Journal of Materials ProcessingTechnology,2009,(209):1512-1520 [6] Vapnik V N. 统计学习理论[M]. 许建华,张学工,译. 北京:电子工业出社,2009Vapnik V N. Statistical learning theory[M]. Xu J H,Zhang X G,Translator. Beijing: Electronic IndustryPress,2009 (in Chinese) [7] 邓乃扬,田英杰. 数据挖掘中的新方法-支持向量机[M]. 北京:科学出版社,2004Deng N Y,Tian Y J. New methods in data mining-SVM[M]. Beijing:Science Press,2004 (in Chinese) [8] 李方方,赵英凯. 基于 Matlab 的最小二乘支持向量机的工具箱及 其 应 用[J]. 计 算 机 应 用,2006,26:-360Li F F,Zhao Y K. The least squares SVM based onmatlab toolbox and its application [J]. Journal ofComputer Applications,2006,26:358-360 (in Chinese) [9] 张彤,殷菲,倪宗瓒. 基于支持向量机的婴儿死亡率预测模型[J]. 中国卫生统计,2004,21(2):78-80Zhang T,Yin F,Ni Z Z. Infant mortality predictionmodel based on SVM[J]. Chinese Journal of HealthStatistics,2004,21(2):78-80 (in Chinese) [10] 赵万生,刘晋春. 电火花加工技术工人培训教材[M]. 哈尔滨:哈尔滨工业大学出版社,2000Zhao W S,Liu J C. Electrical discharge machiningtechnology workers training materials[M]. Harbin:Harbin Institute of Technology Press,2000 (in Chinese)
点击查看大图
计量
- 文章访问数: 140
- HTML全文浏览量: 30
- PDF下载量: 2
- 被引次数: 0