摘要
应力强度因子是表征材料断裂的重要参量,与应力大小,裂纹的形状和裂纹长度有关。对应力强度因子进行分析,基于最小二乘支持向量机原理,结合粒子群优化,建立以应力大小和裂纹长度作为输入值,应力强度因子为输出值的模型,从而对应力强度因子进行分析和预测。模型预测值与理论值进行分析比较,结果显示,基于最小二乘支持向量机结合粒子群优化算法建立的数学模型,模型拟合优度为0.994 9,可通过应力大小和裂纹长度预测应力强度因子,预测值与精确值的相对最大误差为0.186 4,可证明该模型的适用性与精确性。
The stress intensity factor （SIF） is known to be an important factor inducing the deterioration and consequently the breakage of workpiece, which is closely dependent on stress, shape and length of crack. In this work, the SIF was investigated. Based on least squares support vector machines （LS-SVM） modeling approaches coupled with particle swarm optimization （PSO） optimization strategy, efficient statistically predictive models were developed. Taking stress and length of crack as input parameters and SIF as output parameters, the results indicated that the PSO- LSSVM models are capable of capturing the complex nonlinear relationship between the input and output variables, coefficient of determination （ R2 ） values was 0. 9949. The SIF could be predicted by stress and length of crack, and the relative maximum predicting error was 0. 1864, which verified the reliability and validity of the proposed model.
出处
《机械设计与研究》
CSCD
北大核心
2016年第4期125-127,共3页
Machine Design And Research
基金
部级预研项目
作者简介
E-mail：maqingyan@nuc．edu．cn； 马清艳（1980-），女，讲师，博士研究生；主要研究方向：机械制造及其自动化，已发表论文8篇。
