基于改进粒子群算法的高强钢Y-U本构模型参数反演

Y-U CONSTITUTIVE MODEL PARAMETERS INVERSION OF HIGH STRENGTH STEEL BASED ON MODIFIED PARTICLE SWARM OPTIMIZATION ALGORITHM

建立Y-U本构模型参数反演的数学优化模型,针对本构模型迭代求解等特点,对传统粒子群算法的参数设置、初始群体生成等方面进行改进,并融合遗传算法中的变异思想,用于此优化问题的求解。根据高强钢材料WELDOX1100的拉压循环试验数据对模型中的7个参数进行反演。最后进行折弯回弹有限元计算和试验,验证了反演结果的准确性。结果表明,该反演算法收敛性好、精度高,反演结果对提高高强钢成形的数值模拟精度具有重要意义。

Mathematic optimal model of parameters inversion in Y-U constitutive model was established. According to complex iterative process of Y-U constitutive model, a modified particle swarm optimization algorithm was adopted, which employed a mutation method of genetic algorithm and was improved on the aspects of parameters setting and initial group generation. A group of seven parameters was inversed based on tension-compression experimental data of high strength steel WELDOX1100. The accuracy of inversion parameters was verified through bending experiment and finite element analysis. The inversion results show refined convergent behavior and high precision, and are significant to increase the simulation accuracy of high strength steel forming.