基于RBF-HDMR的高强钢材料参数反求

Inverse Determination of High-Strength Steel Parameters Based on RBF-HDMR Metamodel Technique

将RBF(radial basis function)-HDMR(high dimensional model representation)近似模型技术应用到高强钢DP600的Johnson-Cook(JC)模型参数反求中,在落锤试验测得位移-时间曲线的基础上,建立了以计算机仿真结果与试验数据之间的误差为目标输出,待求参数为设计变量的近似模型。结合遗传算法反求出DP600钢板的JC模型参数,将反求的参数输入模型,计算得到的仿真位移曲线与试验结果对比表明,该近似模型方法具有较高的精度。

The RBF（radial basis function）HDMR metomodel technique was applied to the Johnson-Cook（JC）parameter inverse determination of a high-strength steel DP600.On the basis of the displacement-time curve measured during the experiments,the metamodel was established,which took the errors between simulation results and test data as the target outputs while took the parameters as the design variables.Combined with genetic algorithm（GA）,JC model parameters of DP600 was inversely determined by using the metamodel,and the prediction based on the parameters shows high degree of consensus with the test results.