The reliable parameter identification method of the constitutive behavior under large strain conditions is essential for accurate simulation in the hot stamping and warm-cutting process. In this paper, the linear stress-strain curve identification (LSSCI) method is extended to identify the viscoplastic behavior of boron steel at large strain based on the full-field measurements in the Gleeble system. An inverse LSSCI method to evaluate the strain rate sensitivity under different temperatures based on full-field measurements is presented. To overcome the fluctuation of the identified stress-strain curve at large strain, a constrained optimization method is proposed to solve the linear system. The numerical and experimental cases are performed to verify the accuracy and reliability of the proposed LSSCI method. The verification result shows that the identification error of the extended LSSCI method is generally less than 1%, and the maximum error is not more than 4%. The influence of strain rate, specimen shape, constitutive model, noise and discretization on the identification curves have been studied, the results show that the extended LSSCI method is robust when applied to viscoplastic material identification. To accurately describe the viscoplastic behavior of Boron steel at large strain, a unified viscoplastic (UVP) model is proposed based on the dislocation density evolution under elevated temperature. Finally, the tensile testing of shear and notched specimen are carried out to verify the accuracy and reliability of the proposed UVP model, the comparison of load-stroke curve shows that the large deformation behavior of the boron steel under various conditions can be accurately predicted by the UVP model with the parameters from the proposed LSSCI method.

在大应变条件下，本构行为的可靠参数识别方法对于在热冲压和热切割过程中进行精确模拟至关重要。本文将线性应力-应变曲线识别（LSSCI）方法扩展到基于Gleeble系统的全场测量来识别大应变下硼钢的粘塑性行为。提出了一种基于全场测量来评估不同温度下的应变速率敏感性的逆LSSCI方法。为了克服所确定的应力-应变曲线在大应变下的波动，提出了一种约束优化方法来求解线性系统。通过数值和实验案例验证了所提出的LSSCI方法的准确性和可靠性。验证结果表明，扩展LSSCI方法的识别误差一般小于1％，最大误差不大于4％。研究了应变率，试样形状，本构模型，噪声和离散度对识别曲线的影响，结果表明，扩展的LSSCI方法在粘塑性材料识别中具有较强的鲁棒性。为了准确地描述硼钢在大应变下的粘塑性行为，基于高温下位错密度的演化，提出了一个统一的粘塑性（UVP）模型。最后，对剪切试样和缺口试样进行了拉伸试验，以验证所提出的UVP模型的准确性和可靠性，