Objective    In this study, multi-ratio biaxial tensile tests are conducted on abdominal aortic specimens, aiming to systematically clarify the mechanical response rules of blood vessels in axial and circumferential directions, and verify the applicability of the polyconvex polynomial hyperelastic constitutive model for arterial tissue.   Methods   Healthy abdominal aortic specimens collected from Wuzhishan pigs were taken as research objects, which were commonly adopted in cardiovascular biomechanical studies due to their high similarity to human vascular anatomy and physiology. Planar biaxial tensile tests were carried out under three stretch ratios of 1:1, 1:3 and 3:1 to obtain complete mechanical response data under multi-axial loading conditions. A four-fiber hyperelastic constitutive model based on polyconvex invariants was adopted to fit the experimental stress-stretch data from each loading case. The nonlinear least squares method was used to calibrate the material parameters of the model, and the coefficient of determination R2 was employed as the index to quantitatively evaluate the fitting accuracy.   Results   Experimental results shown that the abdominal aorta presented obvious nonlinear anisotropic hyperelastic characteristics. Under equibiaxial tensile loading, the circumferential stiffness of the vessel wall was significantly higher than the axial stiffness, presenting a distinct directional difference in load-bearing capacity. Under asymmetric loading conditions, the stress response in the high-strain direction significantly enhanced, and the tangent stiffness risen remarkably with the increase of stretch degree. The overall average coefficient of determination R2 reached as high as 0.9457, reflecting favorable global fitting performance. Especially under asymmetric loading, the model achieved higher fitting accuracy, with the maximum R2 up to 0.9871.   Conclusions   The polyconvex constitutive model can accurately describe the complex multi-axial mechanical behavior of abdominal aorta under various loading conditions, and possesses good numerical stability and predictive ability. This study provides reliable experimental data and theoretical support for vascular biomechanical simulation, finite element analysis and structural optimization of interventional devices, and also offers a mechanical reference for further research on the pathological mechanism of vascular diseases.