Abstract: A self-healing smart composite material solution based on Shape Memory Alloy (SMA) fibers was proposed to address the problem of crack formation in composite laminates under complex loads, where traditional repair techniques were found to be limited in effectiveness. The number of embedded SMA fibers and environmental temperature were selected as variables for analysis using both simulation and experimental methods. Based on the Brinson phase transformation theory and the Hashin failure criterion, a SMA-Hashin combined UMAT user subroutine was developed, and numerical simulations of the stress-strain behavior of the material were conducted using ABAQUS software. Glass fiber composite laminates with different numbers of embedded SMA fibers (4, 6, or 8 fibers, the volume contents of SMA fibers were 5.2%, 7.8%, and 10.4%, respectively) were fabricated through hot press molding. Pre-cracks were introduced by water jet cutting. The tensile mechanical properties and self-healing performance at room temperature and 80℃ (above the martensite reverse phase transformation temperature) were evaluated using an Electro-Hydraulic Fatigue Testing System (EHFTS), an Environmental Monitoring System (EMS), and a Video Image Correlation -3D (VIC-3D). Test results indicate that in terms of mechanical properties, the tensile strength of the specimens increases to varying degrees with the incorporation of SMA fibers. The best improvement rate is observed in specimens with 6 embedded SMA fibers, while a clear saturation effect in strengthening efficiency occurs when 8 SMA fibers are embedded. Regarding self-healing performance, at 80℃, a crack recovery rate of over 95% is achieved with only 4 embedded SMA fibers.