This article proposes an approach for modeling the electromechanical coupling in industrial robot joints and an approach for identifying the electrical parameters of surface-mounted permanent magnet synchronous motors (SPMSMs) used in robot joints. First, a more accurate joint modeling method is presented that considers both PMSM friction and gear reducer friction, with the PMSM rotor friction modeled using an improved Coulomb-viscous model and the gear reducer friction represented using a simple method that ensures better accuracy for parameter identification. Second, a method is proposed for separately identifying the SPMSM inductance in the \(d-q\) frame and the resistance and flux linkage in the \(\alpha -\beta \) frame using a Recursive Least Squares (RLS) algorithm, addressing the rank deficiency issues that arise when using the q-axis voltage equation alone to identify multiple parameters. The proposed approaches enable more precise modeling of industrial robot electromechanical coupling and more accurate identification of SPMSM electrical parameters, which establish the foundation for further research on the electromechanical coupling phenomenon of industrial robot joints.