Abstract: A parameter identification method based on an improved dung beetle optimization (DBO) algorithm was proposed to address the current issues of insufficient accuracy and low efficiency in multi-parameter identification of permanent magnet synchronous motors (PMSM). Firstly, the PMSM full rank discrete equation was established in the synchronous rotating coordinate system. Secondly, to improve the precision of DBO algorithm in identifying PMSM model, Tent mapping and reverse learning strategy were used to enhance the uniformity of initial population, and a nonlinear update factor was designed. The factor was adjusted as the number of iterations increased, which could improve the search and development capabilities of DBO. Finally, to address the slow convergence problem of DBO, the location update strategy for obstacle-free ball-rolling beetles were improved to accelerate its convergence, and the PMSM parameter identification fitness function was designed and DBO algorithm was improved and applied for parameter identification flowchart. Simulation and experiment verification show that the improved DBO algorithm has faster speed and higher accuracy in parameter identification of permanent magnet synchronous motors.