A Novel Dead-Time Adaptive Control Method for Motor Drives Based on an Improved GaN HEMT Switching Model

Gallium nitride high electron mobility transistors (GaN HEMTs) enable high switching frequencies in motor drives due to their fast switching speed and low on-resistance, thereby enhancing power density. However, the switching characteristics of these devices, coupled with nonlinear parasitic parameters, make it difficult for traditional constant dead-time designs to achieve optimal performance under varying load conditions. To address this challenge, this paper proposes an adaptive dead-time control method based on an improved GaN HEMT switching model. The model incorporates the dynamic effects of parasitic inductance and nonlinear capacitance to accurately model the switching process in motor drives. The proposed method is validated through MATLAB numerical simulations and double-pulse experiments, optimizing the dead-time setting strategy for dynamic adaptive control. Experimental results on a 1kW GaN-based permanent magnet synchronous motor (PMSM) drive platform show that, compared to a constant dead-time approach, the new method improves efficiency by 0.85% at a switching frequency of 100 kHz and reduces total harmonic distortion (THD) by 1.44%.