DSP ‐Implementable RBF ‐Based Adaptive Gain Sliding‐Mode Speed Control for PMSM Motor Drives

This paper presents an RBF‐based adaptive‐gain sliding‐mode speed controller (RBF‐AG‐SMC) for permanent magnet synchronous motor (PMSM) drives. A three‐neuron RBF neural network is embedded in the sliding‐mode control (SMC) loop to self‐adjust the switching gain online, significantly reducing chattering while maintaining robustness to speed reference and load disturbances. The adaptive rule is derived via Lyapunov analysis, ensuring bounded error convergence. The algorithm is evaluated through MATLAB simulation under four representative scenarios: constant speed, 5‐Hz sinusoidal tracking, rapid command toggling, and load‐torque steps. Compared with a conventional PI controller and a fixed‐gain second‐order SMC, the proposed RBF‐AG‐SMC consistently achieves faster rise and settling times, smaller steady‐state error, and lower current/torque ripple. The discrete‐time formulation requires only three RBF Gaussian evaluations and two adaptive gain updates per 1 ms control cycle. On‐chip closed‐loop validation on a TMS320F28379D DSP confirms correct execution within the 1 ms speed‐loop budget with margin to spare, demonstrating practical implementability on standard motor‐drive platforms. © 2026 Institute of Electrical Engineers of Japan and Wiley Periodicals LLC.