Data‐Model‐Driven Predictive Control Using Dual Event‐Triggered Mechanism for Modular Multilevel Converters

This study proposes a dual event‐triggered data‐model hybrid‐driven predictive control method to tackle model mismatch and high switching frequency (SF) challenges in model predictive control (MPC) strategies for modular multilevel converters (MMCs). First, a data‐driven model is constructed using dynamic linearization (DL) techniques and updated online to compensate for system uncertainties via adaptive estimation methods. This model is then combined with the arm current model to form a hybrid‐driven predictive architecture, reducing prediction errors caused by model mismatch. Next, a dual event‐triggered control framework is developed by merging delay‐compensated predictive values with error convergence criteria. This framework optimizes the control law to reduce the triggering frequency and maintain the controlled variables within acceptable error bounds, thereby lowering the SF. Additionally, the convergence of the tracking error under the proposed method is proven using Lyapunov stability theory. Finally, simulations and hardware‐in‐the‐loop (HIL) experiments were conducted to validate the effectiveness of the proposed strategy.