Mustapha Asnoun

This paper presents a grid-voltage-sensorless control scheme for a grid-connected modular multilevel converter based on an adaptive-linear-neuron quadrature signal generator for virtual-flux estimation (VF-ADALINE-QSG). The proposed estimator identifies the DC offset component and extracts the in-phase and quadrature fundamental components online without requiring offline training or specific initialization of the adaptive weights. By reconstructing VF exclusively from the identified fundamental components, the proposed method limits drift induced by DC offset and attenuates higher-order harmonics while avoiding the additional dynamic delays associated with conventional filtering-based estimators. The VF-ADALINE-QSG is integrated into a deadbeat predictive control framework incorporating grid-current regulation, circulating current suppression, and submodule-capacitor-voltage balancing. Its performance is compared to that of low-pass-filter (LPF) and second-order generalized-integrator (SOGI)-based VF estimators through real-time simulations on an OPAL-RT OP5700 platform. The evaluation considers startup, active-power-reference variations, a 20% DC offset, voltage distortion comprising 12.5% fifth- and 7.5% seventh-order harmonics, and the simultaneous occurrence of these disturbances. The proposed estimator achieves a settling time approximately one-sixteenth of that obtained with the LPF-based estimator. Under DC offset conditions, the estimated VF total harmonic distortion is limited to 0.26%, compared with 7.38% and 0.84% for the LPF- and SOGI-based estimators, respectively. The results demonstrate a faster transient response, improved estimation accuracy, and enhanced robustness against both DC offsets and harmonic disturbances.

The widespread development of medium-voltage and high-voltage direct current transmission systems has highlighted the modular multilevel converter (MMC) as a crucial enabling technology. However, the overall performance and lifetime of the MMC strongly depend on the integrity of its submodules (SMs), making online capacitance condition monitoring a critical requirement. Unlike recent related studies that rely on computationally heavy matrix-based algorithms or “black-box” artificial neural networks requiring massive offline training datasets, this paper proposes a parametric, adaptive linear neuron network. Mapped directly to the physical equations of the MMC, the method simultaneously exploits the arm current, SM switching state, and capacitor voltage to identify online parametric variations caused by aging or harsh conditions. The proposed scheme is fully non-intrusive, requiring no additional hardware sensors or signal injections, thereby reducing implementation complexity. The simulation results obtained in MATLAB/Simulink (vR2024b) demonstrate the method’s fast convergence and a quantified steady-state estimation error within ±1%. Furthermore, the estimator exhibits strong robustness under severe operating conditions, successfully maintaining accuracy during a 20% capacitance reduction, a 100% active power step variation, dc-link voltage fluctuations, measurement noise, grid unbalances, and harmonic perturbations.