Adaptive Filtering for Energy Control of a Modular Multilevel Converter Operated as a Virtual Synchronous Machine Under Unbalanced Conditions

This paper presents a control system implementation for operating a Modular Multilevel Converter (MMC) as a Virtual Synchronous Machine (VSM) under unbalanced conditions. Furthermore, a frequency-adaptive filtering strategy is proposed for improving the dynamic response of the MMC energy control loops included in the control system. The filtering strategy is based on Second Order Generalized Integrators (SOGIs) configured as adaptive notch filters, which are made frequency-adaptive by utilizing the virtual speed of the VSM. Compared to a conventional energy control implementation with fixed width moving average filters (MAF) in the feedback signals, the presented approach provides two advantages, i.e., i) reduced equivalent delay in the feedback signal and thereby more damped response for a given controller tuning, and ii) consistent performance independently from grid frequency variations. The presented control system implementation is applicable for both grid-connected operation under unbalanced faults and islanded operation with unbalanced loads. The improvement in dynamic performance compared to a benchmark implementation of MMC energy control with fixed width MAFs is demonstrated by examples of simulation results for a typical HVDC converter terminal.