Skip to content

An Adaptive Frequency-Domain Multiharmonic Voltage Injection System for AC Filter Capacitors With Noise-Driven System Identification

2026 · IEEE Transactions on Instrumentation and Measurement · Vol 75, pp. 9008410-9008410 · 0 citations · 26 references

Abstract

AC filter capacitors in high-voltage direct current (HVDC) converter stations are often subjected to complex multiharmonic voltage excitations, which may lead to vibration and acoustic noise problems. Accurate reproduction of multiharmonic voltage conditions is therefore essential for investigating the harmonic characteristics of these capacitors. However, conventional harmonic injection methods typically regulate individual frequency components sequentially, resulting in low efficiency and limited capability in handling multiple harmonics simultaneously. Recent frequency-domain iterative methods have shown promise, yet they rely on an accurate inverse system response for control gain design, making them sensitive to model-plant mismatches and parameter drifts. This article presents a robust multiharmonic voltage control strategy based on noise-driven system identification and adaptive spectral feedback. The proposed approach employs broadband white noise excitation to rapidly identify the system’s frequency-response function (FRF) as an initial estimate. An error-driven adaptive spectral gain adjustment mechanism is then introduced, which iteratively refines the control input based on spectral error. This mechanism effectively reduces sensitivity to modeling inaccuracies and ensures robust convergence even when system parameters fluctuate during operation. Experimental results verify the effectiveness of the proposed approach, demonstrating accurate multiharmonic voltage synthesis and reference-voltage switching capability. The developed system provides an efficient platform for studying capacitor vibration and noise characteristics under complex harmonic operating conditions.

View source

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.