Online broadband grid impedance estimation method based on multi-objective optimized random PWM

作者：Du, Yan; Yang, Shiyou; Gao, Chunran; Yang, Xiangzhen^*; Wu, Houbo; Su, Jianhui

来源：Journal of Power Electronics, 2023, 23(12): 1867-1879.

DOI：10.1007/s43236-023-00676-2

• Grid-connected inverter
• Grid impedance estimation
• Switching frequency
• Multi-objective optimized
• Genetic algorithm

Grid-connected inverter (GCI)-based online grid impedance estimation (GIE) can be used in the fields of grid state monitoring, fault diagnosis, and the stability control of grid-connected equipment, which can collectively improve the intelligence of GCI. However, the available frequency range of GCI-based GIE is limited by the bandwidth of the controller because a perturbation signal is injected into the reference current. To broaden the bandwidth of grid impedance estimation, an online GIE technology based on multi-objective optimized random pulse width modulation (MOO-RPWM) is proposed in this paper. Besides taking advantage of the harmonic dispersion characteristic of RPWM, the distribution of the switching frequency is further optimized by the constraints on the measurement accuracy and the total harmonic distortion (THD) of the grid current. Moreover, the frequency boundary is constrained by the frequency range of GIE, the stability of the system, and power loss. A genetic algorithm (GA) is applied to obtain the switching frequency sequence that satisfies the optimization targets. A single-phase grid-connected inverter system is built on a StarSim hardware-in-the-loop (HIL) platform to verify that MOO-RPWM-based GIE has the advantages of a broad GIE range, high accuracy, and low disturbance.