Decentralized Control of Cascaded H-Bridge Inverters for Medium-Voltage Grid Integration

Abstract

This paper introduces a fully decentralized control strategy that allows for individualized active and reactive power control of H-bridge inverters connected in series. After formulating a model of the system power flows, we outline a communications-free approach where each inverter can act on independent commands. Inverter-level feedback is accomplished locally via a measurement of current flowing through the stack.A key feature is the lack of any centralized control routines or phase-locked loop. In this framework, the reactive power control loops achieve voltage synchronization among the converters and the grid, and the active power control loops modulate the terminal voltage amplitudes that influence power delivery. Furthermore, pole placement via state feedback is used to ensure stability of voltage synchronization while also enabling tracking of bidirectional power commands. Simulations are presented to validate the proposed approach.

A decentralized control strategy for single‐phase cascaded photovoltaic inverters with MPPT ability

“…Further, the literature [22] proposes a φ-ω (Power Factor Angle Droop Control) control strategy, which can be applied to any line impedance characteristics. However, the premise of [21,22] is that all inverter modules are the same or equivalent, and the amplitude of output voltage is fixed, so it is not suitable for cascaded inverters of different capacities.…”

“…Decentralized strategies of cascaded inverters can overcome these drawbacks of the centralized control mentioned above, because each module only needs its local information to achieve synchronization and power sharing without communication. Due to the peer-to-peer structure, decentralized control has more advantages in terms of cost and reliability, so it is increasingly valued [18][19][20][21][22][23]. In [19], a virtual oscillator-based controller is implemented to synchronize the inverter voltages and share the load.…”

“…However, this approach is only applicable to the island mode. Based on the idea of droop control, a series of decentralized control strategies are proposed in [20][21][22], which can be applied to the grid-connected mode for cascaded inverters. In [20], the virtual impedance method is applied to change the line impedance to resistive, and then Q-ω and P-V inverse droop control method is utilized to decouple reactive and active power respectively.…”

“…It has an advantage of reactive power sharing, but the active power control is essentially constant power control (a PI controller is included in the P–V control loop), so it is not suitable for some islanding applications. In [21], the traditional P – ω droop control is utilized to regulate the output active power of cascaded inverters, but it can only be applied when the line impedance is inductive. Further, the literature [22] proposes a φ – ω (Power Factor Angle Droop Control) control strategy, which can be applied to any line impedance characteristics.…”