Search

This article presents a highly compact direct-injection power flow and quality (f/q) control topology that employs a novel hybrid switching scheme to eliminate the line transformers and discrete high-current inductors in the high-current series injection, which otherwise dominate size and cost. Moreover, the proposed design is especially suitable for the lower voltage levels from the distribution grid onward using the latest low-voltage high-current semiconductors, specifically fast-switching gallium–nitride transistors operating even with only clip-on magnetics, supported by silicon devices for sufficient over-load and fault tolerance. These floating modules are directly inserted in series with each phase and omit any ground connection, thereby only dealing with a small portion of the voltage difference, unlike series back-to-back converters that have to handle the full power. In consequence, by employing low-power components and omitting large line transformers and inductors, the circuit demonstrates remarkable compactness. This makes it highly suitable for installation in on-street utility boxes. Specifically, it occupies a mere 20% of the volume associated with transformer-injection circuits, 30% of that of series back-to-back converters, and 60% the volume of our previous purely silicon direct-injection configurations.

Modular multilevel converters (MMCs) enjoy growing popularity primarily due to modularity and scalability, but their applications are still constrained by the complexity of hardware and control. MMCs with symmetrical-half-bridge (SHB) submodules can achieve bipolar operation via the simplest structure and reduced cost. However, the imbalance of submodule capacitor voltages remains as a challenge. As a solution, this letter proposes a novel MMC with three-active-switch SHB (3AS-SHB) submodules, which benefits from simple structure and parallel connectivity. Through parallel connection of capacitors among submodules, the proposed MMC allows sensorless voltage balancing, and hence the removal of many voltage sensors, thereby reducing both hardware and control complexity as well as lowering the system cost. Moreover, inherent voltage balancing avoids overmodulation and output waveform distortion. To achieve parallel connectivity, we propose a novel modulation strategy to generate complementary driving signals with adjustable double-edge deadbands. Finally, simulations and experiments validate the effectiveness of the proposed MMC and voltage balance scheme.

This article addresses the common-mode (CM) resonance problem in parallel inverters with separate dc inputs and improved LCL filters. Unlike single inverter system, parallel system presents a more challenging picture. Especially when the grid impedance and parasitic capacitance are considered, the parallel inverter interactions introduce more complex coupling relationships. In this situation, intricate CM resonances are excited at various frequencies, which is different from the resonance characteristics of a single inverter. Therefore, the control methods based on the single inverter system is insufficient in the parallel system and cannot behave as expected. To solve this issue, the CM resonance suppression problem in parallel system with separate inputs and parasitic parameters is being considered and analyzed for the first time. Firstly, the Thevenin's CM equivalent model of the parallel system is established. Compared with a single inverter model, the increase of CM loops triggers another two CM resonance peaks, which further deteriorates the system stability. Then, the internal and external CM excitations between each inverter are analyzed. The results show that the CM current of each inverter is coupled with the CM excitation sources of all inverters and the resonance characteristics are affected by the number of inverters. Furthermore, the influence of grid impedance and parasitic capacitance on the CM resonance characteristics of the system is analyzed. Accordingly, a CM resonance controller and a control parameter design method are proposed. By damping the CM excitation sources, the CM resonance suppression and CM current attenuation can be achieved simultaneously.

Grid-forming converters are gradually becoming the enablers of more-electronic power systems. However, the challenge of fault-ride through prevents the widespread adoption of grid-forming converters. Although the existing literature investigates active power-transfer limitations and transient stability problems of grid-forming converters during fault-ride through, few of them touch reactive power-transfer limitations. In this article, we reveal that grid-forming converters also face transient voltage stability problems, which are caused by the violation of reactive power absorption limitations between converters and grids. This is particularly of interest for grid-forming static compensators and different from conventional power system voltage stability problems. Moreover, this article derives the equilibrium points, reactive power absorption limitations, and stable regions of grid-forming converters with two typical reactive power controllers, namely, reactive power proportional–integral (PI) and voltage-droop control. To improve fault-ride through, this article further proposes and designs voltage inertia control, which smoothens the voltage amplitude change and extends the stable region. In addition, this article finds that the active power control deteriorates transient voltage stability. Finally, the experimental results verify the analysis and the impact of control on the transient voltage stability of grid-forming converters.

This article presents a novel distributed secondary control method for dc microgrids. The method improves existing results by considering the effects of transmission line impedances and constant power loads (CPLs) on system stability. Specifically, a new parameter named “virtual current derivative (VCD)” is defined, which simplifies system analysis and facilitates the improved control design. It is shown that through the dynamic averaging of VCDs, the dc microgrid system model can be reduced to a second-order one regardless of the number of converters connected. This second-order model facilitates the stability analyses of the dc microgrid, thereby leading to a sufficient condition for determining the allowable CPL upper bound. Simulation and experiment studies are conducted to verify the proposed method.