The modular multilevel dc-dc converter (MMDC) has attracted much interest in medium voltage dc applications, but it has to face the main issue of balancing submodule (SM) capacitor voltages. This paper proposes a dual-circulant modulation and mathematically proves it possessing the inherent balancing capability for arbitrary operation cases. Two sets of circular switching patterns are preset and combined to complete the full constrains on SM voltages for balancing, and then reallocated to keep the advantage of SM uniformity. The associated polynomial of circulant matrix is introduced and the coprime of polynomials demonstrates the full-rank feature of the extended switching matrix that promises the inherent balancing. Experiments are conducted on prototype to validate the inherent voltage balancing capability of the dual-circulant modulation.
Submodule (SM) capacitor is one of the critical components in modular multilevel converter (MMC) and will gradually age with the working time increasing, which brings potential challenges for the systems reliability. This article proposes a novel online submodule capacitance monitoring scheme. Under normal conditions, the estimated capacitor voltage stays in line with the measured capacitor voltage. However, when the capacitor is aged, there will be a deviation between the estimated and measured values. Based on the voltage variations on rising and falling edge of switching function, the capacitance of aging capacitor can be estimated in real-time. Meanwhile, the proposed capacitance monitoring scheme is based on the mathematical model of the systems and has good applicability. It has no any impact on the normal operation of the MMC systems, no special requirement for the modulation method and no need for additional hardware circuits.
In this paper, power quality regionalization control method with improved piecewise linear representation (IPLR) and improved gray relational analysis (IGRA) was proposed to improve the power quality and operation efficiency with big power quality data. According to the variations of adjacent and interval sampling points of time series, the characteristic points of power quality time series are extracted. The IPLR is applied to reduce the dimension of sequence. With the processed sequence, the coupling relationship of power quality between nodes can be quantified into correlation degree by IGRA method. Based on the correlation degree, a regionalization strategy was defined to Figure out the regionalization results and master nodes by hierarchical clustering method. With the proposed method, the power quality control can be realized by regionalization control and the algorithm efficiency can be improved. The standard IEEE 14 bus system simulation was constructed to verify the effectiveness of the proposed method.
Aircraft have strict requirements for lightweight. The high-voltage power supply (HVPS) is an important structure of the Electro-aerodynamic unmanned aerial vehicles (EAD-UAV) propulsion system. However, the heavier weight of the magnetic components and insulating materials causes the power-weight ratio of HVPS to be lower than the minimum value of 1.2kW/kg. The proposed method uses a magnetic shunt to fully integrate the LCC resonant inductor and uses distributed transformer, which can reduce the weight of the whole machine. Finally, a power supply of 40kV/600W is tested. Compared with the external inductor scheme, the power-weight ratio of the integration scheme is increased by 18.76% to 1. 42kW/Kg, and its weight is only 423. 9g. The experimental results show that the proposed method is correct.
Due to the uncertainty states of diodes in DC-DC converters, they may operate in various continuous conduction mode (CCM) and discontinuous conduction mode (DCM). However, the traditional manual analysis method is not only inefficient as a number of equations are required to be manually formulated and solved, but also incomplete since only a certain set of operations is generally considered. Aiming to avoid these disadvantages, a computer-aided automatic analysis method of DC-DC converters is proposed in this paper. It combines electrical network theory (ENT) with fundamental circuit law (FCL) and power electronics principle (PEP) to automatically explore all the workable operations of the converter and obtain their corresponding steady-state performance analysis results, which contributes to a comprehensive understanding for DCDC converters. Then, the feasibility of the proposed computer-aided automatic analysis method was demonstrated with a single-input dual-output (SIDO) Cuk converter.
