Search

In this article, two continuous control set model predictive control (CCS-MPC) methods with one-step prediction horizon for three-level flying capacitor boost converter (FCBC) that can provide constant switching frequency are presented. The first one is explicit model predictive control (EMPC), which can provide explicit solution. In order to further reduce the complexity and the computational burden of EMPC while meeting the control requirements, a simple model predictive control (SMPC) is proposed as the second CCS-MPC method. The proposed CCS-MPC will be used as the inner-loop controller to control the inductor current and flying capacitor (FC) voltage. Besides, a suitable outer-loop control strategy for the output voltage is proposed to provide a suitable inductor current reference to improve the dynamic response of the output voltage while avoiding overcurrent. Compared with the previous control method for three-level FCBC, the proposed control method has good dynamic performance and stronger ability to resist input voltage disturbance. The effectiveness of the proposed method is verified by simulation and experimental results.

Prompt detection of power switch failure is crucial for fault-tolerant operation of multilevel converters. This article presents a new technique of detecting, identifying, and locating a device fault in a multilevel dc–dc flying capacitor (FC) buck converter (FCBC). This is realized by continuously analyzing the magnetic component (inductor or transformer) voltage harmonics for failure detection and subsequently identifying failure type and location with a look up table method. The harmonics analysis is performed by the existing FPGA controller using discrete Fourier transform (DFT). This new approach uses an auxiliary winding in the existing inductor core without adding any new hardware and can identify the location and fault type with good accuracy and speed. The experimental results validate the proposed concept on a three-level GaN-based FCBC prototype.

Cascaded boost-buck PFC (CBBPFC) converters offer a wide voltage conversion ratio and a near-unity power factor but require a large output electrolytic capacitor, leading to poor reliability and power density. In this paper, a coordinated two-stage operation and control strategy is proposed to significantly minimize the capacitor requirement without any other hardware changes. In a conventional CBBPFC converter, the boost and buck stages either operate independently nor complementally. In contrast, the proposed method operates the two stages in a concerted manner, so it is possible to use the dc-link capacitor with certain voltage fluctuation to buffer the power imbalance between the AC input and DC output. A new coordinated control strategy and a fluctuation-ratio based design consideration are developed to coordinate the operation of the two stages, further complement the system design. A 200W CBBPFC prototype based on the design concept exhibits a maximum reduction of the output capacitor by 83%, a peak efficiency of 95.8%, and a power factor of 0.99.

A comprehensive investigation on local structures in iron melts and their role in nucleation under various cooling rates was performed by means of large-scale molecular dynamics simulations. The embedded atoms method (EAM) was adopted to describe the interactions between iron atoms. Connections between short-range order (SRO), medium-range order (MRO), and crystalline nucleation from iron melts were constructed using several structural analysis techniques, including the radial distribution function, common neighbor analysis method, the Voronoi tessellation, and bond order analysis. The simulation results showed that abundant types of atomic clusters with SRO, mainly including the icosahedral-like (ICO-like) and fcc-like clusters, were predominant in undercooled iron melts. The obtained microstructures were determined by the competition between the ICO-like and crystal-like configurations. There existed a critical cooling rate, below which the fcc-like configurations gain the advantage upon cooling and where crystallization could take place; otherwise, the ICO-like configurations are favored and the glass phases could be obtained. Furthermore, it was proved that the crystal nucleation could be divided into three stages: first, a fluctuation and competition between crystal-like and ICO-like clusters in undercooled melts; second, the formation and growth of MRO clusters via the transformation of atomic configurations from ICO-like to crystal-like; finally, the nucleation of bcc nuclei from the core of steady MRO clusters. This process agrees with the Ostwald's step rule and the findings from other investigations. Based on the analysis of the compositional origin of MRO clusters, we further found that the MRO clusters were mainly composed of fcc-like instead of ICO-like configurations, indicating a negative role of ICO-like configurations in crystal nucleation.

In this paper, a simple model predictive control (MPC) based on state space averaging model for three-level flying capacitor boost converter (FCBC) is presented. The proposed MPC with constant switching frequency is used as the inner loop controller to control the inductor current and flying capacitor voltage. A new outer loop control strategy for the output voltage is proposed to provide a suitable inductor current reference to improve the dynamic response of the output voltage. Compared with the previous control method for three-level FCBC, the proposed control method shows better dynamic performance. The proposed control method is demonstrated through simulations and experimental results.

Prompt detection of power switch failure is crucial for fault-tolerant operation of multilevel converters. The paper presents a new technique of detecting, identifying, and locating a device fault in a multilevel DC-DC flying capacitor (FC) converter. This is realized by continuously analyzing the magnetic component (inductor or transformer) voltage harmonics for failure detection and subsequently identifying failure type and location with look up table method. The harmonics analysis is performed by the existing FPGA controller using discrete Fourier transform (DFT). This new approach uses only existing an auxiliary winding in the magnetic core without adding any new hardware, and can identify the location and fault type with good accuracy and speed. The experimental results validate the proposed concept on a 3-level SiC based FC buck converter prototype. 1

Crystal nucleation is among the most important processes in the synthesis of materials. Here we study the competitive and multistep nucleation process of body centered cubic (bcc) and face centered cubic (fcc) crystals using phase-field crystal simulations. The initial state is a non-equilibrium liquid. This transforms into an amorphous phase composed of clusters with short-range order (SRO) and medium-range order (MRO). Crystal nucleation begins with the formation of MRO clusters structurally similar to the subsequently nucleated crystal. The formation of bcc and fcc nuclei from MRO clusters involves the following steps: (1) formation of a small thin platelet with MRO; (2) its subsequent growth into 3D MRO clusters; (3) development of crystal embryos from MRO clusters; and (4) crystal embryos transformation into a stable crystal nuclei. In addition, the role of bcc precursors in the formation of fcc nuclei is clarified. In particular, {112} surfaces and steps on {110} surfaces of bcc precursors serve as energetically favorable sites for fcc nucleation. This is also reflected in the resulting orientation relationships, i.e., Pitsch, Nishiyama-Wassermann and Kurdjumov-Sachs, between the bcc precursor and fcc nucleus.

As one of the most attractive postsilicon power semiconductor devices, SiC bipolar junction transistor (BJT) has been studied extensively and commercialized in the past few years. However, SiC BJT has not been widely accepted in the market partially because of high driver consumption in the on-state, which is induced by a relatively large constant base current in order to ensure it is fully turned on . In this paper, a new proportional base driver with a simple circuitry is proposed in order to reduce the steady-state base driver consumption of SiC BJT. It utilizes a silicon small-signal mosfet as a variable resistor placed between the driver and the base of the SiC BJT. The collector current of the SiC BJT is measured via a Hall current sensor whose output is connected to the gate of the mosfet . When the collector current decreases, the gate voltage of the mosfet decreases and its on-state resistance increases, resulting in the decrease of the actual base driver current of the SiC BJT. The operation principle of the proposed proportional base drive technique has been verified by theoretical analysis and experimental demonstration of a 1200 V 10 A-SiC BJT-based dc–dc boost converter. Theoretical calculation, PSpice simulation, and experimental results have shown that the steady-state power consumption of the SiC BJT's proportional base driver is reduced by approximately 70% under the light load condition compared to conventional base driver with a constant base current.

Eutectic solidification under rapid solidification conditions has enormous applications, as it can produce microstructure-refined and interface-stable combined composite structures with low cost. However, investigations reported that the coupled interfaces will be destroyed under the condition of large undercoolings. For further understanding of the mechanisms of uncoupling growth, we investigated the uncoupling growth process of binary eutectics during rapid solidification using atomistic simulations. We find that both the nucleation rate and the crystallization velocity for the first phase of eutectics are very high; however, nucleation of the second phase is very inactive: its nucleation rate is low and nucleation incubation time is very long. As the nucleation of the eutectic second phase is severely suppressed during rapid solidifications, the crystallization of the second phase lags far behind, therefore we speculate that the uncoupling growth of eutectics during rapid solidification is nucleation-i...

The paper presents a new multi-objective modulated model predictive control (MMPC) method for power converters with four different model states. The proposed method is based on a new concept of composing an error vector with fundamental state vectors. It can significantly minimize voltage and current ripples while retaining the nonlinear and dynamic performance of conventional MPC methods. Simulation and experimental results on a SiC-based 3-level flying capacitor(FC) converter are provided to validate this concept, and indicate a much reduced control error with the new method.

This paper proposes a multilevel flying capacitor bridgeless PFC converter which allows the use of cost-effective 100V GaN transistors. This new PFC design offers several distinct advantages over the popular two-level totem-pole bridgeless PFC converters using much more expensive 650V GaN devices: lower component cost, lower dv/dt, reduced power loss, and mitigation of device reliability concerns. A 1.6kW, 5-level PFC converter prototype is designed and built, and experimentally demonstrates an efficiency of 99.18% and a power factor of 0.99. Operation principle, design considerations, control strategy, and experimental results are discussed in this paper.

Crystallizations often pass through multiple intermediate structures before reaching the final state, such as amorphous precursors, polymorphs, or denser liquid droplets. However, the atomistic pathways from these metastable phases to final crystals still remain unclear. Here, we investigated the structure evolution process from liquid to final crystals of homogeneous nucleation by atomic-scale simulations and analyzed the intrinsic mechanisms that influence the nucleation pathways. Three different pathways of two-step nucleation were found by visualizing the precursors' evolutions, and some new micromechanisms of two-step nucleation are revealed. We suggest that the solid bond fluctuations can trigger the formation of intermediate precursors, while the precursors' packing density dominates the structural transformation pathways from intermediate phases to crystals. These findings not only shed light on the mechanisms of nucleation but also provide guidance for future refinements of two-step nucleation theory.

This paper proposes a new energy management method using the internal capacitor for energy buffer in the boost cascaded buck-boost (BoCBB) power factor correction (PFC) converter. The theory and existing shortcoming of the conventional PFC are first analyzed. A new energy management method realized by model predictive control (MPC) is proposed to maximize the benefit of the internal capacitor with a considerably reduced capacitance. A 500W PFC is used as a case study. It is found that the energy management method offers more effective control of the output voltage and input current with a much-reduced capacitance. Experimental results validate the proposed concept on a GaN FET based BoCBB PFC prototype.

Credit markets play essential roles in financial institutions. In developing countries, especially in emerging market countries which are in the process of restructuring of industry, it is important to identify the reasons of credit constrains by an efficient way to facilitate economic transaction. On common phenomena in China rural credit markets is the informal credit market develop rapidly, it is very common for rural households to both through their networks. However it is inadequate in addressing rural credit needs. This research has 3 specific objectives: (1) Explore the relative importance of formal vs. informal rural credit markets in China (2) Identify factors determining households' choice of credit sector and the size of loan rural households borrowed from formal or informal credit markets. (3) Explore the degree to which households are constrained in credit access and identify the underlying factors. This paper also has some policy implications based on the empirical studies. The government can release the households constrain by making the formal credit more flexible and increasing the informal credit such as the microfinance for women.

The mechanism of formation of the large-dimension rare-earth(RE) inclusions in steels is discussed by TEM(Transmission Electron Microscope) taking steel 5CrNiMo with addition of RE as example. The two or several large-dimension RE inclusions come out independent nucleation which grow up into a 0.3~0.5 micron in length, then attract and integrate into the large-dimension with 1~5 micron in length. Though the dimension difference between the RE inclusions is varied, the shape is spherical or approximate spherical or elliptical, but no sharply corneous, because RE can promote the inclusion shape variation through metamorphism.

As one of the most attractive post-silicon power semiconductor devices, SiC bipolar junction transistor (BJT) has been studied intensively and commercialized in the past few years. However, SiC BJT has not been widely accepted in the market because of high driver loss in the on-state, which is induced by a relatively large constant base current in order to ensure its fully turning on. In this paper, a new proportional base driver for the SiC BJT is proposed in order to reduce its driver loss. It utilizes a silicon MOSFET as a variable resistor between the base driver and the base of the SiC BJT. The gate voltage of the silicon MOSFET decreases with the collector current of the SiC BJT via a Hall device, thus the on-state resistance of the MOSFET increases and the actual base driver current of the SiC BJT decreases. Comparing to the conventional SiC BJT driver, the proposed new proportional base driver can achieve the decrease of the base current with the decreasing of the collector current of SiC BJTs and has simple circuitry, resulting in the reduction of base driver loss. The simulation and experiment have verified that the proportional base driver of SiC BJT can reduce the driver loss by approximately 70% under the light load condition.

Two-dimensional materials with unique and excellent physical and chemical properties have attracted much attention in recent years. Among the two-dimensional materials, graphene or grapheme-like materials with honeycomb structure can be mainly prepared by the chemical vapor deposition (CVD) method. The key of this method is to select the substrates and control the nucleation and growth process of honeycomb structures. Graphene prepared by CVD contains many structure defects and grain boundaries, which mainly arise from nucleation process. However, the nucleation mechanism of graphene prepared by CVD method is not very clear. In addition, more than ten kinds of metal substrates can be used as substrate materials in CVD methods, such as Cu and Ni, which have nearly always face-centered cubic (FCC) structures and similar functions in the preparation process. In order to better describe the nucleation of graphene and understand the influences of metal substrates, we introduce the structural order parameter into the three-mode phase-field crystal model to distinguish the low-density gas phase from condensed phases. Nucleation processes of graphene on substrates with different symmetries are studied at an atomic scale by using the three-mode phase-field crystal model, which can simulate transitions between highly correlated condensed phases and low-density vapor phases. Simulation results indicate that no matter whether there is a substrate in the nucleation process, firstly gaseous atoms gather to form amorphous transitional clusters, and then amorphous transitional clusters gradually transform into ordered graphene crystals, with continuous accumulation of new gaseous atoms and position adjustment of atoms. In the nucleation process, five membered ring structures act as a transitional function. When grown on the substrate with a good geometric match with the honeycomb lattice, such as (111) plane of FCC metals, the graphene island has small structural defects. However, when grown without a substrate or on the substrate with a bad geometric match, such as (100) plane of FCC metals, the graphene island contains many structural defects and grain boundaries, which are not conducive to the preparation of high quality graphene. Compared with the (100) crystal plane of the tetragonal cell, the (110) crystal plane of the rectangular cell is favorable for the preparation of graphene single crystals with less defects. Therefore, the appropriate metal substrate can promote the nucleation process of graphene and reduce the formation of distortions and defects during the nucleation and growth of graphene.

Wireless Sensor Networks (WSNs) are widely used in various applications for tracking or surveillance purposes. Border security monitoring and terrorist attack prevention are just two of many crucial homeland security applications which rely upon WSN technologies. In these applications, a great amount of sensors powered by batteries are deployed in a particular region. They detect events, process and relay data to the sink node at the base station. Power conservation is critical to these sensors. In this paper, we propose a routing algorithm which separates the sensor nodes to several scheduling sets and keeps track of the energy level of each sensor. This algorithm balances power consumption among sensors, and therefore prolongs the network lifetime. Simulation results verify that our algorithm outperforms the EECCR algorithm proposed in [1].

Steady-state dendrite growth is controlled by the states of dendrite tips, thus the morphology and growth kinetics of dendrite tip have always been a highly-studied issue in the field of dendrite growth. A crucial problem is to find out the factors influencing the morphology of dendrite tip, and to clarify their working mechanism. Since dendrite tip is at micrometer or even atomic scale, interface energy may probably plays an important role during the formation and stabilization of dendrite tips. Investigations of this issue at atomic scale is of significant because it can help us to clearly observe the morphology evolution of dendrite tip fundamentally, and specifically reveal the relationship between interface energy anisotropy and dendrite tip morphology. We investigate growth kinetics and morphology of dendrite tips at atomic scale by using phase-field crystal simulation. Atomic scale steady-state dendrite tips are obtained at different interface energy anisotropy and growth driving force. It is shown that the morphology of the forefront of dendrite tip varies periodically with time, which is related to the behavior of atom attachment during the steady-state growth of dendrite tip. Moreover, we demonstrate that growth driving force determines the overall shape of dendrite tip, but exerts little influence on the morphology of the forefront of dendrite tip which actually is determined by interface energy anisotropy. The dendrite tip of high interface energy anisotropy often presents sharp shape which is in favour of the formation of dendrite tip. Nevertheless, the shape of dendrite tip approaches to circle with decreasing magnitude of interface energy anisotropy, indicating that it is more difficult for the formation of dendrite tip under the circumstance of lower interface energy anisotropy.