Search

Aim to improve the power efficiency of the dual-active-bridge (DAB) dc–dc converter, an efficiency optimization scheme with triple-phase-shift (TPS) modulation using reinforcement learning (RL) is proposed in this article. More specifically, the Q-learning algorithm, as a typical algorithm of the RL, is applied to train an agent offline to obtain an optimized modulation strategy, and then the trained agent provides control decisions online in a real-time manner for the DAB dc–dc converter according to the current operating environment. The main objective is to obtain the optimal phase-shift angles for the DAB dc–dc converter, which can achieve the maximum power efficiency by reducing the power losses. Moreover, all possible operation modes of the TPS modulation are considered during the offline training process of the Q-learning algorithm. Thus, the cumbersome process for selecting the optimal operation mode in the conventional schemes can be circumvented successfully. Based on these merits, the proposed efficiency optimization scheme using the RL can realize the excellent performances for the whole load conditions and voltage conversion ratios. Finally, a 1.2-KW prototyped is built, and the simulation and the experimental results demonstrate that the power efficiency can be improved by using the optimization scheme based on the RL.

In recent years, resonant converters have been widely used in the fields of electric vehicles and renewable energy systems due to their excellent characteristics in terms of high energy density, high energy efficiency, wide output range, and wide ZVS range. Conventional resonant converters generally adopt pulse frequency modulation (PFM) in a wide range of output applications. This means the frequency varies over a wide range, which results in problems such as high switch loss, large noise, and high EMI. To solve this problem, this paper proposes a hybrid fixed-frequency modulation strategy applied to a wide range of outputs based on the series resonant converter, where asymmetrical voltage-cancellation control (AVC control) and its improved strategy (IAVC control) are applied in the normalized gain ranges of [0.5, 1] and [0, 0.5]. The proposed modulation strategy achieves lower switching loss, higher efficiency, and better ZVS characteristics in wide voltage output and load ranges. Under the same output voltage and power, to realize ZVS, the proposed hybrid modulation has a smaller switching frequency and resonant current, which improves efficiency and reduce losses, especially in the case of low voltage gain and light load conditions. Finally, experimental data obtained from an 80 V input and a 800 W power max prototype circuit are included to verify the proposed hybrid modulation strategy.

This article proposed a current sharing strategy for three-phase series capacitor boost converter for continuous conduction mode operation. The phase current can be automatically balanced based on the charge-balance principle, without any additional sensors or devices. And a uniform output voltage expression in full duty ratio range can be obtained. The mechanism of unbalanced current is disclosed. A current sharing strategy is proposed by adjusting the duty ratio in adjacent phase. The selection of capacitor and inductor is discussed via the analysis of topological operation and current deviation. The related characteristics are presented, including uniform output expression, continuous duty ratio operation, insensitivity to inductor parameters, loos analysis, and device stress. Experimental results on 100-W prototype are provided to confirm the feasibility of the proposed strategy.

To better predict the high-frequency switching operation of the half-bridge circuit in power converters, the value of the parasitic elements of these devices must be accurately evaluated. A new MOSFET-based half-bridge circuit parasitic inductances extraction method using two-port S-parameters is proposed in this paper. By changing the terminal connection of the half-bridge circuit, we can treat it as several different two-port networks, and then detailed network analysis can be performed on it. The parasitic parameters of multi-terminal actual circuits such as half-bridge can be extracted accurately and quickly through the simple measurement steps and calculations of the proposed method.

High voltage step-down converters are very popular in distributed power systems, voltage regular modules, automatic vehicles, etc. To avoid extreme duty cycles, a series capacitor-based buck converter with a coupled inductor is proposed in this paper. In this converter, the voltage stresses of the power switches are clamped to half of the input voltage. Thus, low voltage rating MOSFETs with a low Rds,on could be utilized to improve the performance of the proposed converter. Moreover, synchronous MOSFETs are also used to reduce the conduction losses of the switches. Zero voltage switching (ZVS) for the main power switches is achieved using the coupled-inductor technique. In addition, through adding a bypass flowing path, a negative current is established to guarantee the ZVS condition of the power switches. The operating principle of the proposed converter is described in this paper, and the DC voltage gain ratio, automatic current sharing, soft switching condition and design guideline of the critical parameters are also given. Finally, experimental results obtained from a 60 V/200 W prototype are presented to verify the analysis of the proposed converter.

In order to guarantee the system reliability and enhance post-fault maintenance efficiency of inverters, this paper presents an current observer-based online open-switch fault diagnosis method for the voltage-source inverters (VSI). Analyzing the state space model of VSI when fault occurs, a state observer of the system is built to estimate phase currents. Once the open-switch fault occurs, the current residual vector between the estimated currents on observer and the measured currents, and fault diagnosis is realized by exploiting the featured amplitude and phase of the current residual vector. In order to realize on-line diagnosis with high reliability and speed, the average value of current residuals in a fundamental period is obtained and Clark transformation is performed so as to acquire vector angle. In addition, the sum of absolute values of measured currents is normalized to serve as the sign of fault occurrence. The simulation and experimental results show the effectiveness of proposed method.

In this paper, a series-capacitor-based interleaving buck power factor correction (PFC) converter is proposed, the intermediate energy storage capacitance of which is operated at discontinuous capacitor voltage mode (DCVM), and automatic power factor (PF) correction for this converter is obtained. Moreover, the voltage across the series capacitor of the proposed PFC converter is clamped at the input voltage, and thus peak voltage stress existed in the traditional DCVM buck PFC could be relieved. By effectively regulating the charge/discharge time for the series capacitor, ultrahigh step-down conversion ratio of the proposed converter is achieved, the conversion ratio of which is independent of the duty cycle and only related to the switching frequency and circuit parameters. Therefore, high PF can be obtained due to very narrow deadtime of the input current. Meanwhile, the proposed PFC converter can realize the soft turn- on of part power switch and soft turn- off of part power diodes. In addition, comparing with the traditional discontinuous inductor current mode PFC, the input current of the proposed PFC converter is continuous so that the peak value and root-mean-square value of input current can be reduced. The operational principle, performance analysis, and parameter design principle are given in this paper. The analysis results show that the PF and total harmonic distortion are also independent of the duty cycle. Finally, a 220-Vac input, 60 W/15 V output experimental prototype is built to verify the validity of the theoretical analysis.

Aiming at the limitations of traditional full-bridge dc/dc converters, such as, narrow soft switching range, large circulating loss during the freewheeling period, and voltage oscillation on the rectifier diodes, in this paper, a series capacitor-based full-bridge dc/dc converter is presented to reduce the voltage stress of power switches to half of input voltage, and by introducing the asymmetric pulsewidth modulated (APWM) scheme to the proposed converter, zero-voltage switching for all the switches in the proposed converter can be obtained. Moreover, due to the reduced voltage turned-OFF stress for switches, the energy of commutation process for switches is reduced and turned-OFF switching losses can be decreased. Therefore, wide soft switching range for the proposed converter could be achieved on the condition of variable input voltage and load power. Furthermore, a simple auxiliary circuit on the secondary side of the transformer is utilized to clamp the voltage stress of rectifier diodes and voltage oscillation is eliminated. Moreover, during the freewheeling period, the current through the primary side of the transformer can be decay to zero; circulating losses are eliminated and zero current switching for the lagging switches could be achieved. The circuit configuration, operating principle, and relevant analysis results of the proposed converter are analyzed, and a design example is given in this paper. Finally, experimental results from a 400-W 300–380-V/42-V prototype are presented to verify the analysis of the proposed converter.

To improve the speed and position observation performance of a Surface-Mounted Permanent Magnet Synchronous Motor (SM-PMSM), an improved sliding mode observer (SMO) based on the adaptive synchronization filter (ASF) and quadrature phase-lock loop (QPLL) is proposed. An adaptive synchronous filter based on feedback rotor position estimation from a QPLL is proposed to extract the fundamental signal of the back electromotive force (back-EMF). Then, a QPLL is linked to calculate the rotor position and speed according to the fundamental signal of the back-EMF, in order to effectively suppress the estimated harmonics and chattering and improve the observation accuracy of speed and position. Finally, the effectiveness and feasibility of the proposed observer approach are verified by simulation and dSPACE.