The time-domain analysis on single-stage LLC power factor correction (PFC) converter is performed in this article, which reveals that PON mode may occur around the peak point of instantaneous input power and cause instability issue. In order avoid PON mode, the power boundary between PO and PON mode is calculated. For the common design methods of LLC PFC converter with average mode control strategy, the power boundary requires to be higher than the instantaneous input power at its worst point. However, higher power boundary brings higher rms value of resonant current, and causes higher reactive power loss. Therefore, it is a wasteful designing methodology, as the other points of instantaneous input power do not need such high power boundary. In order to improve the efficiency, a power boundary control strategy is proposed in this article. By controlling the input current to make the instantaneous input power follow the power boundary, PON mode is eliminated. Consequently, the required power boundary is overall reduced, and the minimal rms current of resonant tank can be achieved. Furthermore, the optimal parameter design is also presented. A 90–240 VAC input and 250 W/48 V output experimental prototype has been built to verify the analysis results.

The low-frequency oscillation (LFO) issue has occurred frequently in single-phase electric train and traction network interactive systems (hereinafter train-network systems). On a single-phase electric train, the second-order generalized integrator (SOGI) based phase-locked loops (PLLs) have been widely used and also have significant impacts on the LFO. In the SOGI-PLL, the frequency signal is dynamically feedback from PLL to SOGI for adapting frequency variety, which causes complex nonlinearity and may deteriorate the small-signal stability of train-network systems. In previous works, the small-signal impedance model of train supposes that the feedback frequency is quasi-steady-state, which ignores impacts of dynamic frequency feedback loop (DFFL) on system-level stability and causes inaccurate analysis results. Therefore, this article builds an improved impedance model of train in dq frame with considering the DFFL. Compared with fixed frequency feedback loop (FFFL) based SOGI-PLL, the DFFL-based SOGI-PLL results in an additional negative resistance behavior, which is more likely to induce the LFO. Moreover, based on proposed impedance model of train, three FFFL-based SOGI-PLLs are compared among the filtering capability, phase-locked error and negative resistance range. Finally, experiment results by hardware in the loop (HIL) platform and real low-power experimental platform are employed to verity the correctness of theoretical analysis results.

This paper proposes a hybrid control strategy for single-stage interleaved totem-pole LLC AC-DC converter. Since the totem-pole power factor correction (PFC) front-cell provides bi-direction inductor current, this converter is restricted by operating in continuous conduction mode (CCM) or critical conduction mode (CRM). The traditional control strategy shows variable duty cycle of the resonant tank voltage to the instantaneous AC input voltage. Therefore, the line-frequency variation from the AC input-end will be easily interfaced to the load, and brings a large fluctuation of output voltage. To tackle this issue, this paper proposed a novel control strategy that utilizes pulse width modulation (PWM) to shape the input current, and pulse frequency modulation (PFM) to regulate the output voltage. As a positive consequence, the large output voltage ripple brought by the variation of duty cycle for traditional PWM-based control is therefore effectively compensated by the variation of switching frequency, and the power decoupling between the AC pulse input power and constant DC output power is achieved. Numerical-based calculation method is utilized to develop the time-domain analysis of the converter, and a 500W/150V experimental prototype has been built to verify the proposed control strategy and theoretical analysis.

LLC resonant converter is a good candidate for single-stage AC-DC converter, which can realize both functions of power factor correction (PFC) and output voltage regulation, and contribute the improvement of conversion efficiency and power density. However, voltage mode controlled single-stage half-bridge LLC PFC converter has the disadvantages of large dead-zone of input current and hard-switching operation, since there is no current control. Besides, the voltage ripple of input filter brings asymmetrical operation of the converter between the positive and the negative half switching-cycle. In order to solve these problems, this paper presents a modified numerical calculation method, which can realize accurate characterization on the asymmetrical behavior of resonant current and voltage. Based on that, the switching frequency of LLC resonant converter for unity PF is derived. A method of fitting the switching frequency is proposed to simplify the circuit implementation. A 250W/48V prototype has been built to verify the analysis results. Experimental results show that the proposed voltage-mode variable-frequency controlled LLC resonant PFC converter can achieve high efficiency up to 94.08% and power factor of 0.95 over the input voltage range of 180~260VAC.

Frequency-domain admittance models have been widely utilized to analyze the low-frequency oscillation (LFO) issue in train-network systems. Since existing dq -frame or αβ-frame admittance models of train-network systems are usually built from ac-port, impacts of dc-side passive impedance on the LFO cannot be explicitly characterized. Additionally, admittance models built from ac-port are multiple input and multiple output (MIMO) systems, which is inconvenient for admittance measurement and stability margin estimation like a single input and single output (SISO) system. As a result, this paper builds a three-port admittance model of train-network systems. From the perspective of dc-port, a novel SISO stability criterion is derived to analyze the LFO. It is verified that the proposed SISO stability criterion and traditional MIMO stability criterion in the dq -frame are different reduced-order forms of the three-port admittance model, and all of them are equivalent under some preconditions. Moreover, impacts of dc-side passive impedance on the LFO are explicitly revealed based on the proposed SISO stability criterion, a virtual resistor control based on the feedback of dc-link capacitor current is then proposed to suppress the LFO. Finally, the experiment results by StarSim hardware in the loop (HIL) are employed to verify the correctness of theoretical analysis results.

Reducing carbon emissions is a crucial way to achieve the goal of green and sustainable development. To accomplish this goal, electric vehicles (EVs) are considered system‐schedulable energy storage devices, suppressing the negative impact of the randomness and fluctuation of renewable energy on the system's operation. In this paper, a coordination control strategy aimed at minimising the carbon emissions of a distribution network between EVs, energy storage devices, and static var compensators (SVCs) is proposed. A model‐free deep reinforcement learning (DRL)‐based approach is developed to learn the optimal control strategy with the constraint of avoiding system overload caused by random EV access. The twin‐delayed deep deterministic policy gradient (TD3) framework is applied to design the learning method. After the model learning is completed, the neural network can quickly generate a real‐time low‐carbon scheduling strategy according to the system operating situation. Finally, simulation on the IEEE 33‐bus system verifies the effectiveness and robustness of this method. On the premise of meeting the charging demand of electric vehicles, this method can optimise the system operation by controlling the charge‐discharge process of EVs, effectively absorbing the renewable energy in the system and reducing the carbon emissions of the system operation.