The significant stochasticity introduced by renewable energy and flexible loads challenges traditional deterministic stability methods in power system. To address the stability challenges posed by stochastic disturbances, by integrating physics-informed neural network (PINN) with power system characteristic equation, a characteristic oriented physics-informed neural network (COPINN) method is proposed to enhance the assessment accuracy of stability probability in stochastic power system. Unlike traditional PINNs that treat all regions equally, COPINN introduces a spatial attention mechanism guided by the physical characteristic curves of the system. These characteristic curves, derived from the governing equations of stability probability, pinpoint the critical regions in the state-space that most influence the stability outcome. Based on the semi-physical power system simulation platform, the operating data of the power system are obtained and its stability probability is statistically acquired. Following this data acquisition, the physical model of the stochastic power system stability probability is deduced through the stochastic averaging method. Finally, an efficient assessment of the stability probability is achieved through the implementation of the physics-informed neural network incorporating stochastic power system characteristic equation. This method incorporates the stability probability characteristic equation of the power system into the physics-informed neural network through a characteristic oriented module and characteristic weight parameters, enabling the network to focus on the key characteristic regions that determine the stability probability of the power system. A further residual-driven adaptive sampling strategy enhances the density of data points in regions with high error rates during training, thereby achieving efficient allocation of computational resources. Experimental analysis demonstrates that, compared to conventional methods, the proposed COPINN approach can accurately capture the evolution of stability probabilities in power system, and effectively enhance the assessment accuracy of stability probabilities in stochastic power system. Furthermore, the proposed COPINN method also reveals trends in probabilistic stability evolution across key parameters, stochastic initial energies, and short-circuit faults, which provides a certain reference for the probabilistic stability analysis of stochastic power system.

The five-level switched-capacitor hybrid clamped converter (5L-SCHC) exhibits lower low-frequency voltage ripple in DC-link capacitors and voltages passive self-balancing of switched-capacitors (SCs), offering clear advantages over other multilevel clamped converters. However, the discrete output characteristics of the 5L-SCHC circuit make it difficult to maintain capacitor voltages balancing with traditional modulation methods, leading to suboptimal harmonic performance. To address this issue, this paper introduces an improved carrier redistribution pulse width modulation (ICR-PWM). Through carrier redistribution and variable reference modulation voltage strategy, the optimal distribution of output voltage pulses is achieved within two carrier cycles, thereby enhancing the output performance of the 5L-SCHC converter. On this basis, duty cycle modification is applied to balance the DC-link center capacitor voltage in the first carrier cycle, while the SC voltages are balanced in the other one. Additionally, the upper and lower DC-link capacitor voltages are balanced over the two carrier cycles via optimal zero-sequence voltage (ZSV) injection. Finally, validation is conducted on the experimental platform, experimental results validate that the proposed ICR-PWM method ensures capacitor voltages balance under various conditions, exhibiting remarkably improved harmonic performance over traditional methods.

The hybrid clamped converter cannot suppress low-frequency voltage fluctuations of the dc-link capacitors at low power factors. Rather, it exhibits a slow dynamic response and high output harmonics in such scenarios. To address this limitation, this study presents a novel hybrid modulation method that maintains the natural capacitor voltage balance while combining two basic modulation strategies: one minimizes neutral-point (NP) current to suppress low-frequency fluctuations, and the other maximizes NP current to enhance the dynamic response. The method further bolsters harmonic performance through carrier redistribution. Then, the proposed method dynamically switches between two basic modulation methods based on real-time NP current optimization while calculating the required zero-sequence voltage for active voltage balancing. The proposed hybrid modulation method thus inherits the merits of these basic modulations. Finally, the method is validated under various experimental conditions, with the results highlighting its advantages.

Carrier-based modulation strategies with natural capacitor voltage balancing characteristics are widely employed in five-level stacked multicell (SM) converters, owing to their straightforward implementation and reduced capacitor requirements. However, active voltage balancing of DC-link capacitors usually requires zero-sequence voltage (ZSV) or harmonic voltage injection, which introduces significant common-mode voltage (CMV) and degrades system performance. To overcome this limitation, this article proposes a variable reference voltage pulsewidth modulation (PWM) (VRV-PWM) strategy. The proposed method leverages a novel neutral-point (NP) current expression, derived from the SM converter's switching states, to regulate the NP current via duty cycle adjustment, thereby operating without ZSV or harmonic injection. Consequently, this approach maintains an average phase voltage free of low-frequency components while achieving zero average CMV, which significantly suppresses the instantaneous CMV and enhances the output phase voltage performance. An optimal NP current regulation scheme is further derived to achieve independent and decoupled voltage control across all capacitors. Theoretical analysis shows that the proposed strategy complements traditional methods in NP voltage balancing, effectively overcoming their drawback in suppressing low-frequency voltage fluctuations under low power factor conditions. Experimental results confirm the effectiveness of the proposed VRV-PWM, demonstrating its superior performance in CMV reduction and output phase voltage harmonic quality.

To address output stability and transmission efficiency degradation in WPT systems caused by coupling misalignment and load variations, this paper introduces a Predictive Neural Network-based Deadbeat Predictive Control (PNN-DPC) approach. Initially, a transfer function model is constructed via the T-type equivalent circuit of the dual-LCC resonant topology, and the mathematical link between the phase-shift angle and output voltage is derived through output performance parametric analysis. Subsequently, the model is reformulated to reduce traditional model predictive control's computational complexity and parameter reliance. A neural network estimator is incorporated to identify system uncertainties and external disturbances, boosting controller robustness. Finally, a predictive controller merging deadbeat control with neural network-estimated model data is designed, and Lyapunov stability theory confirms the error system's asymptotic stability under input constraints. Experiments show that PNN-DPC maintains constant voltage output across diverse coupling coefficients and loads, outperforming conventional methods in dynamic response and disturbance rejection.

This study proposes a novel carrier redistribution pulse-width modulation (CRPWM) technique and its associated active voltage balancing method with low-frequency voltage ripple suppression for the five-level stacked multicell (SM) converter. While CRPWM offers simplicity and ease of implementation with optimal harmonics, it encounters significant neutral point (NP) voltage imbalance issues, particularly at low power factors. NP voltage imbalance is a primary cause of low-frequency voltage ripples in DC-link capacitors. Therefore, the proposed CRPWM utilizes new switching states to minimize NP current generation and ensures the natural voltage balance of floating capacitors to minimize voltage ripples. Consequently, NP current significantly diminishes at low power factors, and unbalanced NP current is almost completely suppressed under the active voltage balancing method theoretically. The proposed CRPWM method, with lower unbalanced NP current and equivalent line voltage harmonic performance compared with other CRPWM methods, exhibits the lowest low-frequency voltage ripples with optimal output performance. Experimental results reveal about 88% and 65% reduction in low-frequency voltage ripples, and about 0.5% and 0.02% reduction in line voltage total harmonic distortion (THD), particularly at lower power factors and highest modulation index with fundamental frequencies of 10 and 50 Hz, respectively.

Clamped multilevel inverters suffer significantly from neutral point (NP) voltage unbalancing in low power factor applications, which will lead to severe device overvoltage problems and affect its output performance. The five-level hybrid clamped (HC) inverter is a state-of-the-art topology with additional redundant switching states, which makes it possible to suppress the NP low-frequency voltage ripples better. Therefore, a novel and simple modulation for five-level HC inverter with enhanced NP voltage balancing capability is proposed. Based on the influence of NP currents on the dc-link capacitor voltages, the different switching states combinations with capacitor voltages naturally balancing constraints are analyzed comprehensively, and the relationship between the duty ratio of average NP currents and the reference modulation voltage is designed as a bi-triangular function to minimize the currents flowing out of the NPs. Compared with traditional modulation, the equivalent adjustment ability of NP currents is greatly enhanced by the zero-sequence voltage injection, and the low-frequency voltage ripples of DC-link capacitors are suppressed significantly at whole power factors. Moreover, the active balancing methods are explored to achieve independent and decoupling control of all capacitor voltages. Simulations and experiments verify the viability of the modulation strategy and the capacitor voltage balancing method.

Neutral point (NP) voltage imbalance blocks the deployment of carrier-based clamped converters for specific scenarios due to the limited adjustment capability of NP current by zero-sequence voltage (ZSV) injection. Therefore, a novel six-level switched-capacitor (SC) hybrid clamped (6L-SCHC) converter is proposed, which has compromised voltage stress, and utilizes the discrete voltage boost and capacitor voltage self-balancing features of the SC circuit to simplify control complexity. On this basis, a carrier-based hybrid modulation is developed which simplifies the complexity of SC voltage control, and ensures the voltage natural balance of SC and DC-link center capacitor to minimize voltage ripples. Furthermore, a reasonable functional relationship between average NP current and reference modulation voltage is explored. As the result, the proposed 6L-SCHC converter has extremely powerful NP voltage balancing capability compared to other carrier-based clamped converters in theory, especially at high modulation indexes and low power factors. Capacitor voltage ripple models are also derived to achieve voltage independent controls of all capacitors. Comparative simulations and experiments demonstrate that the proposed topology and modulation greatly reduce the lowfrequency voltage ripples of DC-link upper and lower capacitors, and provide excellent NP voltage balancing performance over a wide range of modulation indexes and power factors.

针对光伏、风机等具有不受控性和波动性的分布式发电系统(DGS)大规模并入电网造成的储能单元出力不均匀和荷电状态(SOC)不均衡的问题,基于多代理系统提出了一种基于梯度补偿的改进型一致性算法.将DGS、退役动力电池储能系统(DPBESS)和电网共同组建成多代理微电网系统.在改进一致性算法基础上设计了SOC均衡控制方法,加快了无中心控制器的储能系统SOC均衡速度,保证了微电网内能量的供需平衡.通过仿真验证了所提控制方法的有效性.

基于MMHC-BESS,将MPC算法与占空比调制思想结合,提出一种可有效减小系统计算负担的模型预测优化控制策略.控制策略利用MPC算法实现功率控制,通过变换器的离散模型反推出最优输出电压,采用载波移相调制方法实现开关信号输出.同时考虑电网电压不平衡状况,利用MPC算法对负序电流及功率波动进行抑制,具备较好的冗余容错运行能力.最后通过仿真模型验证了所提控制方法的有效性和可行性.

针对五电平有源中点箝位型(ANPC)变换器共模电压大的问题,提出一种抑制共模电压的极简空间矢量控制.首先将参考电压投影到线电压坐标系中,得到对应区域内的所有矢量,分析矢量与共模电压和中点电压之间的联系,然后以最小化共模电压为目标对矢量选择范围进行合理约束的同时,保证剩余矢量能对中点电压进行均衡控制.在此基础上,针对中点电压控制提出一种极简序列选择法,从而降低空间矢量调制策略的复杂度.最后对所提共模抑制算法的正确性和有效性进行了实验验证.

The neutral-point-clamped (NPC) multilevel converter generally adopts the zero-sequence voltage injection method to realize the voltage balancing of the dc link capacitors. However, the relationship between the reference modulation voltage and the duty ratio of phase neutral point current is often a triangle or trapezoidal function. When the power factor is low, the required optimal zero-sequence voltage will be out of bounds, which will lead the capacitor voltage ripples to become larger or even divergent. Therefore, a novel switched-capacitor-based hybrid clamped (SCB-HC) converter is proposed in this article. The new redundant switching states are obtained by introducing the boost characteristics of the switched-capacitor (SC) circuit. A hybrid phase-shifted pulsewidth modulation (PS-PWM) is developed to extend its functional relationship. This ensures that the peak current that flows out of the neutral points is fewer, and then, the capacitor voltage drift is reduced significantly. Moreover, the proposed hybrid PS-PWM method not only ensures that the SC circuit switches series and parallel patter once during a carrier period but also achieves the voltage balancing of SC and dc link midpoint capacitor naturally under ideal conditions. On top of that, the capacitor voltage balancing method is designed to ensure independent and decoupled control of all capacitor voltages. Simulations and experiments show that the proposed topology, the modulation method, and the voltage control method are practical and feasible for wide power factor applications.

在高比例可再生能源渗透环境下,电力系统呈现低惯性阻尼特性,传统基于PI控制器的同步参考坐标系锁相环更易受到新能源波动性的影响,造成锁相输出结果较差,从而引发跟网型风电变流器的一系列振荡问题.提出基于一阶线性自抗扰控制器的锁相环,通过对比分析一阶线性自抗扰锁相环和传统锁相环的频域特性,说明其抗扰性能的优势.提出一种简易的参数设计方法,简化锁相环参数设计.在相同带宽尺度比较下说明高阶自抗扰控制器应用于并网变流器锁相环的局限性.通过锁相环仿真和基于混合多电平并网变流器的实验结果表明,所提自抗扰锁相环在受扰动环境下具有更准确的锁相结果,更加适用于高比例可再生能源参与的新型电力系统.

针对智能煤矿建设的需求和煤矿井下电网存在的问题,提出了一种含有储能模块的储能型模块化多电平混合式配电变压器(MMC-HDT).首先,对提出的储能型MMC-HDT进行建模,并考虑电网故障和储能荷电状态(SOC)给出四种工作运行模式.其次,根据MMC-HDT交流侧的端口受控耗散哈密顿模型(PCHD)设计无源控制器.最后,设计四种工作运行模式的切换控制策略.实验结果表明,所提控制策略在储能型MMC-HDT中,馈网电流稳态总谐波畸变率进一步降低,有效限制电压故障时馈网电流无限制突增;在任何电网电压故障状态下,均可以实现负荷电压保持在额定值处,可以实现不间断供电的功能,提高了配电系统的可靠性.

针对电压型无线电能传输系统(wireless power transfer,WPT)在实际应用中受外界或系统内部扰动时出现输出电压不稳定以及如何保证控制器对扰动情况迅速做出反应的问题,提出一种基于线性自抗扰控制(linear active disturbance rejection control,LADRC)的WPT系统.首先根据WPT系统SS型谐振网络的T型等效电路建立传递函数模型,分析了各个参数对系统输出的影响,并推导出移相角度与输出电压的关系;其次结合系统模型及自抗扰控制理论,设计一阶LADRC控制器对系统输出电压进行闭环控制,并确定各个环节待整定参数的调节方向;最后搭建了基于LADRC的无线电能传输装置,在多种扰动的情况下比较了 LADRC控制器与PI控制器的控制效果.实验结果证明,LADRC控制器能够更好的应对系统扰动的不确定性,保证了 WPT系统的恒压输出,具有良好的控制特性.

模块化多电平复合变换器电池储能系统(modular multi-level hybrid convert-battery energy storage system,MMHC-BESS)适用于中低压电网,有利于解决可再生能源并网问题.但随着子模块数量的增加,系统的可靠性面临着巨大的挑战.为增强储能系统的容错运行能力,针对子模块故障下的不平衡运行状态,提出无硬件冗余容错的控制策略,基于载波移相调制,通过调制波重构算法维持线电压平衡及并网功率恒定.此外,为提升储能电池容量利用率,针对容错运行模式下的荷电状态(SOC)均衡策略进行了研究,通过零序电压注入及动态调节相内SOC均衡因子构成相间、相内两级SOC均衡策略.最后基于PLECS软件搭建仿真模型,验证了所提策略的可实施性及有效性.

由于电网中可再生能源发电的间歇性和易变性、用户侧负荷的不确定性以及能量的双向性,需要一种合理的微电网协同控制方案来实现多BESS微电网的可靠运行.鉴于模块化多电平复合变换器(modular multi-level hybrid converter,MMHC)电池储能系统(battery energy storage systems,BESS)的效率高、成本低等优点,根据MMHC-BESS的拓扑结构对其进行了建模和SOC估计,并在此基础上提出了一种MAS控制微电网MMHC-BESS PQ控制的SOC协同控制方案,实现了并网微电网内能量的供需平衡、容量不同的MMHC-BESS的SOC协同控制以及MMHC-BESS逆变器输出功率的限幅,最后通过仿真验证了所提方法的有效性.

Inductively coupled power transmission (ICPT) systems cause fluctuations in output power and output voltage when the coil is offset.The existing anti-misalignmemt methods have the problems of relying too much on the system modeling and poor adaptivity. Most of the methods do not consider the impact on the system output characteristics when the coupling coefficient changes continuously. To address the above problems, taking the dual-coupled LCL topology ICPT system based on DDQ coils as the research object, the paper proposes a strong anti-misalignmemt method based on fuzzy adaptive control in variable universe. Firstly, the output power expression of the dual-coupled LCL topology ICPT system is derived from the circuit analysis, and the relationship between the output power, the coupling coefficient and system parameters is obtained. Secondly, by applying the finite element analysis software ANSYS, a three-dimensional magnetic field modeling of the DDQ coil is carried out to obtain the correspondence between the coupling coefficient and the coil offset. On the basis, the corresponding values of the three sets of offsets and the coupling coefficients are used as the data, and the square sum of deviations of the system output power fluctuations is taken as the objective function. Hence, the parameter optimization method of ICPT system based on adaptive particle swarm is proposed to obtain a set of optimal parameter values of the system under the minimum output power fluctuation and improve the anti-misalignmemt performance of the system to a certain extent. Finally, a fuzzy adaptive control method based on variable universe is used to achieve the purpose of quickly adjusting the load terminal voltage by dynamically adjusting the correction value of the PID control coefficient to make the system output higher power. The simulation results show that the method solves the problem of poor adaptivity of existing methods, adapts well to the working conditions with continuous changes of coupling coefficients, has better adaptability and control effect, improves the strong anti-misalignmemt performance of ICPT system, and maintains the basic constant output voltage at the load side.

以能源互联网的最核心环节—能源路由器为研究背景,将前端级联H桥和双有源全桥(DAB)级分别进行独立控制时,将DAB级等效为输入独立输出并联拓扑进行研究.为提高其输出端负载情况突变及输入电压突变时输出电压的动态性能,借鉴直接功率思想,结合闭环控制和前馈控制,以H桥内的超前桥臂与滞后桥臂之间的移相角为控制变量,提出了一种复合控制策略.通过实验将所提复合控制与传统电压闭环控制进行性能对比验证.实验表明:该控制方法极大提升了变换器的动态性能,其控制实现也相对简单且对电路参数依赖性较小,有较好的兼容性和可移植性,在其他类型变换器动态响应研究中有借鉴的潜力.

针对无线电能传输(wireless power transfer,WPT)系统处于不同负载、耦合系数和频率的运行状况进行分析,基于T型等效电路和二端口网络,得到了 SS、LCL和LCC 3种谐振网络的传递函数模型.将3种谐振网络的传输特性进行分析,实验结果表明3种谐振网络均为恒流输出,其中LCC网络具有随负载和耦合系数变化呈单调增的关系,有效地解决SS和LCL网络中存在传输功率过小等问题,并抑制系统处于偏谐振或低耦合状态下出现功率过载的情况.最后搭建了一台基于移相全桥(phase shift full bridge,PSFB)控制的LCC型WPT实验样机,能适应宽负载变化范围,当逆变器处于不同的移相角下均能保持零相角(zero phase angle,ZPA)的条件,使得系统具有较高的传输效率.