When a doubly-fed induction generator(DFIG)is connected to a weak grid,the coupling between the grid and the DFIG itself will increase,which will cause stability problems.It is difficult to maintain the tracking accur...When a doubly-fed induction generator(DFIG)is connected to a weak grid,the coupling between the grid and the DFIG itself will increase,which will cause stability problems.It is difficult to maintain the tracking accuracy and robustness of the phase-locked loop(PLL)in the weak grid,and the risk of instability of the current-controlled DFIG(CC-DFIG)system will increase.In this paper,a new type of voltage-controlled DFIG(VC-DFIG)mode is adopted,which is a grid-forming structure that can independently support the voltage and frequency with a certain adaptability in the weak grid.A small-signal impedance model of the VC-DFIG system is also established.The impedance of DFIG inevitably generates coupling with the grid impedance in the weak grid,especially in parallel compensation grids,and results in resonance.On the basis of the VC-DFIG,impedance stability analysis is performed to study the influences of the control structure and short-circuit ratio.Then,a feedforward damping method is proposed to modify the impedance of the VC-DFIG system at resonance frequencies.The proposed fractional order damping is utilized,which can enhance the robustness and rapidity of resonance suppression under parameter fluctuations.Finally,the experimental results are presented to validate the effectiveness of the proposed control strategy.展开更多
The utilization of wind generation equipment, such as DFIGs （double fed induction generators）, interconnected to islanded power generation and distribution systems is investigated in order to determine their effects...The utilization of wind generation equipment, such as DFIGs （double fed induction generators）, interconnected to islanded power generation and distribution systems is investigated in order to determine their effects on the overall system operating characteristics and stability. The use of a stable power station （with high speed machines） will be critical in achieving fast and reliable transient response to network events, in particular, when large transient loads are expected on a continuous basis, i.e., industrial mining and mineral processing equipment. Simulation results of this paper assist in understanding how small power stations and wind generation equipment respond to large transients in an islanded network. In particular, detailed simulations and analyses will be presented on impacts of distributed wind generation units （1.5 MW DF1G） on the stability of a small weak network. The novelty of this paper is on detailed analyses and simulation of weak networks with interconnects DFIG＇s including their impacts on system stability under various transient operating conditions.展开更多
新能源发电并网系统的振荡问题是电网关注的重要研究课题,揭示振荡发生的机理并明确影响振荡的关键因素是首要研究任务。为此,建立了含静止无功发生器(static var generator,SVG)的光伏并网系统的小信号状态空间模型,通过特征值分析研...新能源发电并网系统的振荡问题是电网关注的重要研究课题,揭示振荡发生的机理并明确影响振荡的关键因素是首要研究任务。为此,建立了含静止无功发生器(static var generator,SVG)的光伏并网系统的小信号状态空间模型,通过特征值分析研究了光伏并网系统的主导振荡模式及与其强相关的动态环节,分析了交流侧系统对光伏并网系统振荡特性的影响;采用频率扫描法从电路谐振角度对光伏并网系统次同步振荡(subsynchronous oscillation,SSO)现象进行了解释,并通过时域仿真验证了特征模式分析结果的正确性。研究结果表明:含SVG的光伏并网系统存在一个主导SSO模式,光伏电站并网变流器的控制参数对该模式的稳定性具有重要影响;此外,SVG与光伏发电系统之间也存在相互控制作用,需要合理设计二者的控制参数,以提高系统整体的稳定性。研究内容可为光伏并网系统的运行和控制设计提供指导。展开更多
当前,深度主动学习(Deep Active Learning,DAL)在分类数据标注工作中获得成功,但如何筛选出最能提升模型性能的样本仍是难题.本文提出基于弱标签争议的半自动分类数据标注方法(Dispute about Weak Label based Deep Active Learning,DWL...当前,深度主动学习(Deep Active Learning,DAL)在分类数据标注工作中获得成功,但如何筛选出最能提升模型性能的样本仍是难题.本文提出基于弱标签争议的半自动分类数据标注方法(Dispute about Weak Label based Deep Active Learning,DWLDAL),迭代地筛选出模型难以区分的样本,交给人工进行准确标注.该方法包含伪标签生成器和弱标签生成器,伪标签生成器是在准确标注的数据集上训练而成,用于生成无标签数据的伪标签;弱标签生成器则是在带伪标签的随机子集上训练而成.弱标签生成器委员会决定哪些无标签数据最有争议,则交给人工标注.本文针对文本分类问题,在公开数据集IMDB(Internet Movie DataBase)、20NEWS(20NEW Sgroup)和chnsenticorp(chnsenticorp_htl_all)上进行实验验证.从数据标注和分类任务的准确性2个角度,对3种不同投票决策方式进行评估.DWLDAL方法中数据标注的F1分数比现有方法Snuba分别提高30.22%、14.07%和2.57%,DWLDAL方法中分类任务的F1分数比Snuba分别提高1.01%、22.72%和4.83%.展开更多
针对在弱电网下直驱风电机组引起的次同步振荡(subsynchronousoscillation,SSO)现象,提出基于一阶总扰动偏差控制的微分前馈线性自抗扰控制器(linear active disturbance rejection control,LADRC),采用全改进LADRC控制策略抑制SSO现象...针对在弱电网下直驱风电机组引起的次同步振荡(subsynchronousoscillation,SSO)现象,提出基于一阶总扰动偏差控制的微分前馈线性自抗扰控制器(linear active disturbance rejection control,LADRC),采用全改进LADRC控制策略抑制SSO现象(“全”是指电压外环、电流内环以及PLL锁相环3个环节都采用相应的控制)。首先,建立直驱风电机组并网数学模型;其次,结合风电机组并网系统对改进LADRC控制器进行设计并对其进行特性分析,该控制器相较于传统LADRC,不仅减小系统的跟踪误差且抗干扰性能更强;最后,通过PSCAD/EMTDC仿真软件将本文策略与全PI、全传统LADRC进行仿真对比。结果表明:相较于全传统LADRC,本文方法在降低1.62%超调量的同时,缩短0.129 s系统调节时间,有效抑制SSO现象并且具有较好的适应性。展开更多
新能源汇集经柔性直流输电(voltage source converter based high voltage direct current,VSC-HVDC)技术送出是促进新能源消纳的有效方式。但新能源渗透率的持续增加导致电网强度不断下降,采用传统跟网型(grid-following,GFL)换流技术...新能源汇集经柔性直流输电(voltage source converter based high voltage direct current,VSC-HVDC)技术送出是促进新能源消纳的有效方式。但新能源渗透率的持续增加导致电网强度不断下降,采用传统跟网型(grid-following,GFL)换流技术已无法满足系统稳定运行需求。为提高系统弱电网适应性,满足大规模新能源接入弱同步支撑柔直系统应用场景需求,提出在柔直系统送端换流站采用VSG控制策略。首先,建立整流侧控制小信号数学模型,利用根轨迹法深入研究虚拟阻抗对系统稳定性的影响。其次,提出利用交流电压变化率及电压差值等电气量构建虚拟电抗自适应调整项的改进VSG控制算法,在保证系统等效阻抗呈感性的同时,可提高送端交流系统的等效短路比,达到改善系统整体性能的效果。最后,通过PSCAD/EMTDC电磁暂态仿真验证所提控制策略的有效性。展开更多
基金supported by the National Natural Science Foundation of China(No.51877063).
文摘When a doubly-fed induction generator(DFIG)is connected to a weak grid,the coupling between the grid and the DFIG itself will increase,which will cause stability problems.It is difficult to maintain the tracking accuracy and robustness of the phase-locked loop(PLL)in the weak grid,and the risk of instability of the current-controlled DFIG(CC-DFIG)system will increase.In this paper,a new type of voltage-controlled DFIG(VC-DFIG)mode is adopted,which is a grid-forming structure that can independently support the voltage and frequency with a certain adaptability in the weak grid.A small-signal impedance model of the VC-DFIG system is also established.The impedance of DFIG inevitably generates coupling with the grid impedance in the weak grid,especially in parallel compensation grids,and results in resonance.On the basis of the VC-DFIG,impedance stability analysis is performed to study the influences of the control structure and short-circuit ratio.Then,a feedforward damping method is proposed to modify the impedance of the VC-DFIG system at resonance frequencies.The proposed fractional order damping is utilized,which can enhance the robustness and rapidity of resonance suppression under parameter fluctuations.Finally,the experimental results are presented to validate the effectiveness of the proposed control strategy.
文摘The utilization of wind generation equipment, such as DFIGs （double fed induction generators）, interconnected to islanded power generation and distribution systems is investigated in order to determine their effects on the overall system operating characteristics and stability. The use of a stable power station （with high speed machines） will be critical in achieving fast and reliable transient response to network events, in particular, when large transient loads are expected on a continuous basis, i.e., industrial mining and mineral processing equipment. Simulation results of this paper assist in understanding how small power stations and wind generation equipment respond to large transients in an islanded network. In particular, detailed simulations and analyses will be presented on impacts of distributed wind generation units （1.5 MW DF1G） on the stability of a small weak network. The novelty of this paper is on detailed analyses and simulation of weak networks with interconnects DFIG＇s including their impacts on system stability under various transient operating conditions.
文摘新能源发电并网系统的振荡问题是电网关注的重要研究课题,揭示振荡发生的机理并明确影响振荡的关键因素是首要研究任务。为此,建立了含静止无功发生器(static var generator,SVG)的光伏并网系统的小信号状态空间模型,通过特征值分析研究了光伏并网系统的主导振荡模式及与其强相关的动态环节,分析了交流侧系统对光伏并网系统振荡特性的影响;采用频率扫描法从电路谐振角度对光伏并网系统次同步振荡(subsynchronous oscillation,SSO)现象进行了解释,并通过时域仿真验证了特征模式分析结果的正确性。研究结果表明:含SVG的光伏并网系统存在一个主导SSO模式,光伏电站并网变流器的控制参数对该模式的稳定性具有重要影响;此外,SVG与光伏发电系统之间也存在相互控制作用,需要合理设计二者的控制参数,以提高系统整体的稳定性。研究内容可为光伏并网系统的运行和控制设计提供指导。
文摘当前,深度主动学习(Deep Active Learning,DAL)在分类数据标注工作中获得成功,但如何筛选出最能提升模型性能的样本仍是难题.本文提出基于弱标签争议的半自动分类数据标注方法(Dispute about Weak Label based Deep Active Learning,DWLDAL),迭代地筛选出模型难以区分的样本,交给人工进行准确标注.该方法包含伪标签生成器和弱标签生成器,伪标签生成器是在准确标注的数据集上训练而成,用于生成无标签数据的伪标签;弱标签生成器则是在带伪标签的随机子集上训练而成.弱标签生成器委员会决定哪些无标签数据最有争议,则交给人工标注.本文针对文本分类问题,在公开数据集IMDB(Internet Movie DataBase)、20NEWS(20NEW Sgroup)和chnsenticorp(chnsenticorp_htl_all)上进行实验验证.从数据标注和分类任务的准确性2个角度,对3种不同投票决策方式进行评估.DWLDAL方法中数据标注的F1分数比现有方法Snuba分别提高30.22%、14.07%和2.57%,DWLDAL方法中分类任务的F1分数比Snuba分别提高1.01%、22.72%和4.83%.
文摘新能源汇集经柔性直流输电(voltage source converter based high voltage direct current,VSC-HVDC)技术送出是促进新能源消纳的有效方式。但新能源渗透率的持续增加导致电网强度不断下降,采用传统跟网型(grid-following,GFL)换流技术已无法满足系统稳定运行需求。为提高系统弱电网适应性,满足大规模新能源接入弱同步支撑柔直系统应用场景需求,提出在柔直系统送端换流站采用VSG控制策略。首先,建立整流侧控制小信号数学模型,利用根轨迹法深入研究虚拟阻抗对系统稳定性的影响。其次,提出利用交流电压变化率及电压差值等电气量构建虚拟电抗自适应调整项的改进VSG控制算法,在保证系统等效阻抗呈感性的同时,可提高送端交流系统的等效短路比,达到改善系统整体性能的效果。最后,通过PSCAD/EMTDC电磁暂态仿真验证所提控制策略的有效性。
