适用于大容量架空线输电的C-MMC型柔性直流技术研究

发布时间：2018-09-16 20:55

【摘要】：相比较于传统的两电平或三电平换流器,模块化多电平换流器(Modular Multilevel Converter, MMC)具有输出电压波形质量高、扩展性好、低损耗和开关频率低等优点,已成为柔性直流输电领域的主流拓扑。早期MMC的桥臂由大量半桥子模块(HBSM)级联构成,无法自主切断直流故障电流,难以扩展到架空线领域。采用箝位双子模块的MMC(简称C-MMC)是最新的换流器改进拓扑,具有直流故障自清除能力,非常适合于架空线输电,可大大扩展柔性直流系统的应用场合。然而,目前涉及至C-MMC的文献很少,因此对其进行深入系统的研究具有工程意义。本文主要围绕C-MMC运行特性及其在大容量架空线输电场合应用的关键问题展开。主要工作如下： (1)研究了C-MMC阻抗频率特性和直流故障穿越机理。建立了稳态运行下MMC连续数学模型和状态空间方程,证明了系统在运行平衡点具有渐进稳定性。推导了C-MMC交直流侧阻抗的等效表达公式,设计了基于测试信号法的阻抗计算方法和详细实现流程,结果表明：交流侧阻抗因非线性调制作用呈现波动特性,直流侧阻抗可等效为单调谐滤波器；运行工况和控制模式对交直流侧阻抗影响均很小,环流抑制会影响直流侧阻抗。建立了C-MMC直流故障下的等值电路,揭示了其直流故障闭锁机理本质。为提高C-MMC稳态运行效益和故障暂态特性,提出了两点改进措施：a)桥臂可由HBSM和CDSM混合而成,b)在子模块内部箝位二极管处串联阻尼电阻。 (2)构建了C-MMC子模块故障多层次冗余保护体系。将子模块划分为三个保护区域,引入了双晶闸管保护方案；设计了子模块内部容错控制流程,以处理不同区域内元件失效问题。在换流器层次,设计了改进型最近电平调制和电容电压平衡策略。该方法动态调节电容电压参考值,通过引入故障因子对电容电压测量值进行修正,以防止故障元件参与投切。提出了动态备用和实时安全裕度的概念,并采取功率调整、挡位动作等措施扩大动态备用元件数量。提出了抑制直流电流波动的电压补偿策略,在故障桥臂电压参考波上附加补偿电压分量,使交流相电流在上下桥臂近似平均分配。 (3)设计了C-MMC-HVDC的正常启动和故障后重启动策略。将正常自励充电过程分为两个基本阶段即静态充电阶段和动态充电阶段。利用电路原理方法化简了静态充电阶段的系统等值电路,推导了最大充电电流的估算公式和限流电阻的选取原则。提出了分组可控有序充电方法,在该方法中模块电容分组依次充电,以保证每个电容获取足够的能量。设计了直流故障后的系统自动重启动控制时序。 (4)研究了C-MMC串并联容量扩展技术和单元投退策略。为实现高电压大容量输电,设计了采用C-MMC组合式换流器的双极柔性直流拓扑,每极由若干换流单元串并联构成,接地极线从上下正负极结构中间引出。将换流单元投退过程划分为并联类和串联类,分别设计了详细的投退控制流程。提出了基于投旁通对的串联类单元投退控制策略,实现了待投退单元的快速旁路。 (5)研究了LCC-C-MMC混合直流稳态运行控制和直流低电压穿越策略。混合直流逆变侧采用组合式C-MMC改进型拓扑,以达到与传统直流相匹配的大容量要求。为解决混合直流系统因整流侧交流电网故障引起的直流低电压穿越问题,提出了带投旁通对的后备定电流和后备定电压两种控制策略。二者分别以逆变侧和整流侧为控制主导站,通过引入三次谐波注入调制和无功功率动态调整策略,扩展了系统的稳态和暂态调节范围；故障严重时,系统借助旁通对快速旁路高位端的换流单元,进入半压运行模式,剩余健全的换流单元维持功率的持续输送。 (6)提出了一种MMC阀损耗计算通用方法,可统一分析不同子模块结构：HBSM、 CDSM和全桥子模块(FBSM)。该方法基本步骤为：第一步,基于系统运行参数和调制控制策略计算各子模块元件的电流电压时域变化波形；第二步,利用厂商提供的特性曲线对半导体器件特性参数进行拟合；第三步,结合器件电流、电压波形和开断次数计算其损耗和结温。本方法能够计及优化电容电压附加控制,便于编程实现,可快速计算各种工况下的换流器功率损耗分布。
[Abstract]:Compared with traditional two-level or three-level converters, modular multilevel converters (MMCs) have become the mainstream topology in the field of flexible HVDC because of their high output voltage waveform quality, good scalability, low loss and low switching frequency. The MMC (C-MMC) with clamped double sub-modules is the latest converter improved topology with DC fault self-cleaning capability. It is very suitable for overhead line transmission and can greatly expand the application of flexible DC systems. In this paper, the operation characteristics of C-MMC and its application in large-capacity overhead line transmission are discussed. The main work is as follows:
(1) The impedance frequency characteristics and DC fault traversing mechanism of C-MMC are studied. The continuous mathematical model and state-space equation of MMC under steady-state operation are established, and the asymptotic stability of the system is proved. The equivalent expression formula of AC-DC impedance of C-MMC is deduced, and the impedance calculation method based on test signal method is designed. The results show that the AC-side impedance fluctuates due to nonlinear modulation, and the DC-side impedance can be equivalent to a single tuned filter. The operating conditions and control modes have little effect on the AC-DC side impedance, and the circulating current suppression will affect the DC-side impedance. In order to improve the steady-state operation benefit and transient characteristics of C-MMC, two improvements are proposed: a) the arm can be made up of a mixture of HBSM and C DSM, and b) the damping resistance in series at the diode clamped inside the sub-module.
(2) A multi-level redundancy protection system for C-MMC sub-module faults is constructed. The sub-module is divided into three protection zones and a double thyristor protection scheme is introduced. The internal fault-tolerant control flow of sub-module is designed to deal with component failure in different zones. The method dynamically adjusts the reference value of capacitor voltage and corrects the measured value of capacitor voltage by introducing fault factors to prevent the faulty components from participating in switching. The fluctuating voltage compensation strategy adds a compensation voltage component to the voltage reference wave of the fault leg, so that the AC phase current is approximately equally distributed between the upper and lower arm.
(3) The normal start-up and restart strategy of C-MMC-HVD C is designed. The normal self-excited charging process is divided into two basic stages: static charging stage and dynamic charging stage. Principle. A grouped controllable ordered charging method is proposed, in which the module capacitors are grouped and charged sequentially to ensure that each capacitor gets enough energy.
(4) The C-MMC series-parallel capacity expansion technology and unit switching-back strategy are studied. To realize high voltage and large capacity transmission, a bipolar flexible DC topology using C-MMC combined converter is designed. Each pole is composed of several commutation units in series and parallel, and the grounding poles are drawn from the upper and lower positive poles. Detailed control flow is designed for the cascade and cascade classes respectively, and a cascade control strategy based on the cast-by-pass pair is proposed to realize the fast bypass of the cascade.
(5) The LCC-C-MMC hybrid DC steady-state operation control and DC low-voltage traversal strategy are studied. A combined C-MMC topology is adopted in the hybrid DC inverter side to meet the large capacity requirement matched with the traditional DC. To solve the DC low-voltage traversal problem caused by rectifier-side AC power grid fault in the hybrid DC system, the band is proposed. The backup constant current control strategy and the backup constant voltage control strategy with bypass pair are adopted. The inverter side and the rectifier side are used as the control dominant stations respectively. By introducing the third harmonic injection modulation and reactive power dynamic adjustment strategy, the steady-state and transient regulation range of the system is extended. The converter unit enters the half pressure operation mode, and the remaining sound converter unit maintains the continuous transmission of power.
(6) A general method for calculating the loss of MMC valves is proposed, which can analyze different sub-module structures: HBSM, CDSM and FBSM. The basic steps of this method are as follows: First, calculating the time-domain variation waveforms of current and voltage of each sub-module component based on system operating parameters and modulation control strategy; second, using the special features provided by manufacturers. In the third step, the loss and junction temperature are calculated by combining the device current, voltage waveform and the number of interruptions.
【学位授予单位】：浙江大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TM721.1

【参考文献】