基于BP神经网络的MMC电容电压控制研究

模块化多电平换流器(Modular Multilevel Converter,MMC)作为一种新型的电压源换流器,被广泛应用于直流输电领域.电容的电压平衡问题在MMC中非常重要,随着MMC子模块个数的增多,电容电压的平衡变得越来越困难.本文提出了一种基于BP-PID的双闭环控制策略,通过控制调制波来控制子模块的工作状态...     

考虑子模块均压约束的混合型模块化多电平换流器损耗分析

随着柔性直流输电工程电压等级及输送容量不断升高,模块化多电平换流器的损耗问题引起诸多关注.目前,常用的损耗分析方法都是针对全桥和半桥型模块化多电平换流器进行分析计算.基于混合型模块化多电平换流器的工作特性,文中提出了 一种考虑子模块均压约束的混合型MMC损耗分析方法.该方法根据混合型MMC桥臂电压和电流在一个控制周期内...     

模块化多电平换流器电容电压的滑模观测器研究

模块化多电平换流器（modular multilevel converter,MMC）子模块电容电压的稳压控制是MMC的一个关键难题,在实际应用中,需要大量的电压传感器以获取各子模块电容电压。对此提出了一种电容电压滑模观测器（sliding mode observer,SMO）,根据MMC电路中的桥臂电流和开关信号观测出子模块电容电压,并结合电压闭环控制,实现电容电压的稳定,采用双曲正切函数tanh作该滑模观测器的切换函数,可削弱系统抖振。仿真和实验结果表明该滑模观测器可准确观测出电容电压。     

基于Z源网络的模块化多电平逆变器拓扑设计

传统多电平逆变器存在着不能升压的缺点而使其的应用受到限制。根据Z源网络可实现升降压的功能,提出将Z源网络应用到模块化多电平逆变器(MMC)中。通过对Z源网络和MMC的原理及运行状态的分析,设计出含Z源网络的MMC拓扑结构、升压策略、包含MMC子模块电容电压平衡控制的调制策略,通过控制Z源网络的直通占空比就可以灵活控制输出电压。仿真结果验证了新型拓扑结构和控制策略的正确性和优越性。     

一种高精度可调电平接口电路设计

为了满足各种仪器或电路所需的接口电平,设计了一种可生成频率范围在25 MHz～3.2 GHz、电平差模范围为0～1.9 V、共模范围为-4 V～4 V可调电平的接口电路。外部驱动源驱动差模电路产生一个频率为25 MHz～3.2 GHz的差模信号,由数模转换器产生的共模信号通过电阻与差模信号耦合输出电平信号,通过两个信号参考端隔离的办法实现电平的共模电压和差模电压解耦调节,差模和共模信号电平值通过电平控制模块来设定。选择接口输出为标准LVDS、RS485和PECL电平进行实验测试,测试结果表明,该电平接口电路输出的电平信号稳定,精确度高,电平误差小于5 mV。     

高电平MMC子模块电容电压控制策略研究

为了提高高电平MMC均压控制的排序速率,提出一种适用于高电平MMC子模块的电容电压分块均压控制策略。首先根据子模块电容电压最值划分分块区域,由排序的复杂度和均压控制效果确定分块区域的数目,然后根据桥臂电流大小和需投子模块个数,从分块区域内选择相应的子模块投入,最后引入分块区域的重新划分判据以减少计算量。在PSCAD/EMTDC环境下搭建21电平MMC系统仿真模型。仿真结果表明,该控制策略提高了分块区域的利用率,降低了IGBT的开关损耗。     

电网电压不平衡时模块化多电平换流器控制策略研究

在电网电压不平衡状态下,负序电流的存在引起模块化多电平换流器(MMC)内部环流包括负序、零序和正序二倍频分量,导致传输有功功率、无功功率及直流侧电压二倍频波动,严重时会影响装置内外特性。为提高电能传输质量及确保换流器安全运行,研究MMC内外不平衡控制策略尤其关键。文中从传统方法、改进方法和创新方法三个方面分别介绍关于电网电压不平衡时MMC外特性和内特性不平衡控制策略的研究进展,并比较其优缺点;总结MMC不平衡控制策略及探讨其未来的研究方向。     

基于模块化多电平光伏并网动态建模与仿真研究

鉴于传统三电平逆变电路存在谐波含量高的缺点,在分析模块化多电平换流器(MMC)工作原理的基础上,提出使用模块化多电平电路实现光伏并网逆变功能,采用微分-跟踪器法实现光伏阵列最大功率点跟踪(MPPT)及逆变器PQ解耦控制,实现了光伏系统以单位功率因数并网.在PSCAD中建立光伏并网系统动态仿真模型,仿真结果表明,所建模型具有开关损耗低、谐波量小的优点,验证了所提方法的正确性和可行性.     

连铸坯视觉定重线结构光平面调整方法研究

鉴于传统三电平逆变电路存在谐波含量高的缺点, 在分析模块化多电平换流器(MMC)工作原理的基础上, 提出使用模块化多电平电路实现光伏并网逆变功能, 采用微分-跟踪器法实现光伏阵列最大功率点跟踪(MPPT)及逆变器PQ解耦控制, 实现了光伏系统以单位功率因数并网。在PSCAD中建立光伏并网系统动态仿真模型, 仿真结果表明, 所建模型具有开关损耗低、谐波量小的优点, 验证了所提方法的正确性和可行性。     

双星型模块组合型多电平变换器的PWM控制策略

双星型模块组合型多电平变换器(MMC)具有良好的可扩展性、模块化、故障无损特性,是一种能很好满足未来高压或中压功率变换需求的拓扑。然而,很少文章对其PWM调制策略及参考电压指令生成方法进行详细的分析和讨论。基于MMC的电路结构特点和基本工作原理,本文提出一种优化的载波移相SPWM控制策略和电压控制策略,无需额外的硬件电路,即可实现对MMC功率单元直流电容电压良好的均压控制效果。通过一台5电平三相MMC的仿真结果,详细分析了其动态特性、静态特性,验证了本文控制策略的有效性以及MMC的可行性。     

单模光纤模场分布特性

介绍光纤的概念及其种类,光纤损耗和传输优点,并应用matlab软件数值模拟了单模光纤模场的分布特性,结论为光纤器件的制作提供参考.     

Polarization mode dispersion in a single mode fiber

The Jones matrix method is used to measure the polarization mode dispersion (PMD) of a large variety of single mode fibers in the 1500 nm range. The dependence of PMD on wavelength, time, and temperature are studied in two different regimes: adiabatic and isothermal. In the adiabatic regime, time dependent stresses are introduced in the fiber by subjecting it to large and rapid changes in temperature. In this regime it is shown that the probability density function of PMD, obtained as a function of temperature and wavelength, fits very closely the theoretically predicted Maxwellian function. In the isothermal regime, the temperature of the fiber is held constant and the stresses are allowed to relax to their long term steady state conditions. In this regime the PMD exhibits a strong dependence on wavelength but otherwise is a bounded function which is nearly stationary with time. Test and analysis of the deterministic PMD in a specially constructed polarization maintaining fiber are used to study the dependence of PMD on temperature and wavelength. Finally, the system implications of this PMD study are described briefly     

Polarisation mode dispersion measurements in long single mode fibres

A frequency domain determination of polarisation mode dispersion in single mode fibres is described, which allows one to obtain directly the group delay difference for long fibres. The results of such measurements are shown on a birefringent fibre with elliptical cross-section which maintains an extinction ratio between linearly polarised eigenstates greater than 10 dB over 1 km.     

Polarisation mode dispersion measurements in long single mode fibres

A new method to measure polarisation mode dispersion is described, in which the accuracy is more than 0.1 ps. 0.17 ps polarisation mode dispersion was obtained for about 1 km length of fibre. The result is compared with the retardation obtained for a short length of the fibre.     

Correlation between mode fluctuations in a multilongitudinal mode laser

A theoretical analysis of the correlation between mode fluctuations in a multilongitudinal mode laser is presented. The rate equations driven by Langevin noise sources are used to derive the correlation coefficients. By taking nonlinear terms into account, this analysis also applies to lasers with few modes and permits the derivation of the correlation coefficients from the the amplitude of these modes. They are found to be different from one pair of modes to another and to have not only negative, but also positive values. The fluctuations of a given mode are found to be concentrated in the low-frequency range, and the associated time constants are of great importance in the estimation of the partition noise for a signal propagating through a dispersive fiber.     

Networking a la mode [asynchronous transfer mode]

The asynchronous transfer mode (ATM) was developed at the behest of the telecommunications industry, but is being deployed principally within private data networks. In this paper, the author explains this apparent paradox     

Asynchronous transfer mode or synchronous transfer mode or both

Some of the problem areas attendant upon the asynchronous transfer mode (ATM) concept are addressed. Concern is chiefly with the use of small, fixed-length packets and the issue of congestion control. It is argued that broadband ISDN services should not be tied down to ATM as presently standardized, because ATM has several problems when embedded in a wide area network with diverse services     

Pure mode versus mixed mode interfacing between analog circuits

The concepts of transmission error and mismatch factor ^–1 are introduced to evaluate the effect of impedance mismatch on the accuracy of broadband signal transmission between two feedback amplifiers. It is shown that, in comparison with the error introduced by the feedback amplifiers, becomes negligible for pure V.M. (voltage mode) or C.M. (current mode) signal transmission; in the mixed mode, in which a V.M. output circuit feeds a C.M. input circuit, or vice versa, may become significant. Computer simulations show that the pure mode also yields reduced T.H.D., improved bandwidth and improved transient response. It is also shown that a particular combination between the kind of feedback and the active circuit to which it is applied, termed enhancing combination, further increases the accuracy of signal transmission.1. The notation (OL) relates to the open-loop value of the parameter involved.2. It can be shown that rule 2 does not apply to a low-noise preamplifier.     

Analyses of mode partition and mode hopping in semiconductor lasers

Mode power fluctuations in semiconductors laser due to mode partition and mode hopping are discussed. The power dropout probability P_e in the mode partition was measured for a wide range, 1.6×10^-6⩽P_e⩽1, which decreased by increasing I/I_th (I and I_th are the DC injection current and its threshold value respectively). The duration time t_d of the power dropout expressed at t_d=3.7×10 ⁴⁸ exp[-118 (I/I_th)] for 1.065⩽I/I_th⩽1.104. Power fluctuations exhibited specific characteristics around the threshold, which were similar to the critical slowing down in the phase transition phenomenon. An increase in the variance of the power fluctuations was observed when the laser oscillating condition was converted from mode partition to mode hopping. The unified stochastic model based on the Fokker-Planck approach described well both mode partition and mode hopping     

双曲余弦高斯光束的模相关和模结构分析

基于二阶矩方法和正交模系展开法 ,推导出双曲余弦高斯光束的光束传输M2 因子和模相干系数的解析表达式 ,从而可对双曲余弦高斯光束的模相关和模结构进行分析。提出了一种在实验室中产生双曲余弦高斯光束的简单方法。