【作者】 陈众；

【导师】 徐国禹； 王官洁；

【作者基本信息】 重庆大学 ， 电气工程， 2003， 博士

【摘要】 统一潮流控制器（UPFC）是柔性交流输电系统（FACTS）的重要组成部分。它具有能够同时控制线路上潮流及接入点母线电压的特点，是FACTS家族中最具有吸引力的元件之一。UPFC概念出现十余年来，一直得到了电力系统运行与控制领域专家的广泛关注，在结构设计、控制策略和设备制造等方面曾进行了较为深入的研究。在关于UPFC控制方式的研究中，常用控制策略（如PI控制、线性最优控制）将电力系统的量测信息（如母线电压和线路上的有功功率、无功功率等）直接作为控制系统的反馈，从而将电力系统的非线性和强耦合特性引入到UPFC控制系统。而智能预估控制方法虽然意识到了这一问题，却未能提出简洁的解决方案。本论文将智能控制领域的重要分支——仿人智能控制（HSIC）引入到UPFC的控制中。在对比仿人智能控制理论的分层递阶结构与电力系统自动化的分层次结构的基础上，指出二者具有极其相似之处，存在紧密的类似于理论与实践之间相互的促进关系，进而提出了基于HSIC的UPFC智能控制系统结构。它将整个系统划分为两个相对独立的子系统：非线性系统和线性随动系统，用以解决通常UPFC控制方法将非线性和强耦合特点引入控制器中，造成的解耦困难、控制范围缩小等问题。根据HSIC理论多模态控制的特点以及针对控制对象进行控制策略设计的基本方法，论文在分析UPFC直流侧电容电压动态过程以及UPFC常规控制策略对直流侧电容电压控制的共同特点的基础上，确定了UPFC直流侧电容电压强控制的概念，相应地提出了UPFC直流侧电容电压弱控制策略。该策略利用了直流侧电容能够在干扰过程中，吸收或释放系统中部分的多余或不足的能量，从而有利于稳定系统的主要控制目标，如线路潮流和母线电压。论文给出了直流侧电压弱控制策略一个具体的实现方案，通过对系统发生三相对称短路以及调度计划发生变化的情况，运用UPFC直流侧电容电压弱控制策略对系统进行仿真，并与线性最优控制方案进行了对比分析。仿人智能控制的多模态结构为增强低层次控制器的智能性提供了较大的发展空间，使控制器从整体上更接近人的智能行为。本文在低层次智能的实现上，运用仿人智能控制的基本研究方法和手段，提出了基于HSIC思想的NPID控制器的设计原理，给出了比例-积分-微分三个分量改进的基本方法；给出了依据仿人智能控制原理设计NPID的两个实例，通过仿真实例验证了其有效性，并运用于UPFC智能控制系统低层的运行控制级。基于UPFC直流侧电容电压弱控制策略集成在UPFC智能控制系统中，设计<WP=10>了直流侧电容电压弱控制策略与线性最优控制策略相结合的弱-强二阶段控制方式，给出了基于HSIC的UPFC智能控制系统的各级控制方案的数学描述。弱-强二阶段控制方式在大干扰时初始阶段，如果电容不存在过电压现象，则主要完成对系统主要控制目标，如线路潮流和母线电压的控制，在系统稳定后或小干扰情况下才考虑UPFC直流侧电容电压的控制，从而加快了系统稳定的响应速度。此外，论文对于来自发电机控制单元的远端信息可能存在的干扰，如延时和白噪声干扰，讨论了基于HSIC的UPFC智能控制系统在干扰下的工作性能。更多还原

【Abstract】 As an important part of Flexible AC Transmission System (FACTS) family, Unified Power Flow Controller (UPFC) is the most attractive device because of its strong ability to regulate bus voltage and optimize power flow on the transmission line synchronously. Since it was proposed 10 years ago, it has been widely concerned by the experts in power system operation and control fields and extensive researches have been gone along in its structure design, control strategy and manufacture.The control strategies in common use (such as PI control, linear optimal control and so on) feed back the measure information, such as the bus voltage, active power and reactive power, into the control system directly, which brings the nonlinear and strong couple characteristics of power system into the UPFC control system. Although Intelligent Predictive-Evaluating Control method had taken notice of this problem, but did not give a simple and convenient way to solve it.This paper introduces the important offshoot of intelligent control field --- Human-Simulated Intelligent Control (HSIC) theory to the control system of UPFC. Based on the comparison of the hierarchical architecture of HSIC and the hierarchical architecture of power system automation, it analyses the analogous relation like theory and practice between them. Further this paper proposes the intelligent control system structure based on HSIC, which divides the whole system into two comparatively independent parts: nonlinear subsystem and servo subsystem to solve the nonlinear and strong couple problem in normal control strategies used in UPFC control system.According to the multiple modes control feature of HSIC and its basic and control object oriented control strategy design method and based on analyzing the charge and discharge dynamic process and the control method on dc-link capacitor voltage of common control strategies used in UPFC control system, this paper confirms the dc-link capacitor voltage strong control strategy and proposes the soft control concept on dc-link capacitor voltage control relatively. The soft control strategy utilizes the dc-link capacitor to absorb or release the transient energy during the fault or disturb period, which will benefit the main control object such as power flow and bus voltage. This paper provides a detail realization of soft control strategy and compares it with the linear optimal control strategy by simulating the system response under the condition of three phase symmetry short and dispatching task changed. The multiple mode control structure feature of HISC provides the developing<WP=12>scope for the intelligence in different levels, which make the controller approach the human-being behaves as a whole. In low level intelligence, using the basic research way, this paper proposes the nonlinear PID (NPID) controller design method based on HSIC and the improving way in P-I-D control parts. It also provides two design examples of NPID based on HSIC, proves its effectiveness by simulating and utilizes it to the low level of UPFC intelligent control system.Based on the soft control strategy on dc-link capacitor voltage and the basic theory of HSIC, this paper integrated the soft control strategy and linear optimal control strategy into the UPFC intelligent control system, proposed the soft-strong control strategy and provides the detail mathematic describe of each level scheme in the control system. During the beginning of strong disturb, if the capacitor is not over voltage, the soft-strong control strategy aims at the main control object such as the power flow and bus voltage, and consider the dc-link capacitor voltage control just under the condition that the system is nearly stable or is disturbed lightly, which accelerates the stable process.Additionally, this paper discusses the performance of UPFC intelligent control system based on HSIC under the disturbance (such as delay and white noise) on the information from the generator control unit.更多还原