基于非线性规划灵敏度技术的模型预测长期电压稳定控制  

Model Predictive Control for Long-Term Voltage Stability Based on Nonlinear Programming Sensitivity Technique

作　　者：王爽[1] 谢敏[2] 刘明波[2] 王晓东[1] 初艳伟[1] 

机构地区：[1]中国电力科学研究院,北京100192 [2]华南理工大学电力学院,广州510640

出　　处：《华东电力》2013年第12期2472-2478,共7页East China Electric Power

基　　金：国家自然科学基金项目(50907023;50777021)~~

摘　　要：针对长期电压稳定问题,提出了非线性模型预测长期电压控制滚动优化模型和求解方法。以电力系统详细模型为基础,等式约束条件包含微分—代数方程。为提高求解该优化模型的计算效率、降低在线反馈延时,采用非线性规划灵敏度技术求解滚动优化问题。基本思想是:采用排列法将滚动优化问题转化为非线性规划问题,根据下一周期的状态预测值求解此非线性规划问题,并同时给出Karsh-Kuhn-Tucker方程对模型参数的灵敏度。在进入下一控制周期后,再根据状态变量的真实值,采用灵敏度信息修正控制变量值。该方法可以大幅提高计算效率,为紧急电压控制应用于大系统创造了条件。IEEE 17机162节点系统的算例说明了该方法的优越性。In view of long-term voltage stability, this paper propose a receding optimization model for nonlinear model predictive voltage control and its solution method. The proposed model is based on the detailed model of power grid and equality constraints include continuous and discrete time differential-algebraic equations. To improve computational efficiency of the proposed model and decrease online feedback delays, it is proposed to solve the receding optimization problem by nonlinear programming sensitivity algorithm. The basic idea of this approach is as follows： firstly, turn the receding optimization into a nonlinear programming problem by Radau collocation method; then solve this nonlinear programming problem in accordance with the predicting state variable values of the next period, and mean- while calculate the sensitivity of Karsh-Kuhn-Tucker equation of this nonlinear programming problem to model parameters ; in the next control period, modify the values of control variables by sensitivity information according to true values of state variables. The proposed method can improve computational efficiency considerably and create the condition of applying emergency voltage control to large-scale power grid. The superiority of the proposed method is verified by IEEE 17-machine 152-bus system.

关 键 词：长期电压稳定 全过程动态模型 非线性模型预测控制 滚动优化 非线性规划灵敏度 

分 类 号：TM712[电气工程—电力系统及自动化]