节点文献
基于响应的电力系统暂态稳定性实时判别与控制技术的研究
Studies on Real-Time Transient Stability Determination of Power System and Control Technology Based on Responsed Data
【作者】 吴为;
【导师】 汤涌;
【作者基本信息】 中国电力科学研究院 , 电力系统及其自动化, 2014, 博士
【摘要】 基于实时广域测量信息进行了电力系统暂态稳定分析的理论探讨、数学模型的建立、动态特征信息提取、受扰轨迹预测、实时暂态稳定判据、主导失稳模式识别及切机控制措施量化等方面的研究,研究了基于广域实测信息的电力系统暂态稳定响应控制策略。广域测量系统可以获得全网的实时动态响应数据,由于测量数据是海量的,需要从海量测量数据中提取部分动态特征信息进行研究,从而避免对海量数据的处理,提高暂态稳定分析的速度。动态特征信息必须要能全面反映系统受扰后的暂态稳定过程,考虑到多机系统暂态稳定问题的复杂性,单一依靠某种特征信息可能很难准确识别系统的稳定性。因此,基于广域测量信息的暂态稳定研究,需要综合利用多种动态特征信息。电力系统受扰轨迹变化趋势十分复杂,具有很强的非线性特性,要求在暂态稳定受扰轨迹的预测模型必须能实时跟踪曲线的变化趋势,及时调整模型参数以提高预测性能。本文基于预测模型参数的特性,提出了一种预测模型参数自适应调整技术,主要目的是跟踪受扰轨迹的非线性变化,利用最新的测量值,及时调整模型参数从而实现总体最优预测性能。模型参数自适应技术主要分为内外两部分循环计算,内循环计算根据预测模型得到预测结果,外循环计算基于最新测量值,根据模型参数自适应调整原则,及时修正模型参数,提高模型的适应能力。实时暂态稳定性判据是实现暂态稳定实时决策的关键,对多机系统而言,系统发生扰动时输电断面联络线的特征信息能有效反映两侧系统间的摇摆情况。本文基于联络线特征信息提出了一种基于功率-相角-频率的快速判据,仅使用有限的、关键的网络特征信息进行设计,因此具有快速性和实用性。在进行失稳识别时,首先判断系统是否进入不稳定区域,然后判断系统是否有趋于失稳的倾向,最后判断系统是否会越过不稳定平衡点。基于实际系统进行了仿真验证,结果表明运用此判据可准确地判别暂态功角失稳。暂态功角失稳与暂态电压失稳是故障后系统在暂态过程中的两种主要表现形式,暂态功角失稳和暂态电压失稳的主导性识别,是进行暂态稳定控制的前提。本文研究表明,暂态过程中输电断面联络线有功功率的变化量包含两部分分量,一部分与联络线母线电压相角相关,另一部分与联络线母线电压幅值相关。这两部分分量与失稳模式之间存在如下关系:当电压失稳为主导失稳模式时,输电断面功率改变量主要由与母线电压幅值相关的分量引起,当功角失稳为主导失稳模式时,输电断面功率改变量主要由与母线电压相角差相关的分量引起。基于这种关系,本文提出了一种功率全微分的主导失稳模式识别判据,为后续紧急控制提供决策依据。进行暂态稳定紧急控制时,首先需要确定合理的切机量。基于暂态能量函数的切机控制措施量化方法,要求能量函数能精确地反映出导致系统失稳的暂态能量。修正的暂态能量函数与同步坐标下和惯量中心坐标下的暂态能量函数相比,最能反映导致系统失稳时的暂态能量。基于修正的暂态能量函数,本文定义了系统的加速能量函数。基于临界切机量时系统加速能量所满足的特性方程,提出了一种基于系统加速能量的临界切机量求取方法。与基于等面积法则的传统方法不同,该方法不需要计算系统的不稳定平衡点,积分路径拟合及积分计算过程简单。
【Abstract】 This dissertation focuses on the rotor angle stability analysis based on wide-area measurement information, including mechanisms analysis, mathematics modeling, extraction of dynamic characteristic data, prediction of disturbed trajectories, real-time transient instability criterion, recognition of principal mode and quantitative research of generation capacity tripped. Then a response data driven transient stability control strategy based on real-time measurement data is proposed.Response data of network can be collected by wide-area measurement system and it is need to extract the dynamic characteristic data for transient stability analysis in order to avoid dealing with the huge number of PMUs data and improve the speed of stability analysis. Dynamic characteristic data is required to reflect the process of transient stability after disturbance. Considering the complexity of transient stability problem of multi-machine system, it is difficult to identify the stability of power system correctly, so multiple characteristic data is used for transient stability analysis based on wide-area measurement information.The tendency of disturbed trajectories variation is complex and the curves have strong nonlinear characteristics. In the prediction of disturbed trajectories, the parameters of prediction model must dynamically adjust to cope with the variation of nonlinearity of curves in real-time to enhance the prediction ability. Based on detailed analysis on the relation between prediction performance and model parameters, a self-adaption technology for model parameters is proposed in this dissertation. Using the latest measurement data to calculate the parameters and cope with the variation of nonlinearity for the purpose of realizing overall optimal performance prediction. The proposed technology includes inner and outer loop calculations. Inner loop focuses on the calculation of prediction results and the outer loop focuses on the calculation of parameters with the latest measurement data according to principles of adaptive adjustment of model parameters.Real-time transient stability criterion is the key of real-time transient stability analysis. For multi-machine power system, dynamic characteristic information of tie lines in transmission section can reflect the oscillation status between two systems when the oscillation occurs. A power-phase angle-frequency criterion based on the characteristic information of tie lines is proposed. Few and vital network information is used to design this criterion, so it has fast calculation speed and practicality to be used. There are three steps to identify the unstable status with this criterion. Firstly, identify if the system operating point is in the unstable area. Secondly, identify if the system has the tendency to lose stability. Finally, confirm if the system operating point is beyond the unstable equilibrium point. Simulation results based on actual power system show that the criterion can identify transient angle stability correctly.Rotor angle instability and transient voltage instability are two forms of system instability during the post disturbance. Recognition between them is the pre-preparation for power system emergency control. Research shows that the variation of active power of tie lines in transmission section includes two components. The first component is associated with the bus voltage magnitude of tie lines; the second component is associated with the phase angle of bus voltage for tie lines. The relationship between different unstable modes and the components is shown as follows:When the transient voltage instability is the principle mode, the variation of active power in tie lines is mainly determined by the first component. When the rotor angle instability is the principle mode, the variation of active power in tie lines is mainly determined by the second component. A criterion for recognition of principal mode based on this relationship is presented for the subsequent emergency control.Estimating the generation capacity to be tripped suitably is a pre-preparation for power system transient stability emergency control. The methods of estimating generation capacity tripped based on transient energy function requires that the function can accurately reflect the energy which causes the system instability. Based on the corrected energy function, accelerated energy function of power system is defined. With the detailed analysis of accelerated energy of power system, a method to estimate generation critical capacity tripped is presented. Different from the equal-area criterion, this method don’t need to calculate the unstable equilibrium points, integral path determined and calculation process are simple.
【Key words】 Wide-area measurement system; rotor angle instability; transient voltage instability; predictionof disturbed trajectories; real-time transient instability criterion; recognition of principal mode; principal system variable; transient energy function; generator tripping;