【作者】 周成；

【导师】 贺仁睦；

【作者基本信息】 华北电力大学（北京） ， 电力系统及其自动化， 2010， 博士

【摘要】 随着我国经济的快速发展,电力系统也日益庞大与复杂。一方面电网互联使得电网规模越来越大,另一方面科技的进步使得新型设备不断的在电网中投运。保证电力系统安全运行就需要工作人员准确掌握系统的动态特性,这就必须要借助于电力系统动态仿真。因此保证电力系统动态仿真的可信度对于保证系统运行的安全性、经济性起着举足轻重的作用。提高先验仿真结果的可信度需要提高参与仿真计算的模型及参数的有效性。本文按后验仿真验证、模型及参数有效性评估、先验仿真结果可信度评估这个流程对如何评估模型有效性、如何评估仿真结果可信度进行了研究。后验仿真准确度评估是所有工作的基础,目前的评估方法主要是从数值方面来评估曲线相似性,忽略了曲线的动态特性。本文从曲线的数值与曲线的形状两方面来比较时间序列的相似性,进而合理的评估后验仿真准确度。本文分别利用规范系数法与界标分界法计算曲线形状相似度,评估了系统及单个元件的仿真准确度。评估系统仿真准确度时,综合考虑了关键节点的电压幅值、相角、有功、无功,避免评估结果的片面性；评估单个元件的模型的仿真准确度时,利用实测数据对模型进行混合仿真,评估时还考虑了不同时段仿真误差对评估结果的影响。由于系统中元件数目庞大,对所有元件模型进行校核比较费时。本文通过详细的统计分析,从模型的建模理论完善程度和参数来源的可信程度两方面对模型有效性进行等级划分,进而计算出模型及参数的有效性指标值。利用每次系统扰动所记录的实测数据对单个模型或者一个子网进行混合仿真,根据仿真结果的准确度对模型的有效性指标进行动态修正。工作人员根据模型及参数的有效性指标值能提高误差溯源的速度。所有工作的目的就是要合理评估与提高先验仿真的可信度。本文利用模型及参数的有效性指标值,对先验仿真结果进行可信度评估。由于扰动时不是所有元件都参与系统的动态行为,而且不同元件对系统的动态特性影响也不同,所以在评估过程中,同时考虑了模型重要度与模型参与度两种因素。这使得可信度评估结果更合理、科学。更多还原

【Abstract】 The grid becomes more and more large and complex with the national economy develops quickly.One side the scope of the grid is larger than before as the result of the grids are interconneted, the other side new types of equipments are put into operation continually with technological progress. Only the staff of the grid company must learn about the dynamic character of grid accuratly that the grid can run safely, this must resort to the dynamic simulation. So to guarantee the reliability degree of the dynamic simulation result is critical to the safety and economy of the grid operation..The best way to improve the reliability of simulation result is to improve the validity of the model and its parameters used in dynamic simulation. In the paper the problems of how to set up the system of evaluating the validity of model and its paramters and how to evaluate the reliability of the prior simulation result are analyzed following the order that post simulation validation, validity evaluation of model and its parameters, reliability evaluation of prior simulation. The main creative achievements are shown as following:Accuracy assessment of posterior simulation is the base of all work. The current assessment methods are mainly to assess the curve from the numerical similarity aspects and ignore the dynamic aspects. From the two sides of the curve data values and shape to compare the similarity of the time series, so evalating the result of post dynamic simulation is more reasonable.Based on the criterion and landmark dividing method the shape similarity of the measured data and the simulation results is compared, also the difference of each time intervals is also considered.When evaluating accuracy of the system simulation, the physical data of critical nodes, voltage, angle, active power, reactive power are compared generally.Due to the large number of system components, it is time-consuming to validate the entire components model. Based on the detailed investigation, the model is analysed from the perfect degree of modeling theory and the reliability of the way by which the parameters can be get. So the index value of validity of model and its parameters can be computed. By use of the measured data recorded when the power system is disturbed, the hybrid simulation can be run according to a model or a subnet, then the accuracy degree of post simualtion can be evaluated and the index value of validity of model and its paramters can be modified. According to the index of validity, the engineers can improve the speed of error traceability.All aim is to make a reasonable assessment of the credibility of a priori simulation and enhance credibility. By use of the index of validity of model and its paramters, the reliability degree of prior simulation result can be evaluated. During the disturbance, not all the equipments participate in the dynamic behavior of the system and the impact of each device on the dynamic behavior is different, so two factors, degree of model participation and degree of model importance, are considered during the evaluation. This makes the evaluation result more reasonable.:更多还原