【作者】 赵红嘎；

【导师】 薛禹胜；

【作者基本信息】 山东大学 ， 电力系统及其自动化， 2005， 博士

【摘要】 电力系统状态估计为电力系统在线分析和控制功能提供电网实时工况,状态估计的性能直接影响分析的准确性和控制的效果。电网互联和电力市场化不仅使电力系统运行的复杂性和不可预知性增大,而且对分析和决策的准确性以及控制效果提出了更高要求,这就要求状态估计能够提供及时、准确、全面的电网实时工况。实时状态估计通过数据处理技术,滤除实时量测数据误差,获得最接近于系统真实状态的最佳估计值,量测数量、质量以及配置情况都会影响状态估计的性能。传统状态估计主要基于监控和数据采集(SCADA-supervisory control and dataacquisition)系统提供的遥测量。随着相量测量装置(PMU-phasot measurement unit)在电网中的推广应用,PMU量测已成为电力系统重要的数据源之一。PMU量测具有精度高、全网严格同步、更新周期短等优点,并实现了节点状态直接可测。由于成本原因,当前PMU配置还不能保证电网可观。如何在状态估计中有效利用PMU量测是当前必须面对和解决的问题。直流输电技术在远距离输电和电网互联中具有独特优势,在电力工业中得到了越来越广泛地应用,尤其将在我国西电东送和全国联网中起主导作用,交直流互联系统将会越来越多,而当前的状态估计算法还无法有效的考虑直流模型,研究快速可靠的交直流系统状态估计算法具有重要的理论和工程应用价值。本文主要针对PMU量测在状态估计中的应用以及交直流系统状态估计问题进行研究。PMU可以直接测量节点状态相量,如果PMU配置能使系统可观,则不需要进行状态估计。但是当前PMU配置不可能使系统可观,并且在相当长的时间内还难以实现。PMU的可观配置。本文对各种利用PMU量测的状态估计模型进行分析,提出当前可行的研究思路:集成PMU量测和SCADA量测,共同应用于非线性状态估计,利用PMU量测特性来改善状态估计性能,从而明确了深入研究的方向,促进了PMU量测的工程应用。在非线性状态估计中,利用PMU电压相量量测较方便,但直接利用PMU电流相量量测存在困难。本文提出两种利用PMU电流相量量测的新方法:把支路电流相量量测等效变换为支路功率量测;把支路电流相量量测等效变换为相关节点电压相量量测,从而使所有PMU量测都能在状态估计中得到有效利用。PMU可以直接测量所在节点状态量,并且PMU相关节点的状态量也可以直接推算出来,因此配置PMU后,会形成状态可观测的局部区域。本文充分利用山东大学博士学位论文该PMU量测特性,提出在非线性状态估计中利用PMU量测的另一种模型:在估计运算中直接把PMU节点状态量测值及PMU相关节点状态推算值作为估计值,并对该模型进行深入研究和仿真,指出该模型可以降低估计系统规模,提高估计方程的数值稳定性和迭代收敛速度,但由于PMU量测也存在一误差,该模型不一定会提高估计精度。 本文首次提出非线性估计和线性估计结合的二次估计模型:首先把PMU节点的状态量测值及PMU相关节点的状态推算值作为估计值进行非线性估计,然后再利用非线性估计结果和PMU量测进行线性估计。该模型在利用PMU量测提高状态估计收敛速度的同时,可以进一步利用PMU量测来提高估计精度;在当前PMU不可观配置的情况下,这种模型可以作为从非线性估计到线性估计的过渡。 传统分布式状态估计由于边界量测冗余度低,使边界节点估计精度低,边界量测坏数据不易辨识,并且在子系统参考节点协调时给全网估计值引入较大误差。本文首次把PMU量测应用于分布式估计,提出PMU量测新的应用领域:基于PMU量测进行子系统参考节点协调,并通过在子系统边界节点配置PMU,把PMU节点的状态量测值及PMU相关节点的状态推算值作为估计值,使子系统相互之间完全解藕,从而提高边界量测坏数据辨识能力。研究和仿真表明:利用PMU量测的分布式估计模型的估计精度和边界量测坏数据辨识能力都优于传统分布式估计模型。 PMU测点配置情况会影响状态估计的性能。本文基于PMU量测在状态估计中的两种利用模型:把PMU量测作为常规量测;把PMU节点的状态量测值及PMU相关节点的状态推算值作为估计值,分别从提高估计精度和估计方程数值稳定性出发,研究PMU测点优化配置问题,提出优化配置模型,引入禁忌算法实现了PMU测点优化配置。 SCADA量测传输时延的差别将影响状态估计精度;PMU量测和SCADA量测在传输时延、更新周期上差别很大,在状态估计中同时利用两种量测量时,必须考虑两种量测量的协调问题。本文基于对实际SCADA量测采样和传输系统的深入分析,创造性的提出量测传输时延均匀分布模型,基于该模型,分析了量测时延偏差对状态估计精度的影响,提出基于量测时延期望值对不同时延量测进行匹配的方法,使状态估计量测时延问题及PMU量测和SCADA量测协调问题获得了完美解决,并且该方法易于工程实现。 越来越多的交直流联合系统的出现,迫切需要研究适合于工程应用的交直流系统状态估计算法。一个工程实用化的状态估计算法应该具有较少的计算量和内存占用量,并且收敛性好。本文提出一种新的交直流联合系统快速分解状态估一计算法,该算法实现了交流有功、无功和直流方程的解祸,并通过对直流方程用保更多还原

【Abstract】 The power system state estimate (SE) gives system real-time status for power system’s analysis and control. Both the analysis’ veracity and the control’s effect are influenced by the performance of the SE. The interconnection of power grids and the running of power market not only make the power system behavior become more complex and more difficult to anticipate, but also require that the analysis and decision be more accurate and the control be more effective. Then the SE should give exact state data in time and comprehensively. The SE can filter the measurement error by data processing, and gain the best estimation of the power system state. The quantity, quality and placement scheme of the measurements influence the SE performance.The traditional SE utilizes only the measurements of the SCADA-supervisory control and data acquisition system. The PMU-phasor measurement unit, has been placed in the power system expansively, and the PMU measurements have become another important data source. The PMU measurements have high precision, synchronous sampling, and short updating cycle, and can measure the node state directly. For too high cost for PMU placement, the network is not observable only with PMU. The question How to Utilize the PMU measurements in the SE effectively, must be faced and solved now.The direct current transmission (DC) technology has been used more and more extensively in the power system for its advantages in power transmission over long distance and interconnection of alternating current (AC) networks, and especially it will play an important role in the power transmission from west to east and union of the whole country networks in China. Then more and more AC/DC systems will come into being, but the traditional SE arithmetic is not fit for the system with DC model, so exploring a fast and stable SE arithmetic for AC/DC systems has important value for the theory research and practical applications.The paper studies the problem of application of PMU measurements in the SE and the SE arithmetic for AC/DC systems.The PMU can measure the state variables of PMU nodes, then the SE is not necessary if the network is observable only with PMU, but this can’t be realized in a long time. The paper analyzes all kinds of application models of PMU measurements in SE, presents the feasible idea: utilizing both the PMU measurements and SCADAmeasurements in the nonlinear SE, improving the SE performance with PMU measurements characters, so points out the research direction and accelerates PMU measurements’ application.The PMU voltage phasor measurements can be used easily, but it is difficult to use PMU current phasor measurements, especially the current phase measurements in the nonlinear SE directly. The paper presents two new modes to utilize current phasor measurements: transforming the branch current phasors to branch powers; transforming the branch current phasors to the correlative node’s voltage phasors, then all PMU measurements can be utilized efficiently in the SE.The PMU can measure the state variables of PMU nodes, and the state variables of correlative nodes can be calculated with PMU measurements, so the local area is observable with PMU. The paper presents another model of PMU measurements application in SE: regarding the nodes’ status values measured or calculated as SE results of these nodes in the SE operation process, the model can reduce the scale of estimator, improve the SE equations’ numerical stability and convergence speed, however it is not necessary to be propitious to the estimation precision because of the PMU measurements error.The paper presents a double SE model for the first time, which includes both nonlinear and linear SE: at first, the nonlinear SE is done on the condition of taking nodes’ status values measured or calculated by PMU as estimation results, then the linear SE is done with both nonlinear SE results and PMU measurements. The model can improve the SE equations’ convergence speed and estimation precision simultaneously by PMU measurements. The model is the transition更多还原