计及节点时空关联性的电力系统预测辅助状态估计

发布时间：2018-06-22 09:37

本文选题：预测辅助状态估计 + 时间序列分析　；参考：《西南交通大学》2017年硕士论文

【摘要】：电力系统状态估计对电力系统的可靠和安全运行至关重要。电力系统状态估计可以根据量测系统采集到的信息计算出未量测的电气量,得到的结果将被送往能量管理中心用作事故分析、负荷预测和最优潮流计算等。然而,传统的静态状态估计由于缺乏预测能力因而实际中不能得到状态的预测值。该方法的缺陷使研究者们提出了能在下一时刻提供预测状态的预测辅助状态估计。预测辅助状态估计根据系统当前的状态使用物理模型来预测系统下一时刻的状态。一旦得到下一时刻的预测向量,该向量便用于滤波从而得到最优的估计值。预测辅助状态估计有助于发现,识别和排除坏数据从而提高估计效果。目前预测辅助状态估计算法都假定状态转移矩阵是对角的,这意味着状态变量之间被假定为相互独立。这个假设在电力系统中有微电网及分布式和间歇性新能源时是不成立的。由于分布式能源的接入,未来的电网将变得越来越分散化并导致系统节点数量的减少及带来大量的功率注入。因此,系统中枢纽节点将失去它们在电网中的影响力而使状态间的空间关联性变得越来越可观。当系统中某个量测量因量测设备故障而丢失量测信息时而使系统不可观测时,考虑状态量之间的空间相关性可以使系统重建可观测性。基于此,本文提出了一种能考虑节点时空关联性的状态转移矩阵应用于预测辅助状态估计,主要工作如下:(1)本文首先对时间序列分析常用的几种建模方法作了简要介绍,并通过实验作了对比分析。(2)在考虑节点时空关联性的状态预测方面,给出了一个稀疏、非对角及时变的状态转移矩阵用于状态预测。主要内容是对节点间的时间关联性建立自回归模型,空间关联性建立向量自回归模型,并使用该模型在系统正常运行和系统中节点量测信息丢失情况下与目前广泛应用的指数平滑模型做了实验对比。(3)在考虑节点时空关联性的预测辅助状态估计方面,将能考虑节点时空相关性的自回归及向量自回归混合模型用于状态预测。主要内容是在系统正常运行和系统中节点量测信息丢失情况下,使用该模型与席尔瓦预测模型进行预测辅助状态估计,对得到的结果进行了分析及总结。
[Abstract]:The state estimation of the power system is very important for the reliable and safe operation of the power system. The state estimation of the power system can be calculated according to the information collected by the measurement system, and the results will be sent to the energy management center for accident analysis, load forecasting and optimal power flow calculation. However, the traditional static state of the power system is used. As a result of lack of predictability, the predicted value of a state can not be obtained in practice. The defect of this method makes the researchers propose a prediction assistant state estimate that can provide the predictive state at the next moment. The prediction auxiliary state estimation is based on the current state of the system using the physical model to predict the state of the system at the next moment. The prediction auxiliary state estimation helps to discover, identify and exclude bad data to improve the estimation effect. The present predictive auxiliary state estimation algorithms assume that the state transition matrix is diagonal, which means that the state variables are assumed to be phases. It is independent. This assumption is unfounded when there is a micro grid and distributed and intermittent new energy in the power system. Due to the access of distributed energy, the future power grid will become more and more decentralized and lead to a reduction in the number of nodes and a large amount of power injection. Therefore, the hub nodes in the system will lose them in the power grid. The influence of the system makes the spatial correlation between states become more and more observable. When one measure of the system is not observable when measuring the fault of the equipment, the system can reconstruct the observability by considering the spatial correlation between the state. Based on this, this paper proposes a kind of consideration of the spatiotemporal Association of nodes. The main work of the state transition matrix is as follows: (1) first of all, this paper briefly introduces several modeling methods commonly used in time series analysis, and makes a comparative analysis through experiments. (2) a sparse, non diagonal and timely change state is given in consideration of the state prediction of the spatiotemporal correlation of nodes. The transfer matrix is used for state prediction. The main content is to establish an autoregressive model for the time correlation between nodes, and to establish a vector autoregressive model for spatial correlation. The model is compared with the exponential smoothing model which is widely used in the normal operation of the system and the loss of measurement information of nodes in the system. (3) it is considered. In the aspect of prediction aided state estimation of node spatiotemporal correlation, the autoregressive and vector autoregressive hybrid model which can consider the spatiotemporal correlation of nodes is used for state prediction. The main content is to use the model and the Silva prediction model to estimate the auxiliary state estimation under the condition of the system normal operation and the loss of the node measurement information in the system. The results are analyzed and summarized.
【学位授予单位】：西南交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM732

【相似文献】