CN105678337B - Information fusion method in intelligent substation fault diagnosis - Google Patents
Information fusion method in intelligent substation fault diagnosisDownload PDFInfo
- Publication number
- CN105678337B CN105678337BCN201610018296.9ACN201610018296ACN105678337BCN 105678337 BCN105678337 BCN 105678337BCN 201610018296 ACN201610018296 ACN 201610018296ACN 105678337 BCN105678337 BCN 105678337B
- Authority
- CN
- China
- Prior art keywords
- fault
- information
- protection
- fault diagnosis
- network
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Images
Classifications
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/251—Fusion techniques of input or preprocessed data
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Remote Monitoring And Control Of Power-Distribution Networks (AREA)
Abstract
Description
Translated fromChinese技术领域technical field
本发明涉及一种智能变电站故障诊断中的信息融合方法。The invention relates to an information fusion method in fault diagnosis of an intelligent substation.
背景技术Background technique
在电力系统中发生故障时，由于诊断对象的复杂性、测试手段的局限性、知识的不精确性，存在着大量的不确定因素。尤其是对于庞大的电力系统，各个元件之间的联系紧密复杂，其故障可能是多故障，关联故障等复杂形式。面对具有不确定性(包括不完整性)的信息，传统的变电站故障诊断大多是利用保护和断路器告警信息以及故障录波信息，对系统中的一次元件进行定位,在信息冗余度不够的情况下，容错性较低，诊断的精度和深度也不够。随着网络技术的发展，变电站朝着智能化的方向发展。相比于传统变电站二次系统采用硬接线，智能变电站一次系统采用智能装置(IED)，二次系统网络化。其信息的集成应用使信息利用的有效性得到极大提高，网络化使二次系统的各个工作环节可以得到有效监视，可观性和可控性得到极大提高。为实现更加高效、全面、深入的变电站故障诊断和评估方法提供了机会和实现手段When a fault occurs in the power system, there are a lot of uncertain factors due to the complexity of the diagnostic object, the limitation of testing methods, and the inaccuracy of knowledge. Especially for a huge power system, the connection between various components is close and complex, and its faults may be complex forms such as multiple faults and associated faults. In the face of information with uncertainty (including incompleteness), traditional substation fault diagnosis mostly uses protection and circuit breaker alarm information and fault recording information to locate the primary components in the system, and the information redundancy is not enough. In the case of , the fault tolerance is low, and the accuracy and depth of diagnosis are not enough. With the development of network technology, substations are developing towards the direction of intelligence. Compared with the traditional substation secondary system which adopts hard wiring, the intelligent substation primary system adopts intelligent device (IED), and the secondary system is networked. The integrated application of its information greatly improves the effectiveness of information utilization, and the network enables each work link of the secondary system to be effectively monitored, and the observability and controllability are greatly improved. Provides opportunities and means for realizing more efficient, comprehensive and in-depth substation fault diagnosis and evaluation methods
目前，在智能变电站故障诊断方面，神经网络，专家系统，Petri网等智能方法的应用非常广泛，虽然这些方法在一定程度上解决了故障时存在不确定因素对故障诊断的影响，具有一定的容错度，但是在出现较复杂的情况时仍然不能合理地给出诊断结果甚至造成误诊断，究其原因，一方面是系统发生故障时，故障情况的复杂性；另一方面是诊断所用信息源的单一性。因而仅从算法的角度去考虑，利用仅有的故障信息去完备化信息具有一定的局限性，已经不能从根本上提高诊断的可靠度。At present, in the fault diagnosis of intelligent substations, intelligent methods such as neural networks, expert systems, and Petri nets are widely used. Although these methods solve the influence of uncertain factors on fault diagnosis to a certain extent, they have certain fault tolerance. However, when a more complex situation occurs, the diagnosis result cannot be reasonably given or even misdiagnosis is caused. The reasons are, on the one hand, the complexity of the fault situation when the system fails; oneness. Therefore, only from the perspective of the algorithm, using the only fault information to complete the information has certain limitations, and cannot fundamentally improve the reliability of the diagnosis.
发明内容SUMMARY OF THE INVENTION
本发明为了解决上述问题，提出了一种智能变电站故障诊断中的信息融合方法，本方法首先利用贝叶斯网络建立针对元件的故障诊断模型，以系统故障时获取的来自一次系统和二次系统的多源信息为输入，通过贝叶斯网络推理计算获取元件故障概率值，从而判定出故障元件，可以有效适用于智能化变电站故障时故障元件的定位和变电站二次系统装置的诊断和分析。In order to solve the above problems, the present invention proposes an information fusion method in fault diagnosis of intelligent substation. In this method, a Bayesian network is used to establish a fault diagnosis model for components, and the information obtained from the primary system and the secondary system is obtained when the system fails. The multi-source information is input, and the probability value of component failure is obtained through Bayesian network inference calculation, so as to determine the faulty component, which can be effectively applied to the location of faulty components in intelligent substation faults and the diagnosis and analysis of secondary system devices of substations.
为了实现上述目的，本发明采用如下技术方案：In order to achieve the above object, the present invention adopts the following technical solutions:
一种智能变电站故障诊断中的信息融合方法，包括以下步骤：An information fusion method in fault diagnosis of an intelligent substation, comprising the following steps:
(1)根据电力系统中各个元件的保护配置，建立面向元件的贝叶斯网络故障诊断模型，将保护装置和断路器的可靠参数纳入基本输入，确定贝叶斯网络故障诊断模型各节点的之间的信任值；(1) According to the protection configuration of each element in the power system, establish a component-oriented Bayesian network fault diagnosis model, incorporate the reliable parameters of the protection device and circuit breaker into the basic input, and determine the relationship between each node of the Bayesian network fault diagnosis model. trust value between
(2)根据二次网络告警信息定义可靠参数，根据智能变电站的配置信息，建立装置-告警信息的关系矩阵；(2) Define reliable parameters according to the secondary network alarm information, and establish a device-alarm information relationship matrix according to the configuration information of the smart substation;
(3)利用误差反向传播算法进行贝叶斯网络故障诊断模型的网络参数的训练；(3) Using the error back-propagation algorithm to train the network parameters of the Bayesian network fault diagnosis model;
(4)将故障时获得的故障告警信息作为已训练好的故障诊断模型网络的输入，计算目标节点的值，计算元件的故障概率值。(4) The fault alarm information obtained when the fault occurs is used as the input of the trained fault diagnosis model network, the value of the target node is calculated, and the fault probability value of the element is calculated.
所述步骤(1)中，具体方法包括：In the step (1), the specific method includes:
(1-1)依据系统中各个元件的保护配置情况，以及元件故障、保护装置动作和断路器跳闸之间的内在逻辑关系，建立由Noisy-or、Noisy-and节点组成的贝叶斯网络故障诊断模型；(1-1) According to the protection configuration of each element in the system and the inherent logical relationship between element failure, protection device action and circuit breaker tripping, establish a Bayesian network fault composed of noise-or, noise-and nodes diagnostic model;
(1-2)对于每一个保护或者断路器输入节点，对应设置一个正确动作可靠度参数；(1-2) For each protection or circuit breaker input node, set a correct action reliability parameter correspondingly;
(1-3)以保护和断路器的动作信息以及各装置对应的可靠度参数信息作为贝叶斯网络模型的基本输入；(1-3) The action information of the protection and circuit breaker and the reliability parameter information corresponding to each device are used as the basic input of the Bayesian network model;
(1-4)计算Noisy-or、Noisy-and节点取真时的信任值。(1-4) Calculate the trust value when the Noisy-or and Noisy-and nodes are true.
进一步的，所述步骤(1-1)中，正确动作可靠度参数介于[0,1]之间，反应保护装置二次信息。Further, in the step (1-1), the correct action reliability parameter is between [0, 1], and the secondary information of the protection device is reflected.
所述步骤(1-4)中，Noisy-or、Noisy-and节点取真时的信任值计算方法为：In the step (1-4), the trust value calculation method when the Noisy-or and Noisy-and nodes are true is:
其中参数cij是Nj单个前提Ni取真值时对Nj真的认可程度，即从节点Ni到节点Nj条件概率，条件概率通过参数训练得到。The parameter cij is the true recognition degree of Nj when the single premise Ni of Nj takes the true value, that is, the conditional probability from node Ni to node Nj , and the conditional probability is obtained through parameter training.
所述步骤(2)中，具体方法包括：In the step (2), the specific method includes:
(2-1)保护装置和断路器的可靠度参数根据二次网络告警信息进行定义；(2-1) The reliability parameters of the protection device and the circuit breaker are defined according to the secondary network alarm information;
(2-2)根据故障时与各装置保护相关的二次告警信息和与保护装置相关的所有可能的告警信息，定义保护装置正确动作的可靠度；(2-2) Define the reliability of the correct action of the protection device according to the secondary alarm information related to the protection of each device and all possible alarm information related to the protection device at the time of failure;
(2-3)根据故障时与断路器相关的二次告警信息和与断路器相关的所有可能的告警信息，定义断路器正确动作的可靠度；(2-3) Define the reliability of the correct action of the circuit breaker according to the secondary alarm information related to the circuit breaker and all possible alarm information related to the circuit breaker when the fault occurs;
(2-4)根据变电站的配置信息，建立装置-告警信息的关系矩阵，寻找矩阵中与告警信息相关的装置，并统计计算各装置的可靠度。(2-4) According to the configuration information of the substation, establish a device-alarm information relationship matrix, find the devices related to the alarm information in the matrix, and calculate the reliability of each device statistically.
所述步骤(2-1)中，具体方法为：对于保护装置，与其相关的二次网络告警信息包括：保护装置自检信息、SV报文通信链路状态信息和GOOSE跳闸通信链路状态信息；对于断路器，与其相关的二次网路告警信息包括：保护装置自检信息和GOOSE跳闸通信链路状态信息。In the step (2-1), the specific method is: for the protection device, the related secondary network alarm information includes: protection device self-check information, SV message communication link status information and GOOSE trip communication link status information ; For circuit breakers, the related secondary network alarm information includes: protection device self-check information and GOOSE trip communication link status information.
进一步的，所述步骤(2-2)中，定义RP为保护装置动作的可靠度：
其中Si是故障时与保护P相关的二次告警信息，Sn是与保护P相关的所有可能的告警信息。Among them, Si is the secondary alarm information related to the protection P when the fault occurs, andSn is all possible alarm information related to the protection P.
所述步骤(2-3)中，定义RB为断路器装置动作的可靠度：
其中Si是故障时与断路器B相关的二次告警信息，Sn是与断路器B相关的所有可能的告警信息。Among them, Si is the secondary alarm information related to the circuit breaker B when the fault occurs, andSnis all possible alarm information related to the circuit breaker B.
所述步骤(3)中，具体方法为：针对Noisy-or、Noisy-and节点组成的贝叶斯网络故障诊断模型，利用误差反向传播算法进行参数的训练，利用梯度下降算法使得目标变量的测量值和计算值之间的均方差达到最小。In the step (3), the specific method is: for the Bayesian network fault diagnosis model composed of Noisy-or and Noisy-and nodes, use the error back propagation algorithm to train the parameters, and use the gradient descent algorithm to make the target variable The mean square error between the measured and calculated values is minimized.
所述步骤(4)中，具体方法为：对于已训练好的不同类型元件的故障网络模型，将故障时获得的故障告警信息作为网络的输入：当获得的保护或断路器动作时，保护或断路器节点的输入值应该为：RP(RB)；当获得的保护或断路器没有动作时，保护或断路器节点的输入值应该为：1-RP(RB)；当一次系统信息缺失时，保护或断路的状态信息不明确，也将该节点的输入定义为RP(RB)。In the step (4), the specific method is: for the fault network models of different types of components that have been trained, the fault alarm information obtained during the fault is used as the input of the network: when the obtained protection or circuit breaker operates, the protection or The input value of the circuit breaker node should be: RP (RB ); when the protection or circuit breaker is not actuated, the input value of the protection or circuit breaker node should be: 1-RP( RB ); when the primary system When the information is missing, the status information of protection or disconnection is not clear, and the input of this node is also defined as RP (RB ).
所述步骤(4)中，将相应节点的输入值输入到网络中，逐层推理计算出目标节点的信任值从而获得该元件的故障概率值，对于故障时可能出现的所有故障元件都进行计算，并对其故障概率值进行排序，通过已设定的故障概率门槛值判断元件是否故障。In the step (4), the input value of the corresponding node is input into the network, and the trust value of the target node is calculated layer-by-layer inference to obtain the failure probability value of the element, and all the faulty elements that may occur during the failure are calculated. , and sort its failure probability value, and judge whether the component fails through the set failure probability threshold value.
本发明的有益效果为：The beneficial effects of the present invention are:
(1)本发明结合智能变电站的特点，充分利用二次网络信息，通过定义保护装置和断路器正确动作的可靠度参数，解决了一次系统诊断过程中只利用保护和断路器信息时存在的缺陷；(1) The present invention combines the characteristics of the intelligent substation, makes full use of the secondary network information, and solves the defect that only the protection and circuit breaker information is used in the primary system diagnosis process by defining the reliability parameters of the correct action of the protection device and the circuit breaker. ;
(2)本发明从拓宽信息源的角度出发，融合多源信息，同时结合贝叶斯网络的强大容错功能，提高诊断的精确度和可靠度。(2) From the perspective of expanding information sources, the present invention integrates multi-source information, and at the same time combines the powerful fault-tolerant function of the Bayesian network to improve the accuracy and reliability of diagnosis.
附图说明Description of drawings
图1(a)为本发明的线路故障诊断模型示意图；Fig. 1 (a) is the schematic diagram of the line fault diagnosis model of the present invention;
图1(b)为本发明的母线故障诊断模型示意图；Figure 1(b) is a schematic diagram of a bus fault diagnosis model of the present invention;
图1(c)为本发明的变压器故障诊断模型示意图；Figure 1 (c) is a schematic diagram of a transformer fault diagnosis model of the present invention;
图2为本发明的训练贝叶斯网络故障诊断模型的流程示意图。FIG. 2 is a schematic flowchart of training a Bayesian network fault diagnosis model according to the present invention.
具体实施方式：Detailed ways:
下面结合附图与实施对本发明作进一步说明。The present invention will be further described below in conjunction with the accompanying drawings and implementation.
结合智能变电站的特点，当系统中发生故障时，不仅一次系统会发生状态上的变化，网络化的二次系统也会发出很多告警信息，二次告警信息不仅在一定程度上反映了故障特征，还可以反映出装置本身的状态信息。在进行智能变电站故障诊断时将变电站二次系统故障信息考虑在内，可以极大提高信息的冗余度，避免出现由于信息不足造成故障诊断不精确的问题。利用二次告警信息定义保护装置和断路器正确动作的可靠度参数，解决了一次系统诊断过程中只利用保护和断路器信息时存在的缺陷，从拓宽信息源的角度出发，结合一、二次系统的信息，给系统的诊断提供更加全面的信息，提高诊断的精确度。Combined with the characteristics of smart substations, when a fault occurs in the system, not only will the primary system change in state, but the networked secondary system will also issue a lot of alarm information. The secondary alarm information not only reflects the fault characteristics to a certain extent, but also It can also reflect the status information of the device itself. Taking into account the fault information of the secondary system of the substation in the fault diagnosis of the intelligent substation can greatly improve the redundancy of the information and avoid the problem of inaccurate fault diagnosis due to insufficient information. Using the secondary alarm information to define the reliability parameters of the correct action of the protection device and the circuit breaker, it solves the defect that only the protection and circuit breaker information is used in the primary system diagnosis process. The system information provides more comprehensive information for the system diagnosis and improves the diagnosis accuracy.
以融合二次信息的智能变电站故障诊断为核心，现对发明内容做进一步说明。Taking the intelligent substation fault diagnosis integrating secondary information as the core, the content of the invention is further explained.
步骤一：故障诊断模型的建立。Step 1: Establish a fault diagnosis model.
1)如图1(a)、图1(b)和图1(c)所示，依据系统中各个元件的保护配置情况，以及元件故障、保护装置动作和断路器跳闸之间的内在逻辑关系，建立由Noisy-or、Noisy-and节点组成的贝叶斯网络故障诊断模型。1) As shown in Figure 1(a), Figure 1(b) and Figure 1(c), according to the protection configuration of each component in the system, as well as the inherent logical relationship between component failure, protection device action and circuit breaker tripping , to establish a Bayesian network fault diagnosis model composed of Noisy-or, Noisy-and nodes.
以线路为例，线路L发生故障时，理论上线路两侧的保护都应该动作使其相应的断路器跳闸，线路两侧的保护构成Noisy-and节点。对于每一侧来说，保护又可以分为三类：主保护、后备保护和相邻元件的远后备保护。这三类保护中的任一类动作使其对应断路器跳闸，都可以切除故障，因此这三类保护组成的是Noisy-or节点。保护装置正常动作的情况下，保护装置和断路器动作应该是一致的，因此保护和它对应断路器组成Noisy-and节点。Taking a line as an example, when a fault occurs on line L, in theory, the protections on both sides of the line should act to trip the corresponding circuit breakers, and the protections on both sides of the line constitute a noise-and node. For each side, the protection can be divided into three categories: main protection, backup protection and remote backup protection of adjacent elements. Any action of these three types of protection makes the corresponding circuit breaker trip, and the fault can be removed. Therefore, these three types of protection are composed of noise-or nodes. When the protection device operates normally, the action of the protection device and the circuit breaker should be consistent, so the protection and its corresponding circuit breaker form a Noisy-and node.
2)对于每一个保护或者断路器输入节点，对应都有一个正确动作可靠度参数，该参数介于[0,1]之间，反应保护装置二次信息，如图1(a)中的保护节点MLP，通过定义可靠度参数改变了原始的非0即1的输入，其可靠度参数RP的定义及计算在技术方案中已详细描述。2) For each protection or circuit breaker input node, there is a corresponding correct action reliability parameter, which is between [0, 1] and reflects the secondary information of the protection device, such as the protection in Figure 1(a). The node MLP changes the original input of either 0 or 1 by defining the reliability parameter, and the definition and calculation of the reliability parameterRP have been described in detail in the technical solution.
3)当系统中发生故障时，首先根据断电区域识别出可疑故障元件，针对每一个可疑故障元件建立对应的故障诊断模型。3) When a fault occurs in the system, first identify the suspected faulty components according to the power-off area, and establish a corresponding fault diagnosis model for each suspected faulty component.
步骤二：可靠度参数计算。Step 2: Calculation of reliability parameters.
1)如技术方案中详述，RP为保护装置动作的可靠度：
其中Si是故障时与保护P相关的二次告警信息，Sn是与保护P相关的所有可能的告警信息。Among them, Si is the secondary alarm information related to the protection P when the fault occurs, andSn is all possible alarm information related to the protection P.
2)RB为断路器动作的可靠度：
其中Si是故障时与断路器B相关的二次告警信息，Sn是与断路器B相关的所有可能的告警信息。Among them, Si is the secondary alarm information related to the circuit breaker B when the fault occurs, andSnis all possible alarm information related to the circuit breaker B.
由于断路器和保护之间存在着一定的逻辑关系即断路器受保护的控制，因而当出现保护装置错误的告警信息时，相应的断路器正确动作的可靠性也降低，保护和断路器的关系可以通过下述装置-告警信息的关系矩阵反映。Because there is a certain logical relationship between the circuit breaker and the protection, that is, the circuit breaker is controlled by the protection, so when the wrong alarm information of the protection device occurs, the reliability of the correct action of the corresponding circuit breaker is also reduced. The relationship between the protection and the circuit breaker It can be reflected by the following device-alarm information relationship matrix.
3)根据智能变电站的配置信息，建立装置-告警信息的关系矩阵DSm×n3) According to the configuration information of the smart substation, establish the relationship matrix DSm×n of the device-alarm information
其中每一行Di表示变电站的一个装置，其中包括保护，断路器智能终端，交换机等各种IED智能装置，共m维；每一列Sj表示变电站中可能的告警信息，共n维。矩阵中的元素非0即1，当装置Di和告警信息Sj之间存在关联关系时为1，否则为0。Each row Di represents a device in the substation, including various IED intelligent devices such as protection, circuit breaker intelligent terminals, switches, etc., with a total of m dimensions; each column Sj represents possible alarm information in the substation, with a total of n dimensions. The elements in the matrix are either 0 or 1, and are 1 when there is a correlation between the device Di and the alarm information Sj , and 0 otherwise.
同时根据系统的网络拓扑结构可以定义元件-保护矩阵EPa×b，保护-断路器矩阵PBb×c，将元件的保护配置情况通过数学的形式表达出来。如下式所示，行元素表示元件Ei，列元素表示保护Pi，元素1表示元件Ei受保护Pi保护。为了区分保护的类别，例如主保护，后备保护等，可以定义多个这样的矩阵。At the same time, the element-protection matrix EPa×b and the protection-circuit breaker matrix PBb×c can be defined according to the network topology of the system, and the protection configuration of the element can be expressed in mathematical form. As shown in the following formula, the row element represents the element Ei , the column element represents the protectionPi , and the
发生故障时，针对可疑故障元件，通过矩阵搜索的方法找到元件对应的保护及其相关的断路器。同时，对于二次网络中出现大量告警信息，搜寻矩阵DSm×n找到元件对应保护和断路器的相关告警信息，统计并计算各装置正确动作的可靠度。When a fault occurs, for the suspected faulty component, the corresponding protection of the component and its related circuit breaker are found by the method of matrix search. At the same time, for a large amount of alarm information in the secondary network, the matrix DSm×n is searched to find the relevant alarm information of the protection and circuit breakers corresponding to the components, and the reliability of the correct operation of each device is counted and calculated.
步骤三：网络参数的训练。Step 3: Training of network parameters.
针对步骤一中所述的Noisy-or、Noisy-and节点组成的贝叶斯网络故障诊断模型，利用误差反向传播算法进行参数的训练。训练的基本方法已在技术方案中详述，流程图(如图2)具体介绍训练步骤。For the Bayesian network fault diagnosis model composed of Noisy-or and Noisy-and nodes described in
对于建立好的故障诊断模型，利用误差反向传播算法进行网络参数的训练。For the established fault diagnosis model, the error back propagation algorithm is used to train the network parameters.
针对步骤一种所述的Noisy-or、Noisy-and节点组成的贝叶斯网络故障诊断模型，利用误差反向传播算法进行参数的训练。误差反向传播算法通常用于多层神经网络的训练，利用梯度下降算法使得目标变量的测量值和计算值之间的均方差达到最小。下式给出了均方差的计算公式：For the Bayesian network fault diagnosis model composed of Noisy-or and Noisy-and nodes described in
其中
其中η是学习因子，δj是节点Nj的误差。对于输出节点：where η is the learning factor andδj is the error at node Nj. For output nodes:
对隐藏的节点，节点Nj反向传播至其父节点Ni的误差为：For hidden nodes, the error of node Nj backpropagating to its parent node Ni is:
每一次训练根据误差δj计算梯度Δij，用梯度去修正参数，再重新训练计算，直至误差满足要求。In each training, the gradient Δij is calculated according to the error δj , the parameters are corrected by the gradient, and then the calculation is re-trained until the error meets the requirements.
对于每一个不同类型的元件都要进行训练，如表1所示，为线路故障模型的训练样本，通常为了获得更加优化的网络参数，可以根据实际运行情况增加训练样本。For each different type of component, training is required, as shown in Table 1, which is the training sample of the line fault model. Usually, in order to obtain more optimized network parameters, training samples can be added according to the actual operation.
表1.线路故障模型训练Table 1. Line fault model training
步骤四：元件故障概率计算Step 4: Calculation of Component Failure Probability
对于已训练好的不同类型元件的故障网络模型，将故障时获得的故障告警信息作为网络的输入：当获得的保护或断路器动作信息为1(即保护或断路器动作)时，保护或断路器节点的输入值应该为：RP(RB)；当获得的保护或断路器动作信息为0(即保护或断路器不动作)时，保护或断路器节点的输入值应该为：1-RP(RB)；当一次系统信息缺失时，保护或断路的状态信息不明确，也将该节点的输入定义为RP(RB)。For the fault network models of different types of components that have been trained, the fault alarm information obtained during the fault is used as the input of the network: when the obtained protection or circuit breaker action information is 1 (that is, the protection or circuit breaker action), the protection or circuit breaker The input value of the circuit breaker node should be: RP (RB ); when the obtained protection or circuit breaker action information is 0 (that is, the protection or circuit breaker does not act), the input value of the protection or circuit breaker node should be: 1- RP (RB ); when the primary system information is missing, and the status information of protection or disconnection is not clear, the input of this node is also defined as RP (RB ).
将相应节点的输入值输入到网络中，逐层推理计算出目标节点的信任值从而获得该元件的故障概率值。对于故障时可能出现的所有故障元件都要进行计算，并对他们的故障概率值进行排序，通过已设定的故障概率门槛值判断元件是否故障。通常认为故障概率值明显高于其他元件的为故障元件。The input value of the corresponding node is input into the network, and the trust value of the target node is calculated by layer-by-layer inference to obtain the failure probability value of the element. All faulty components that may occur in the event of a fault should be calculated, and their fault probability values should be sorted, and whether the components are faulty will be judged by the set fault probability threshold value. It is generally considered that the failure probability value is significantly higher than that of other components as the faulty component.
上述虽然结合附图对本发明的具体实施方式进行了描述，但并非对本发明保护范围的限制，所属领域技术人员应该明白，在本发明的技术方案的基础上，本领域技术人员不需要付出创造性劳动即可做出的各种修改或变形仍在本发明的保护范围以内。Although the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, they do not limit the scope of protection of the present invention. Those skilled in the art should understand that on the basis of the technical solutions of the present invention, those skilled in the art do not need to pay creative work. Various modifications or deformations that can be made are still within the protection scope of the present invention.
Claims (8)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610018296.9ACN105678337B (en) | 2016-01-12 | 2016-01-12 | Information fusion method in intelligent substation fault diagnosis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610018296.9ACN105678337B (en) | 2016-01-12 | 2016-01-12 | Information fusion method in intelligent substation fault diagnosis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105678337A CN105678337A (en) | 2016-06-15 |
| CN105678337Btrue CN105678337B (en) | 2020-02-04 |
Family
ID=56300189
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610018296.9AExpired - Fee RelatedCN105678337B (en) | 2016-01-12 | 2016-01-12 | Information fusion method in intelligent substation fault diagnosis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105678337B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106446393A (en)* | 2016-09-18 | 2017-02-22 | 广东电网有限责任公司电力科学研究院 | Power transmission network component failure diagnosis method |
| CN106646068A (en)* | 2017-01-22 | 2017-05-10 | 国网湖北省电力公司检修公司 | Method for diagnosing defects of intelligent substation secondary system based on multi-parameter information fusion |
| CN110298409A (en)* | 2019-07-03 | 2019-10-01 | 广东电网有限责任公司 | Multi-source data fusion method towards electric power wearable device |
| DE102020120539A1 (en)* | 2020-08-04 | 2022-02-10 | Maschinenfabrik Reinhausen Gmbh | Device for determining an error probability value for a transformer component and a system with such a device |
| CN112415331B (en)* | 2020-10-27 | 2024-04-09 | 中国南方电网有限责任公司 | Power grid secondary system fault diagnosis method based on multi-source fault information |
| CN113807461A (en)* | 2021-09-27 | 2021-12-17 | 国网四川省电力公司电力科学研究院 | Transformer fault diagnosis method based on Bayesian network |
| CN114662585B (en)* | 2022-03-18 | 2025-09-26 | 南方电网科学研究院有限责任公司 | A fault detection method and system for primary and secondary fusion equipment |
| CN116008730B (en)* | 2023-02-06 | 2025-09-16 | 东北电力大学 | DC power distribution network fault diagnosis method based on Bayesian network information fusion |
| CN117169717A (en)* | 2023-09-11 | 2023-12-05 | 江苏微之润智能技术有限公司 | Motor health assessment method and device based on single chip microcomputer and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102255764A (en)* | 2011-09-02 | 2011-11-23 | 广东省电力调度中心 | Method and device for diagnosing transmission network failure |
| CN102497024A (en)* | 2011-12-16 | 2012-06-13 | 广东电网公司茂名供电局 | Intelligent warning system based on integer programming |
| CN103986238A (en)* | 2014-05-28 | 2014-08-13 | 山东大学 | Intelligent substation fault diagnosis method based on probability weighted bipartite graph method |
- 2016
- 2016-01-12CNCN201610018296.9Apatent/CN105678337B/ennot_activeExpired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102255764A (en)* | 2011-09-02 | 2011-11-23 | 广东省电力调度中心 | Method and device for diagnosing transmission network failure |
| CN102497024A (en)* | 2011-12-16 | 2012-06-13 | 广东电网公司茂名供电局 | Intelligent warning system based on integer programming |
| CN103986238A (en)* | 2014-05-28 | 2014-08-13 | 山东大学 | Intelligent substation fault diagnosis method based on probability weighted bipartite graph method |
Non-Patent Citations (1)
| Title |
|---|
| 基于贝叶斯网络的电网故障诊断方法;霍利民;《华北电力大学学报》;20040625;第31卷(第3期);第30-34页* |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105678337A (en) | 2016-06-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105678337B (en) | Information fusion method in intelligent substation fault diagnosis | |
| CN104297637B (en) | The power system failure diagnostic method of comprehensive utilization electric parameters and time sequence information | |
| CN109633372B (en) | Membrane system-based automatic power system fault diagnosis method | |
| CN103840967B (en) | A kind of method of fault location in power telecom network | |
| CN103308824B (en) | Power system fault diagnostic method based on probability Petri net | |
| CN113900844B (en) | Fault root cause positioning method, system and storage medium based on service code level | |
| CN105183952B (en) | A kind of power transmission network method for diagnosing faults based on separation time Fuzzy Petri Net | |
| CN107656176A (en) | A kind of electric network failure diagnosis method based on improvement Bayes's Petri network | |
| CN111008454B (en) | Intelligent substation reliability assessment method based on information physical fusion model | |
| CN103001328A (en) | A Fault Diagnosis and Evaluation Method for Smart Substation | |
| CN103901320A (en) | Method for diagnosing power system fault considering multi-source data | |
| CN103197168B (en) | Realize the method for fault diagnosis control based on the event set chain of causation in electric system | |
| CN104766246A (en) | Power system fault diagnosis method based on timing order fuzzy Petri net | |
| CN103986238A (en) | Intelligent substation fault diagnosis method based on probability weighted bipartite graph method | |
| CN118473910B (en) | Power Internet of Things fault detection method and system based on edge-cloud collaboration | |
| Ren et al. | Research on fault location of process-level communication networks in smart substation based on deep neural networks | |
| CN110018390A (en) | Hierarchical fuzzy petri net electric network failure diagnosis method based on comprehensive variable weight | |
| CN106557607A (en) | A kind of data summarization method of power transmission and transformation fault detection system | |
| CN104764979A (en) | Virtual information fusion power grid alarming method based on probabilistic reasoning | |
| CN104600680A (en) | Intelligent alarming method based on data fusion | |
| CN111695231A (en) | Practical reliability analysis method for complex power distribution network information physical system | |
| CN112684300B (en) | Active power distribution network fault diagnosis method and device using bidirectional monitoring information | |
| CN113740666B (en) | A method for locating root faults of alarm storm in data center power system | |
| CN114252727A (en) | A rapid diagnosis method for power grid faults based on artificial intelligence technology | |
| CN112285484B (en) | Power system fault diagnosis information fusion method and device based on deep neural network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | Granted publication date:20200204 | |
| CF01 | Termination of patent right due to non-payment of annual fee |