CN103675600B - Based on the Fault Diagnosis of Distribution Network system and method for topological knowledge - Google Patents

Abstract

The invention relates to a distribution network fault diagnosis system and method based on topology knowledge and topology search method. The system includes a resource layer, a knowledge layer and a diagnosis layer. The resource layer includes a fault recorder, a monitoring terminal, and a power distribution equipment ledger. , topology files and databases, monitoring terminals are used to collect switch status information after faults, fault recorders are used to obtain waveforms after faults, topology files are used to provide electrical wiring diagrams of the entire power grid, and power distribution equipment ledgers are used to obtain power distribution The operating condition of the equipment when it fails; the knowledge layer includes a knowledge base that provides information for the diagnosis layer; the diagnosis layer includes a fault diagnosis server, and a fault diagnosis program module is arranged on the fault diagnosis server. This method implements topology knowledge formation and fault diagnosis based on the above system. The system and method can analyze accidents after failures, provide references for operating personnel, save manpower, financial resources, and time, and the system operates efficiently and quickly, and has good use effects.

Description

Distribution Network Fault Diagnosis System and Method Based on Topological Knowledge

technical field

The invention relates to the technical field of distribution network fault diagnosis, in particular to a distribution network fault diagnosis system and method based on topology knowledge and a topology search method.

Background technique

The traditional distribution network fault diagnosis is based on fault information such as measurement and control, protection, and wave recording, and adopts a certain diagnostic mechanism to determine the fault feeder and fault location, isolate the fault area, minimize the power outage area and shorten the power outage time. The power quality of the power grid is analyzed, excluding the accident analysis function after the failure, such as the analysis of the cause of the accident (human factors, natural factors, equipment failure, etc.), the scope of the accident, and the power outage loss caused by the accident. This requires manual inspection of the accident site, investigation and statistics to give a corresponding report. The distribution network is a part where accidents occur frequently. Therefore, accident analysis consumes a lot of manpower and time.

With the construction of the distribution network automation system, the amount of data that can be used for fault diagnosis is increasing, and the sources are becoming more and more extensive. These data contain redundant information, and they will flow into the dispatching center within a period of time after the fault. It is very difficult for operators to diagnose faults manually, and how to effectively use these data is a problem that needs to be solved urgently. Therefore, the traditional fault diagnosis framework still has a lot of room for improvement in terms of speed and accuracy.

In addition, grid topology analysis is the basis of fault diagnosis, and a good grid topology data structure can greatly improve the efficiency of fault diagnosis. Traditional grid topology representation methods include matrix method and adjacency table method, etc. Matrix-based grid topology analysis method needs to identify the sub-electrical islands formed by the entire distribution network after the circuit breaker and other switching devices trip. Different branch linked lists need to be constructed for each node of the power grid to form a node tree. The entire search process requires three types of data structures in the form of a matrix, an electrical island, and a linked list. The search process is cumbersome and inefficient. However, the traditional adjacency table method needs to input the node-branch information of the entire power grid when establishing the initial node-branch relationship table and branch-access flag-node relationship table, so the workload is very huge.

Contents of the invention

The purpose of the present invention is to overcome the deficiencies of the prior art and provide a distribution network fault diagnosis system and method based on topology knowledge. Time, and the system runs efficiently and quickly, and the use effect is good.

In order to achieve the above purpose, the technical solution adopted by the present invention is: a distribution network fault diagnosis system based on topology knowledge, including a resource layer, a knowledge layer and a diagnosis layer;

The resource layer includes a fault recorder, a monitoring terminal, a power distribution equipment ledger, a topology file and a database. The monitoring terminal is used to collect switch status information after a fault, and the fault recorder is used to obtain waveforms after a fault. The topology file is used to provide an electrical wiring diagram of the entire power grid, and the power distribution equipment ledger is used to obtain the operating conditions of various power distribution equipment when they fail;

The knowledge layer includes a knowledge base that provides information for the diagnosis layer, and the knowledge base includes a topology knowledge base and a fault knowledge base; the topology knowledge base is formed as follows: two classes are created on the basis of topology packages in IEC61970: The Vertex class and the adjNode class correspond to the head node class and the table node class of the adjacency list respectively; the three fields of the Vertex class are the access flag, the vertex name, and the first node pointer, and the Vertex class includes busbars, lines, transformers and power supplies , the adjNode class reflects the electrical components that are connected to the Vertex class, that is, the ConnectivityNode connected to a Vertex device contains n terminals, and the corresponding singly linked list in the adjacency list has n-1 adjNodes, and each adjNode class has three The domains are the name of the associated electrical component, the connected circuit breaker, and the next domain; in addition, according to the definition of the ConnectivityNode class in the topology package in IEC61970, a Y node is added between two directly connected circuit breakers; the judgment method of the Y node is : If a ConnectivityNode object connects two circuit breakers and a line end, it will be reserved as a Y node; the disconnector in the "closed" state is regarded as a wire;

The diagnosis layer includes a fault diagnosis server, and a fault diagnosis program module is arranged on the fault diagnosis server.

The present invention also provides a diagnosis method corresponding to a distribution network fault diagnosis system based on topology knowledge, including topology knowledge formation and fault diagnosis;

The formation of the topology knowledge is carried out in the following steps:

(1) When a fault occurs, the resource layer at the bottom collects various fault information, collects the switch status information after the fault from the monitoring terminal, obtains the waveform after the fault from the fault recorder, and obtains various distribution equipment from the power distribution equipment ledger. The operation of electrical equipment in the event of a fault, and the electrical wiring diagram of the entire grid is obtained from the topology file;

(2) Provide RDF data files conforming to CIM-based model rules, or CIM-XML model data files;

(3) Integrate SVG graphic elements to form wiring diagrams of each plant station and distribution network feeder, and provide an interface to the upper layer in the form of CIS services;

(4) Query the wiring diagrams of each plant station and distribution network feeder through LINQtoXML, find the names, connection points, and connection components of each electrical component, and store them in the topology library; the topology library exists in the form of DataSet in the memory , is an offline container of data defined in .NET, does not contain the concept of database connection, but is just a set of data sets in memory; the topology library is stored in the DataSet in the form of an adjacency list, and is distributed with the system in the form of a server start up;

(5) Search the topology of the entire network to obtain the topology of the entire network;

The fault diagnosis is carried out as follows: after obtaining the topological structure of the whole network, combined with the switch state information after the fault obtained by the monitoring terminal of the resource layer, the feeders disconnected due to faults and the feeders on these feeders are obtained through the fault diagnosis program module. Connected distribution transformers, the information of these outage distribution transformers is counted through the distribution equipment ledger; the fault diagnosis program module establishes an interface with the power distribution GIS, and calls the geographic information of the power outage line from the power distribution GIS, that is, the power outage feeder The geographic location, and then transmit the geographical information to the weather forecast system and lightning positioning system to obtain the weather conditions and lightning conditions in the fault area at the time of the fault, and then use the power distribution equipment ledger to obtain the power distribution equipment in the fault area at the time of the fault Finally, according to the weather, lightning, and equipment operation, a preliminary judgment on the cause of the failure is given.

Further, in step (204), the formation process of the topology library is:

(401) The application starts an empty DataSet according to the DataSet template;

(402) Search for relevant data according to the definition of the DataSet template from the wiring diagram of each plant and station, and fill the topology library; this part of the function is provided to the application layer in the form of an interface in the data filling layer;

(403) Use the same method to search for feeders, distribution transformers, and connected switchgear.

Compared with the prior art, the beneficial effects of the present invention are:

1. The proposed fault diagnosis and analysis method makes up for the shortcomings of the traditional fault diagnosis without the fault analysis function, and makes a preliminary judgment on the cause of the fault, the severity of the fault, and the scope of the fault, so as to provide reference for the operator and save manpower, financial resources and time .

2. The proposed representation method of power grid topology in memory in fault diagnosis directly maps the grid wiring diagram conforming to the CIM model into an adjacency table, which solves the problem of complex initialization, improves the search efficiency of the plant wiring xml file, and can Represents various complex wiring methods, which improves the speed of grid topology.

3. Use LINQtoXML to find the names, connection points, and connection elements of each electrical component for the wiring diagram of the plant and station, and store it in the topology database. The calculation based on the power grid topology can be obtained by searching the topology database, thus avoiding the need to search the XML of the wiring diagram of the plant and station. Frequent operation of files can not only improve the running speed of the system, but also facilitate the distributed maintenance and management of data.

4. The established adjacency list starts with the system in the form of a server, and its advantages are mainly reflected in: 1) The traversal algorithm based on the adjacency list is quite mature, and each client can write different algorithms according to different needs to get the desired data . 2) It starts once and stays in memory for a long time, without repeated generation, and provides the topology data required by different clients at the same time in a multi-threaded access mode, which improves the parallel capability of the system.

Description of drawings

FIG. 1 is a structural diagram of a diagnostic system of an embodiment of the present invention.

FIG. 2 is a UML mapping diagram of an adjacency list in an embodiment of the present invention.

FIG. 3 is a diagram of the correspondence between the 3/2 connection and the adjacency list in the embodiment of the present invention.

Fig. 4 is a diagram of the corresponding relationship between the double bus section with bypass connection and the adjacency table in the embodiment of the present invention.

Fig. 5 is a hierarchical structure diagram of power grid topology generation in an embodiment of the present invention.

Fig. 6 is a flowchart of the whole network topology in the embodiment of the present invention.

detailed description

The distribution network fault diagnosis system based on topology knowledge of the present invention, as shown in FIG. 1 , includes a resource layer, a knowledge layer and a diagnosis layer.

The resource layer includes fault information collection devices such as fault recorders, monitoring terminals, power distribution equipment ledgers, topology files, and databases. Switch status information, the fault recorder is used to obtain the waveform after the fault and expressed in the form of COMTRADE, the topology file is used to provide the electrical wiring diagram of the entire power grid, and the power distribution equipment ledger is used to obtain various power distribution equipment operation at the time of failure.

The knowledge layer includes a knowledge base that provides information for the diagnosis layer, and the knowledge base includes a topology knowledge base and a fault knowledge base, which provide basis for the diagnosis layer to function.

The topology knowledge base in this layer is formed as follows: IEC61970 reflects the connection relationship between devices through Terminal (terminal) and ConnectivityNode (connection point), and physically connected devices are connected to a common one in CIM through their own Terminal on the ConnectivityNode. The adjacency list establishes a singly linked list for each vertex in the graph, and the nodes in the singly linked list represent all the edges attached to this vertex. After comparison, it can be obtained that each ConnectivityNode in the CIM corresponds to a singly linked list in the adjacency list. Through the above analysis, the present invention creates two classes on the basis of topology package (Topology): Vertex class and adjNode class, corresponding to the head node class and table node class of the adjacency list respectively. The three fields of the Vertex class are access flags, vertex names, and first node pointers. The Vertex class includes busbars, lines, transformers, and power supplies. The adjNode class is used to reflect the electrical components that are connected to the Vertex class, that is, a Vertex device connection The ConnectivityNode contains n terminals, and the corresponding singly linked list in the adjacency list has n-1 adjNodes. The three domains of each adjNode class are the name of the associated electrical component, the connected circuit breaker, and the next domain. Figure 2 shows the UML block diagram of building an adjacency list according to the topology package (Topology).

In addition, as a part of the distribution network, the low-voltage side bus of the substation will adopt complex wiring methods such as 3/2, double bus section with bypass, etc. In order to solve the situation of direct connection of circuit breakers in 3/2 wiring, the present invention is based on IEC61970 The topology package (Topology) defines the ConnectivityNode class, adding a Y node between two directly connected circuit breakers. The judgment method of the Y node is: if a ConnectivityNode object connects two circuit breakers and one line end point, it is reserved as a Y node, as shown in Figure 3 for an example. However, in the double-bus section with bypass wiring, disconnectors are involved, and the circuit breaker is connected to different busbars through different disconnectors in different operating modes. The present invention regards the disconnectors in the "closed" state as wires. An example of the conversion diagram is shown in Figure 4.

The diagnostic layer is the interface of the whole system, which completes the communication between human and system. The diagnosis layer includes a fault diagnosis server and a post-fault accident analysis program module. The fault diagnosis server is equipped with a fault diagnosis program module for making preliminary judgments on the cause and impact of the fault and providing reference for operators.

The invention also provides a diagnosis method corresponding to the distribution network fault diagnosis system based on topology knowledge, including topology knowledge formation and fault diagnosis.

The formation of the topological knowledge proceeds as follows:

(1) When a fault occurs, the resource layer at the bottom collects various fault information, collects the switch state information after the fault from the section switch, isolating switch and circuit breaker through the monitoring terminal, and obtains the waveform after the fault from the fault recorder , Obtain the operation status of various power distribution equipment at the time of failure from the power distribution equipment ledger, and obtain the electrical wiring diagram of the entire power grid from the topology file.

(2) Through RDFSchemeEditor software, provide RDF data files conforming to CIM-based model rules, or CIM-XML model data files.

(3) Integrate SVG graphic elements to form wiring diagrams of each plant station and distribution network feeder, and provide an interface to the upper layer in the form of CIS services. Among them, SVG is a two-dimensional graphics language described by XML. In the standard IEC61970-553-4 part, it is used as the information carrier of substation graphic elements and the standard for generating and exchanging power graphics.

(4) Query the wiring diagrams of each plant station and distribution network feeder through LINQtoXML, find the names, connection points, and connection components of each electrical component, and store them in the topology library. The topology library exists in the form of DataSet in the memory, which is an offline container of data defined in . NET, does not contain the concept of database connection, and is only a group of data sets in the memory. The topology library here provides different interfaces for viewing plant graphics, topology analysis, and power flow calculation.

The process of (2), (3) and (4) above is shown in Figure 5.

The topology library mentioned in the above step (204) is the core part, and its formation process is:

(401) The application program (line fault analysis, power flow calculation, etc.) starts an empty DataSet according to the DataSet template.

(402) Search for relevant data from the wiring diagrams of each plant and station according to the definition of the DataSet template, and fill the topology library. These functions are provided to the application layer in the form of interfaces in the data filling layer. Take the search for a 10kV busbar as an example, the search steps are:

a. Read the entire power grid graphics file. Obtain the sub control area through cim:SubControlArea.

b. For each sub-control area, obtain each plant station through cim:SubControlArea.Contain_Substations.

c. Use LINQtoXML to find the busbar cim:BusbarSection and its corresponding voltage level for the graphic files corresponding to each plant.

d. Take out the switch connected to the bus through the ConnectivityNode of the bus.

e. Determine whether all sites have been accessed, otherwise go to step b.

(403) Use the same method to search for feeders, distribution transformers, and connected switchgear such as section switches and isolating switches.

The topology library is stored in the DataSet in the form of an adjacency list, and is started with the system in the form of a server.

(5) According to the above adjacency table representation method, search the topology of the whole network to obtain the topology structure of the whole network. The algorithm flow is shown in Figure 6.

The fault diagnosis is carried out as follows:

After obtaining the topological structure of the whole network, combined with the switch state information obtained by the monitoring terminal of the resource layer after a fault, the feeder disconnected due to the fault can be obtained through the fault diagnosis service program, and then through the depth-first search method, all the feeders on these feeders can be obtained. connected distribution transformer. The information of these outage distribution transformers is counted through the ledger information of power distribution equipment, such as the distribution transformer capacity, the instantaneous power flowing through the distribution transformer when a fault occurs, etc.

The power distribution GIS provides an integrated "geographical information + attribute information + relationship information" model. The fault diagnosis program module establishes an interface with the power distribution GIS, and calls the geographical information of the power outage line from the power distribution GIS, that is, the geographical location of the power outage feeder. Then the geographical information is transmitted to the weather forecast system and lightning location system to obtain the weather conditions and lightning conditions in the fault area at the time of the fault, and then use the power distribution equipment ledger to obtain the operating conditions of the power distribution equipment in the fault area at the time of the fault, and finally According to the weather, lightning, and equipment operation, a preliminary judgment on the cause of the fault is given for the reference of the operating personnel.

The above are the preferred embodiments of the present invention, and all changes made according to the technical solution of the present invention, when the functional effect produced does not exceed the scope of the technical solution of the present invention, all belong to the protection scope of the present invention.

Claims (3)

1. A distribution network fault diagnosis system based on topology knowledge, characterized in that, comprising a resource layer, a knowledge layer and a diagnosis layer; The resource layer includes a fault recorder, a monitoring terminal, a power distribution equipment ledger, a topology file and a database. The monitoring terminal is used to collect switch status information after a fault, and the fault recorder is used to obtain waveforms after a fault. The topology file is used to provide an electrical wiring diagram of the entire power grid, and the power distribution equipment ledger is used to obtain the operating conditions of various power distribution equipment when they fail; The knowledge layer includes a knowledge base that provides information for the diagnosis layer, and the knowledge base includes a topology knowledge base and a fault knowledge base; the topology knowledge base is formed as follows: two classes are created on the basis of topology packages in IEC61970: The Vertex class and the adjNode class correspond to the head node class and the table node class of the adjacency list respectively; the three fields of the Vertex class are the access flag, the vertex name, and the first node pointer, and the Vertex class includes busbars, lines, transformers and power supplies , the adjNode class reflects the electrical components that are connected to the Vertex class, that is, the ConnectivityNode connected to a Vertex device contains n terminals, and the corresponding singly linked list in the adjacency list has n-1 adjNodes, and each adjNode class has three The domains are the name of the associated electrical component, the connected circuit breaker, and the next domain; in addition, according to the definition of the ConnectivityNode class in the topology package in IEC61970, a Y node is added between two directly connected circuit breakers; the judgment method of the Y node is : If a ConnectivityNode object connects two circuit breakers and a line end, it will be reserved as a Y node; the disconnector in the "closed" state is regarded as a wire; The diagnosis layer includes a fault diagnosis server, and a fault diagnosis program module is arranged on the fault diagnosis server. 2. the diagnostic method corresponding to the distribution network fault diagnosis system based on topological knowledge according to claim 1, is characterized in that, comprises topological knowledge formation and fault diagnosis; The formation of the topology knowledge is carried out in the following steps: (1) When a fault occurs, the resource layer at the bottom collects various fault information, collects the switch status information after the fault from the monitoring terminal, obtains the waveform after the fault from the fault recorder, and obtains various distribution equipment from the power distribution equipment ledger. The operation of electrical equipment in the event of a fault, and the electrical wiring diagram of the entire grid is obtained from the topology file; (2) Provide RDF data files conforming to CIM-based model rules, or CIM-XML model data files; (3) Integrate SVG graphic elements to form wiring diagrams of each plant station and distribution network feeder, and provide an interface to the upper layer in the form of CIS services; (4) Query the wiring diagrams of each plant station and distribution network feeder through LINQtoXML, find the names, connection points, and connection components of each electrical component, and store them in the topology library; the topology library exists in the form of DataSet in the memory , is an offline container of data defined in .NET, does not contain the concept of database connection, but is just a set of data sets in memory; the topology library is stored in the DataSet in the form of an adjacency list, and is distributed with the system in the form of a server start up; (5) Search the topology of the entire network to obtain the topology of the entire network; The fault diagnosis is carried out as follows: after obtaining the topological structure of the whole network, combined with the switch state information after the fault obtained by the monitoring terminal of the resource layer, the feeders disconnected due to faults and the feeders on these feeders are obtained through the fault diagnosis program module. Connected distribution transformers, the information of these outage distribution transformers is counted through the distribution equipment ledger; the fault diagnosis program module establishes an interface with the power distribution GIS, and calls the geographic information of the power outage line from the power distribution GIS, that is, the power outage feeder The geographic location, and then transmit the geographical information to the weather forecast system and lightning positioning system to obtain the weather conditions and lightning conditions in the fault area at the time of the fault, and then use the power distribution equipment ledger to obtain the power distribution equipment in the fault area at the time of the fault Finally, according to the weather, lightning, and equipment operation, a preliminary judgment on the cause of the failure is given. 3. The distribution network fault diagnosis method based on topology knowledge according to claim 2, characterized in that, in step (4), the formation process of the topology library is: (401) The application starts an empty DataSet according to the DataSet template; (402) Search for relevant data according to the definition of the DataSet template from the wiring diagram of each plant and station, and fill the topology library; this part of the function is provided to the application layer in the form of an interface in the data filling layer; (403) Use the same method to search for feeders, distribution transformers, and connected switchgear.