A transmission line trip root cause analysis system and method

Abstract

The present invention belongs to the technical field of power system operation and maintenance, and specifically relates to a system and method for analyzing the root cause of a transmission line trip. The structure of the system is divided into four parts, including a data layer, a risk layer, a fault layer, and a diagnostic layer; first, multi-source data related to the transmission line trip is collected to form a data layer. Secondly, the mapping relationship between risk and time and space is obtained in the risk layer based on expert knowledge, and the multi-source data in the data layer is fused using the DS evidence theory to obtain the initial value of the risk probability. Then, based on the initial value of the risk probability, the conditional probability of tripping under the condition of risk occurrence is calculated in the fault layer by combining the joint occurrence probability of risk and tripping fault in expert knowledge. Finally, the diagnostic layer uses Bayes' theorem to calculate the conditional probability of various risks under the tripping fault, and completes the root cause analysis of the tripping fault.

Description

Technical Field

[0001] The invention belongs to the technical field of power system operation and maintenance, and specifically relates to a transmission line tripping root cause analysis system and method. Background Art

[0002] In recent years, the scale of power grids has continued to expand, the load has continued to increase, and adverse conditions such as natural disasters, external force damage and equipment failures have led to a high incidence of transmission line tripping accidents. Tripping accidents not only affect the power supply of the loaded equipment, but also have an impact on the entire local power grid system, causing the system voltage to drop sharply and causing damage to some equipment. Therefore, a multi-level analysis of the root causes of transmission line tripping is conducted, and the analysis results are used to guide the dispatch of manpower and material resources, carry out targeted operation and maintenance work in the tripping area, effectively reduce the impact of transmission line tripping and outage, improve the power grid's response capabilities in the event of sudden tripping, and save inspection costs.

[0003] At present, there is no perfect root cause analysis method for transmission line tripping. The existing transmission line tripping operation and maintenance mainly rely on fault distance measurement and large-scale manual tower climbing to find the fault point where the tripping occurs. Then the operation and maintenance personnel judge and analyze the cause of the tripping. There are problems such as long search time, large investment of manpower and material resources, and a large degree of reliance on the work experience and skill level of the operation and maintenance personnel, which has great limitations. In addition, there are also probabilistic analysis technologies for tripping caused by a single risk. For example, the technology of calculating the probability of lightning strike and lightning tripping of distribution towers through electrical geometry models and Monte Carlo methods; the technology of calculating the probability of ice tripping by establishing a multi-dimensional coordinate system of environmental data and using heuristic algorithms; and the technical method of analyzing the correlation between meteorological data and tripping through the FP-growth association rule algorithm. However, these methods are only for a certain type of risk data, and cannot reversely analyze the root cause of the fault through the tripping accident. Summary of the invention

[0004] In order to overcome the shortcomings of the existing power transmission line tripping root cause analysis technology and realize the purpose of quickly digging out the root cause of tripping from a multi-dimensional and multi-level perspective, the present invention provides a power transmission line tripping root cause analysis system and method. The structure of the system is divided into four parts, including a data layer, a risk layer, a fault layer and a diagnosis layer. Among them, the data layer is mainly the acquisition of multi-source data; the risk layer includes building a risk knowledge base, fusing the multi-source data of the data layer, and initializing the risk probability; the fault layer includes building a fault knowledge base, initializing the conditional probability of tripping under the condition of risk occurrence; the diagnosis layer mainly calculates the conditional probability of various risks under the tripping fault and performs the root cause analysis of the tripping fault. First, multi-source data related to the transmission line tripping is collected to form a data layer. Secondly, the mapping relationship between risk and time and space is obtained in the risk layer based on expert knowledge, and the multi-source data in the data layer is fused using the DS evidence theory to obtain the initial value of the risk occurrence probability. Then, on the basis of the initial value of the risk occurrence probability, the conditional probability of tripping under the risk occurrence condition is calculated in the fault layer by combining the joint occurrence probability of risk and tripping fault in the expert knowledge. Finally, the diagnosis layer uses the Bayesian theorem to calculate the conditional probability of various risks under the tripping fault and completes the root cause analysis of the tripping fault.

[0005] To achieve the above object, the present invention adopts the following technical solution:

[0006] A power transmission line tripping root cause analysis system, the system comprising a data layer, a risk layer, a fault layer and a diagnosis layer;

[0007] The data layer is used to provide data support for the root cause analysis of power line tripping, including detailed tripping data of the power system, historical maintenance records, operation and maintenance work order data, tower environment data, meteorological warning data and tower location data;

[0008] The risk layer is used to record the types of risks that occur in the transmission line and the mapping relationship between each type of risk and time, month, and space. The risk layer obtains data from the data layer and calculates the probability of occurrence of each type of risk in the risk layer by fusion of multi-source data. At the same time, the risk layer initializes the initialization probability of each risk based on the mapping relationship between the risk type and time, month, and space;

[0009] The fault layer is used to record the tripping fault information and the probability of each type of risk and the tripping fault occurring simultaneously. At the same time, the fault layer obtains the initialized risk probability data from the risk layer and calculates the conditional probability of tripping under each risk occurrence condition;

[0010] The diagnostic layer is used to determine the risk type and corresponding probability value that cause tripping. The diagnostic layer obtains the conditional probability data of tripping under risk occurrence conditions from the fault layer. According to the conditional probability of tripping under each risk occurrence condition and the initialization probability of the corresponding risk, the conditional probability of each risk under the tripping fault is calculated by Bayesian theorem, and the root cause analysis of the tripping fault is performed based on the obtained probability value.

[0011] A method for analyzing the root cause of a power transmission line tripping, comprising:

[0012] S1. Obtain the detailed tripping data, historical maintenance records, operation and maintenance work order data, tower environment data, meteorological warning data and tower location data of the power system to build a transmission line tripping root cause analysis database;

[0013] S2. Define risk i It is used to indicate the risk of transmission line tripping, which includes lightning strike, wildfire, bird damage, wind deviation, ice cover, geological disaster, pollution flashover, external force damage, and tree and bamboo discharge. Each risk type is set to correspond to an i value;

[0014] S3, constructing the risk knowledge base KB1, including defining the mapping relationship between the i-th risk and the time month in step S2 as ft i , and the mapping relationship between the space is fs i ;

[0015] S4. Based on the database of step S1 and the risk types of step S2, the BPA function of multi-source data is calculated according to the DS evidence theory, and then the BPA of data from different sources is integrated and calculated using the Dempster rule in the DS evidence theory to obtain the probability of occurrence of each type of risk. The probability of occurrence of the i-th type of risk is recorded as Pr i At the same time, according to the mapping relationship between risk, time, month and space in the risk knowledge base KB1, the initialization probability of the i-th risk is obtained:

[0016] P i月份 =ft i (Pr i ,x 月份 )

[0017] P i空间 =fs i (Pr i , x 位置 )

[0018] where x 月份 The month information of the current evaluation time, x 位置 is the spatial location information of the area to be evaluated, and the initial probability P of the i-th risk occurrence is calculated by the geometric mean method. i :

[0019]

[0020] S5. Construct the fault knowledge base KB2, including defining the probability of the i-th risk and the tripping fault occurring at the same time as P (risk i ,trip), combined with the initial probability P of the i-th risk i , calculate the conditional probability of tripping under the condition of the occurrence of the i-th risk:

[0021] P(trip|risk i )=P(risk i ,trip) / P i

[0022] S6. Use Bayesian theorem to calculate the conditional probability of various risks under tripping fault, and get the conditional probability of the i-th risk as:

[0023]

[0024] According to the conditional probability values ​​of various risks under tripping faults, the larger the probability value, the greater the impact of this type of risk on the tripping fault, thereby realizing the root cause analysis of the tripping fault.

[0025] The beneficial effects of the present invention are:

[0026] The present invention proposes a transmission line tripping root cause analysis system and method, aiming to quickly dig out the root cause of the tripping from a multi-dimensional and multi-level perspective, apply the analysis results to the transmission line operation and maintenance work, guide the scheduling of manpower and material resources, improve the response capability of the power grid in the event of a sudden tripping accident, and save inspection costs.

[0027] Compared with the prior art, the present invention is a complete root cause analysis system for power transmission line tripping. The structure of the system is divided into four parts, including a data layer, a risk layer, a fault layer and a diagnosis layer. Among them, the data layer mainly acquires multi-source data; the risk layer includes building a risk knowledge base, fusing multi-source data in the data layer, and initializing the risk probability; the fault layer includes building a fault knowledge base, initializing the conditional probability of tripping under the condition of risk occurrence; the diagnosis layer mainly calculates the conditional probability of various risks under the tripping fault and performs the root cause analysis of the tripping fault. First, multi-source data related to the tripping of the transmission line is collected to form a data layer. Secondly, the mapping relationship between risk and time and space is obtained in the risk layer based on expert knowledge, and the multi-source data in the data layer is fused using the DS evidence theory to obtain the initial value of the risk occurrence probability. Then, based on the initial value of the risk occurrence probability, the joint occurrence probability of risk and tripping fault in the expert knowledge is combined in the fault layer to calculate the conditional probability of tripping under the risk occurrence condition. Finally, the diagnosis layer uses Bayes' theorem to calculate the conditional probability of various risks under the tripping fault and completes the root cause analysis of the tripping fault. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 It is a logical structure and flow chart of the present invention. DETAILED DESCRIPTION

[0029] The present invention is further described in detail below with reference to the accompanying drawings.

[0030] like Figure 1 As shown, the present invention provides a transmission line trip root cause analysis system and method, the system includes a data layer, a risk layer, a fault layer and a diagnosis layer, as follows:

[0031] The data layer is used to provide data support for the root cause analysis of power line tripping, including detailed tripping data of the power system, historical maintenance records, operation and maintenance work order data, tower environment data, meteorological warning data and tower location data;

[0032] The risk layer is used to record the types of risks that occur in the transmission line and the mapping relationship between each type of risk and time, month, and space. The risk layer obtains data from the data layer and calculates the probability of occurrence of each type of risk in the risk layer by fusion of multi-source data. At the same time, the risk layer initializes the initialization probability of each risk based on the mapping relationship between the risk type and time, month, and space;

[0033] The fault layer is used to record the tripping fault information and the probability of each type of risk and the tripping fault occurring simultaneously. At the same time, the fault layer obtains the initialized risk probability data from the risk layer and calculates the conditional probability of tripping under each risk occurrence condition;

[0034] The diagnostic layer is used to determine the risk type and corresponding probability value that cause tripping. The diagnostic layer obtains the conditional probability data of tripping under risk occurrence conditions from the fault layer. According to the conditional probability of tripping under each risk occurrence condition and the initialization probability of the corresponding risk, the conditional probability of each risk under the tripping fault is calculated by Bayesian theorem, and the root cause analysis of the tripping fault is performed based on the obtained probability value.

[0035] A method for analyzing the root cause of a power transmission line tripping, specifically comprising:

[0036] S1. Obtain the detailed tripping data, historical maintenance records, operation and maintenance work order data, tower environment data, meteorological warning data and tower location data of the power system to build a transmission line tripping root cause analysis database;

[0037] S2. Define risk i It is used to indicate the risk of transmission line tripping, which includes lightning strike, wildfire, bird damage, wind deviation, ice cover, geological disaster, pollution flashover, external force damage, and tree and bamboo discharge. Each risk type is set to correspond to an i value;

[0038] S3, constructing the risk knowledge base KB1, including defining the mapping relationship between the i-th risk and the time month in step S2 as ft i , and the mapping relationship between the space is fs i ;

[0039] S4. Based on the database of step S1 and the risk types of step S2, the BPA function of multi-source data is calculated according to the DS evidence theory, and then the BPA of data from different sources is integrated and calculated using the Dempster rule in the DS evidence theory to obtain the probability of occurrence of each type of risk. The probability of occurrence of the i-th type of risk is recorded as Pr i At the same time, according to the mapping relationship between risk, time, month and space, the initialization probability of the i-th risk is obtained:

[0040] P i月份 =ft i (Pr i , x 月份 )

[0041] P i空间 =fs i (Pr i ,x 位置 )

[0042] where x 月份 The month information of the current evaluation time, x 位置 is the spatial location information of the area to be evaluated, and the initial probability P of the i-th risk occurrence is calculated by the geometric mean method. i :

[0043]

[0044] S5. Construct the fault knowledge base KB2, including defining the probability of the i-th risk and the tripping fault occurring at the same time as P (risk i ,trip), combined with the initial probability P of the i-th risk i , calculate the conditional probability of tripping under the condition of the occurrence of the i-th risk:

[0045] P(trip|risk i )=P(risk i ,trip) / P i

[0046] S6. Use Bayesian theorem to calculate the conditional probability of various risks under tripping fault, and get the conditional probability of the i-th risk as:

[0047]

[0048] According to the conditional probability values ​​of various risks under tripping faults, the larger the probability value, the greater the impact of this type of risk on the tripping fault, thereby realizing the root cause analysis of the tripping fault.

[0049] The existing root cause analysis technology for power transmission line tripping mainly relies on fault distance measurement and large-scale manual tower climbing to find the fault point where the tripping occurs, and then the operation and maintenance personnel judge and analyze the cause of the tripping. The present invention designs a complete root cause analysis system for tripping, which specifically includes a data layer, a risk layer, a fault layer and a diagnosis layer. The specific implementation method includes building a tripping root cause analysis database, integrating expert knowledge, using DS evidence theory and Bayesian theorem, and establishing a data-driven analysis system, which reduces the time for finding the root cause of tripping and saves inspection costs.

[0050] As for the probability analysis technology of tripping caused by a single risk, the advantages of the present invention are that it collects a wealth of tripping-related data at the data layer, incorporates multiple risk types that may cause tripping at the risk layer, and can simultaneously analyze the impact of multiple risks on tripping failures. At the same time, the diagnosis layer reversely infers the root cause of the tripping and reversely analyzes the probability of various risks that cause the tripping, while the existing technology lacks consideration of this part.

Claims

1. A method for analyzing the root cause of a power transmission line trip, characterized in that: include: S1. Obtain the detailed tripping data, historical maintenance records, operation and maintenance work order data, tower environment data, meteorological warning data and tower location data of the power system to build a transmission line tripping root cause analysis database; S2. Define risk i It is used to indicate the risk of transmission line tripping, which includes lightning strike, wildfire, bird damage, wind deviation, ice cover, geological disaster, pollution flashover, external force damage, and tree and bamboo discharge. Each risk type is set to correspond to an i value; S3, constructing the risk knowledge base KB1, including defining the mapping relationship between the i-th risk and the time month in step S2 as ft i , and the mapping relationship between the space is fs i ; S4. Based on the database of step S1 and the risk types of step S2, the BPA function of multi-source data is calculated according to the DS evidence theory, and then the BPA of data from different sources is integrated and calculated using the Dempster rule in the DS evidence theory to obtain the probability of occurrence of each type of risk. The probability of occurrence of the i-th type of risk is recorded as Pr i At the same time, according to the mapping relationship between risk, time, month and space, the initialization probability of the i-th risk is obtained: P i月份 =ft i (Pr i ,x 月份 ) P i空间 =fs i (Pr i ,x 位置 , where x 月份 The month information of the current evaluation time, x 位置 is the spatial location information of the area to be evaluated, and the initial probability P of the i-th risk occurrence is calculated by the geometric mean method. i : S5. Construct the fault knowledge base KB2, including defining the probability of the i-th risk and the tripping fault occurring at the same time as P (risk i ,trip), combined with the initial probability P of the i-th risk i , calculate the conditional probability of tripping under the condition of the occurrence of the i-th risk: P(trip|risk i )=P(risk i ,trip) / P i S6. Use Bayesian theorem to calculate the conditional probability of various risks under tripping fault, and get the conditional probability of the i-th risk as: According to the conditional probability values ​​of various risks under tripping faults, the larger the probability value, the greater the impact of this type of risk on the tripping fault, thereby realizing the root cause analysis of the tripping fault.

Images