Thermal power plant steam pipeline system weak link positioning method based on data fusion

Abstract

The invention relates to the technical field of thermal power plant equipment state evaluation, and discloses a thermal power plant steam pipeline system weak link positioning method based on data fusion, comprising the following steps: step 1, analyzing multi-source data of a thermal power plant steam pipeline system, and combining a logic mapping relation between the multi-source data and steam pipeline multi-dimensional characteristics to obtain a positioning result; constructing a running state knowledge graph of the thermal power plant steam pipeline system based on multi-source data association analysis; 2, establishing a steam pipeline state characteristic index system; step 3, screening out key characteristic indexes from the steam pipeline state characteristic index system; and 4, fusing the key characteristic indexes by adopting a fuzzy entropy weight method, calculating the health degree of each section of steam pipeline, and sorting, thereby realizing the positioning of the weak link of the steam pipeline system through a health degree sorting result. According to the method, the health state of the steam pipeline can be accurately and efficiently determined, and efficient and reasonable identification of the weak link of the steam pipeline system is realized.

Description

Technical Field

[0001] The invention relates to the technical field of equipment status assessment in thermal power plants, and in particular to a method for locating weak links in a steam pipeline system of a thermal power plant based on data fusion. Background Art

[0002] With the large-scale grid connection of new energy sources with unstable output characteristics such as wind power and photovoltaic power, the peak-to-valley difference of power load continues to expand, and the power system faces unprecedented peak-shaving pressure. As the main basic regulating power source, the peak-shaving role of thermal power plants in the power system has become increasingly prominent, and their reliable operation level has therefore received increasing attention.

[0003] The steam pipeline system is an important part of a thermal power plant. It plays an important role in transporting and distributing high-temperature and high-pressure steam to drive the steam turbine to generate electricity and provide a hot steam source for other auxiliary systems. However, with the increase in the operating time of thermal power plants, problems in the early steam pipeline design, manufacturing, installation and other links have gradually been exposed. Especially for aging pipelines with a long operating time, the long-term high temperature, high pressure and complex stress working environment make their safety issues more prominent. Once leaks or bursts occur, they will seriously threaten the safe and reliable operation of thermal power plants. Therefore, it is of great significance to reasonably evaluate the health level of the steam pipelines of thermal power plants and guide the efficient identification of weak links in the steam pipeline system to ensure the reliable operation of thermal power plants. At the same time, this can also provide a more scientific decision-making basis for determining the maintenance / update priority of steam pipelines at the work site, thereby ensuring the reasonable allocation of maintenance resources.

[0004] However, the existing health level assessment scheme for steam pipes in thermal power plants still has the following defects: (1) In terms of steam pipe data fusion, the current method has not yet conducted in-depth correlation analysis on multi-source data of steam pipes, and lacks a knowledge graph of the operating status of the steam pipe system in thermal power plants, resulting in the failure to fully utilize the potential of multi-source data, thereby limiting the ability to model and evaluate the operating status of the steam pipe system; (2) In terms of steam pipe health status assessment, the existing model has failed to establish a comprehensive status feature indicator extraction method based on data mining technology, and cannot realize quantitative analysis of multi-dimensional features from multiple angles; (3) In terms of locating weak links in the steam pipe system, the current model lacks a complete module design and cannot be embedded and directly applied to the existing thermal power plant steam pipe system monitoring platform. Summary of the invention

[0005] The present invention aims to provide a method for locating weak links in a steam pipeline system of a thermal power plant based on data fusion, which can accurately and efficiently determine the health status of the steam pipeline and realize efficient and reasonable identification of weak links in the steam pipeline system.

[0006] The basic scheme provided by the present invention is: a method for locating weak links in a steam pipeline system of a thermal power plant based on data fusion, comprising the following steps:

[0007] Step 1: Analyze the multi-source data of the steam pipeline system of the thermal power plant, and build an operating status knowledge graph of the steam pipeline system of the thermal power plant based on the association analysis of multi-source data by combining the logical mapping relationship between the multi-source data and the multi-dimensional features of the steam pipeline; the data sources of the multi-source data include the DCS system of the thermal power plant, the inspection records of the on-site personnel of the thermal power plant, the equipment ledger of the steam pipeline, and the factory information of the steam pipeline; the multi-dimensional features of the steam pipeline include operating status features, structural features, reliability features, stability features, and thermal performance features;

[0008] Step 2: Based on the multi-source data collected in step 1, a steam pipeline state characteristic index system is established in combination with the constructed operation state knowledge graph; the steam pipeline state characteristic index system includes steam pipeline failure rate, operation condition, stability index, and structural importance;

[0009] Step 3, using the state characteristic index selection method based on VT-PCA, from the steam pipeline state characteristic index system, select the key characteristic index that can effectively characterize the differential operation state of the steam pipeline;

[0010] Step 4: Use the fuzzy entropy weight method to fuse the key characteristic indicators, and based on the fusion results, calculate the health of each section of the steam pipeline and sort the health, so as to locate the weak links of the steam pipeline system through the health sorting results.

[0011] The working principle and advantages of the present invention are:

[0012] The method for locating weak links in the steam pipeline system of a thermal power plant based on data fusion of the present invention can accurately and efficiently determine the health status of the steam pipeline, and realize the operating status monitoring of the steam pipeline system and the efficient and reasonable identification of weak links, which is of great significance for ensuring the safe and stable operation of the steam pipeline system of the power plant, improving power generation efficiency and increasing economic benefits. The key points are: First, based on multi-source data related to the steam pipeline, this scheme constructs an operating status knowledge graph of the steam pipeline system of a thermal power plant based on multi-source data association analysis, which can make full use of multi-source data, based on richer information dimensions, accurately identify the operating status of the steam pipeline, realize accurate and efficient status monitoring of the steam pipeline system, and provide reliable data input for modeling and evaluation of the operating status of the steam pipeline system. Second, this scheme further constructs a steam pipeline status characteristic indicator system, and can realize quantitative analysis of multi-dimensional features from multiple angles, providing a solution for the quantitative description of the steam pipeline status.

[0013] Third, this scheme proposes a state characteristic index selection method based on VT-PCA (variance threshold method-principal component analysis method), which can remove redundant features, simplify data interpretation, extract key characteristic indicators, and help improve overall calculation efficiency. Fourth, this method uses intuitive health quantitative indications for weak links to accurately locate weak links in the steam pipeline system, which can provide a more scientific decision-making basis for determining the maintenance / update priority of steam pipelines at the work site, thereby ensuring the rational allocation of maintenance resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 It is a schematic diagram of the knowledge graph structure of the operation status of the steam pipeline system of a thermal power plant according to an embodiment of the method for locating weak links of the steam pipeline system of a thermal power plant based on data fusion of the present invention;

[0015] Figure 2 It is a schematic diagram of the method flow structure of an embodiment of a method for locating weak links in a steam pipeline system of a thermal power plant based on data fusion according to the present invention;

[0016] Figure 3 It is a schematic diagram of the input-output relationship between the modules constructed in the method of the embodiment of the method for locating weak links in the steam pipeline system of a thermal power plant based on data fusion of the present invention. DETAILED DESCRIPTION

[0017] The following is a further detailed description through specific implementation methods:

[0018] The embodiment is basically as shown in the attached Figure 2 As shown: A method for locating weak links in a steam pipeline system of a thermal power plant based on data fusion includes the following steps:

[0019] Step 1: Analyze the multi-source data of the steam pipeline system of the thermal power plant, combine the logical mapping relationship between the multi-source data and the multi-dimensional characteristics of the steam pipeline, and construct the operating status knowledge graph of the steam pipeline system of the thermal power plant based on the association analysis of the multi-source data, such as Figure 1 shown.

[0020] The data sources of the multi-source data include the DCS system (distributed control system) of the thermal power plant, the on-site inspection records of the thermal power plant, the steam pipeline equipment ledger and the steam pipeline factory information. The multi-dimensional characteristics of the steam pipeline include operating status characteristics, structural characteristics, reliability characteristics, stability characteristics and thermal performance characteristics.

[0021] The multi-source data includes one type of data, two types of data, three types of data, four types of data and five types of data.

[0022] The first type of data is steam pipeline temperature data and steam pipeline pressure data, that is, based on the pressure monitoring data and temperature monitoring data collected from the temperature sensors and pressure sensors in the steam pipeline system of the thermal power plant, the historical time series of pipeline temperature values ​​and the historical time series of steam pressure values ​​of each section of the steam pipeline are extracted from the DCS system of the thermal power plant; the first type of data is associated with the operating status characteristics of the steam pipeline (corresponding to the operating condition characteristics and operating condition indicators of the steam pipeline).

[0023] Specifically, taking the nth section of steam pipeline as an example, the temperature value time series T extracted by the DCS system is n With pressure value P n The time series can be specifically expressed as:

[0024]

[0025] Where: It represents the average temperature of the nth steam pipeline on the lth day in the historical time series of pipeline temperature values; It represents the average temperature of the nth section of steam pipeline at the tth hour on the lth day in the historical time series of pipeline temperature values; It represents the average pressure of the nth steam pipeline on the lth day in the historical time series of steam pressure values; It represents the average pressure of the nth section of steam pipeline at the tth hour on the lth day in the historical time series of steam pressure values.

[0026] Here, since the pipeline temperature value and steam pressure value mainly reflect the thermodynamic state of the steam pipeline under variable load and working conditions, this type of data can be associated with the steam pipeline operating status characteristics, and the steam pipeline operating condition characteristic quantities can be mined therefrom.

[0027] The second type of data is the connection relationship between the inlet and outlet connection positions of each section of the steam pipeline determined according to the topological diagram of each steam pipeline system recorded in the DCS system of the thermal power plant; the second type of data is associated with the structural characteristics of the steam pipeline (corresponding to the structural importance characteristics and the pipeline vulnerability index).

[0028] Specifically, for a steam pipe system with b connections and m pipe sections, in this embodiment, a two-dimensional correlation matrix I with a size of m×b is used to describe the connection relationship between each steam pipe section and the inlet and outlet positions. The specific rules are as follows:

[0029]

[0030] Where: I(n,k) represents the value of the element in the nth row and kth column of the two-dimensional incidence matrix I.

[0031] Through the above method, any section of steam pipe can be associated with the location of its inlet and outlet, so as to describe and record the topological structure of the steam pipe system in mathematical form. This association matrix form can clearly show the connection relationship between pipe sections and their inlet and outlet locations. This type of data can be associated with the structural characteristics of steam pipes, and the importance characteristics of steam pipe structures can be mined from them.

[0032] The three types of data are state duration matrices; the state duration matrix is ​​constructed according to the following steps: the wear and failure process of the steam pipeline section is divided into four states: normal integrity, crack initiation, crack propagation and burst leakage; according to the inspection records of on-site personnel and the commissioning time of the steam pipeline system, the historical duration of each section of the steam pipeline in the above four states is determined, and then the corresponding state duration matrix is ​​formed; the three types of data are associated with the reliability characteristics of the steam pipeline (corresponding to the pipeline failure rate index).

[0033] Specifically, taking the nth section of the steam pipeline as an example, the specific expression is as follows:

[0034] D n =[D n,1 ,D n,2 ,D n,3 ,D n,4 ]n=1,2,……,m;

[0035] Where: D n represents the state duration matrix of the nth section of steam pipeline obtained from historical patrol records; D n,1 ,D n,2 ,D n,3 ,D n,4 They respectively represent the duration of the nth section of steam pipeline being in the four states of normal integrity, crack initiation, crack propagation and pipe burst leakage.

[0036] Here, since the state duration data mainly reflects the wear of the steam pipe under long-term multiple stresses, and long-term wear will cause the steam pipe to fail, this type of data can be associated with the reliability characteristics of the steam pipe, and the reliability characteristics of the steam pipe can be mined from it.

[0037] The four types of data are displacement indicators; the displacement indicators are obtained according to the following steps: analyzing the manual inspection data of the steam pipeline system, analyzing the displacement of key parts in the steam pipeline, and qualitatively describing the displacement of each section of the steam pipeline, and converting the qualitative description into a numerical quantitative indicator, namely, the displacement indicator; the four types of data are associated with the stability characteristics of the steam pipeline (corresponding to the pipeline displacement indicator).

[0038] Specifically, in this embodiment, the qualitative description may adopt multiple qualitative descriptions such as "normal", "slight deviation", "large deviation" in the manual inspection record. And the qualitative description is converted into a numerical quantitative indicator based on the following formula:

[0039]

[0040] Where: L n It is a numerical quantitative index used to quantitatively describe the displacement of the nth section of the steam pipeline; A kind A is a qualitative description category. If it includes three descriptions: "normal", "slight deviation", and "large deviation", then A kind The value is 3; if the inspection result of the nth section of the steam pipeline is "normal", then a n,kind Taking the value as 1, the displacement of each section of the steam pipe can be described in turn.

[0041] Here, since the displacement data reflects the adaptability of the steam pipeline to external disturbances (such as thermal expansion, vibration, etc.) during operation, this type of data can be associated with the stability characteristics of the steam pipeline, and the steam pipeline stability characteristics can be mined therefrom.

[0042] The five types of data are thermal performance parameters; the thermal performance parameters include the forced convection heat transfer thermal resistance inside the steam pipe, the thermal conductivity thermal resistance parameters of the insulation layer, the convection heat transfer thermal resistance parameters between the outer wall of the pipe and the environment, and the historical average ambient temperature parameters of different steam pipe locations; the five types of data are associated with the thermal performance characteristics of the steam pipe (corresponding to the heat dissipation capacity index).

[0043] Specifically, taking the nth section of the steam pipeline as an example, the thermal performance parameters are specifically expressed as follows:

[0044] Q n =[Q n,1 ,Q n,2 ,Q n,3 ,Q n,4 ]n=1,2,……,m;

[0045] Where: Q n represents the performance parameters and ambient temperature matrix of the nth section of steam pipeline obtained from the steam pipeline equipment inventory system and the manufacturer's factory data; Q n,1 ,Q n,2 ,Q n,3 They represent the thermal resistance of forced convection heat transfer in the nth section of the steam pipeline, the thermal conductivity resistance of the insulation layer, and the convection heat transfer resistance between the outer wall of the pipeline and the environment. n,4 Represents the average ambient temperature at the nth section of the steam pipe.

[0046] Here, since parameters such as pipeline thermal resistance will affect the heat dissipation capacity of the steam pipeline, this type of data can be associated with the thermal characteristics of the steam pipeline, and the thermal performance characteristic quantities related to the heat dissipation capacity of the steam pipeline can be mined therefrom.

[0047] This step also includes: building a multi-source data integration module in the data station of the DCS system of the thermal power plant; the multi-source data integration module is used to execute step 1 and store and manage the multi-source data.

[0048] Step 2: Based on the multi-source data collected in step 1, a steam pipeline status characteristic index system is established in combination with the constructed operating status knowledge graph; the steam pipeline status characteristic index system includes steam pipeline failure rate, operating conditions, stability index, and structural importance.

[0049] The characteristic indicators in the steam pipeline state characteristic indicator system include:

[0050] CV n =[AT n ,VT n ,CT n ,AP n ,VP n ,CP n ,BS n ,λ n (t n ),LN n ,QT n ];

[0051] In the formula, CV n is the characteristic index vector corresponding to the nth section of the steam pipeline.

[0052] AT n VT n , CT n They are the temperature average characteristic index, temperature variance characteristic index and temperature trend characteristic index of the nth section steam pipeline, which can respectively reflect the overall level of the temperature time series of the nth section steam pipeline, the fluctuation amplitude of the temperature time series of the nth section steam pipeline and the trend of the temperature time series of the nth section steam pipeline changing over time.

[0053] Specifically, AT n VT n , CT n Calculate using the following formula:

[0054]

[0055] AP n VP n , CP nThey are respectively the pressure average characteristic index, pressure variance characteristic index and pressure trend characteristic index of the nth section of the steam pipeline.

[0056] Specifically, AP n VP n , CP n Calculate using the following formula:

[0057]

[0058] BS n It is the structural importance index of the nth steam pipeline after normalization and weighting. BS n The larger the value, the more important the pipeline is.

[0059] Specifically, B.S. n Calculate using the following formula:

[0060]

[0061] Where, IN(n) is the importance weight of the subsystem where the nth steam pipeline is located compared with other subsystems; Node(n) is the total number of nodes in the subsystem where the nth steam pipeline is located.

[0062] B n is the betweenness index of the nth steam pipeline; through the betweenness index, the structural importance of each section of the steam pipeline in the normal operation of the system and in the presence of faults can be comprehensively considered. When calculating the betweenness index, based on the established two-dimensional correlation matrix, the betweenness index of each section of the steam pipeline in each steam system is calculated to determine the structural importance of each section of the steam pipeline. Specifically, it can be calculated by the following formula:

[0063]

[0064] In the formula, TB uv represents all directed connected shortest paths from the steam pipe system node (inlet / outlet) u to the steam pipe system node (inlet / outlet) v; NB uv (n) represents the number of paths passing through the nth section of the steam pipe from the shortest path.

[0065] λ n (t n ) represents the nth steam pipeline at its current service time t n The failure rate at time.

[0066] Specifically, when calculating λ n (t n ), including the following sub-steps:

[0067] Based on the state duration data of each section of the steam pipeline in the three types of data (state duration matrix) and the multi-state Markov model, the failure rate of each section of the steam pipeline is calculated as a reliability characteristic indicator.

[0068] Steam pipes are usually repairable electrical components. Therefore, there can be a transition relationship between the four states of normal intact S1, crack initiation S2, crack extension S3 and burst leakage S4. Specifically, the steam pipe will gradually transfer from the normal intact state to the burst leakage state as the operating time increases. In this process, the steam pipe can be returned from the crack extension state to the normal intact state or crack initiation state through overhaul / minor repair operations, and the steam pipe can also be returned from the crack initiation state to the normal intact state. Therefore, the four-state Markov state transfer matrix of the steam pipe can be established first based on the duration of each section of the steam pipe state, as follows:

[0069]

[0070] In the matrix STs, σ i,j (i=1,…,4,j=1,…,4) represents the transfer rate of the steam pipeline from state i to state j, σ i,j 0 means there is no transition relationship between the two states, σ i,j The data in the duration matrix of each section of the steam pipeline state can be used to calculate the following formula:

[0071]

[0072] Where Time n,i,j represents the number of times the steam pipe n is transferred from state i to state j during its operation; It represents the total number of times each section of the steam pipe system transfers from state i to state j during its operation.

[0073] It represents the total time that each section of the steam pipe system is in state i during its operation.

[0074] Based on the four-state Markov state transfer matrix of the steam pipeline established in the above steps, the Runge-Kutta method is used to solve the value of its failure rate with operating time when the pipeline is in different operating states. Taking the nth section of the steam pipeline as an example, assuming that it is in the crack initiation state S2, its failure rate can be expressed as:

[0075]

[0076] In the formula, λ n (t n ) represents the nth steam pipeline at its current service time t nThe failure rate when and It can be obtained by solving the characteristic roots of the Markov state transfer matrix. The solution method adopts the existing technology and will not be explained in detail here.

[0077] Based on the established four-state Markov state transfer matrix of the steam pipeline, the failure rate value of each section of the steam pipeline can be calculated in turn and used as a reliability characteristic index. The larger the index is, the lower the reliability level of the steam pipeline is.

[0078] LN n It is the stability characteristic index of the nth section of steam pipeline after normalization; LN n The larger the value, the more significant the displacement of the steam pipeline section and the worse the pipeline stability.

[0079] Specifically, LN n Calculate using the following formula:

[0080]

[0081] Wherein, L is the numerical data vector of displacement of the m-segment steam pipe in the steam pipe system, i.e., the displacement index.

[0082] QT n It is the average heat flux density of the nth section of the steam pipeline, which can be used to measure the heat dissipation capacity of the steam pipeline. The larger the value, the more significant the heat dissipation of the steam pipeline.

[0083] Specifically, QT n Calculate using the following formula:

[0084]

[0085] This step also includes: building a characteristic index extraction module in the data station of the DCS system of the thermal power plant; the characteristic index extraction module is used to execute step 2.

[0086] Step 3, using the state characteristic index selection method based on VT-PCA (variance threshold method-principal component analysis method), selects key characteristic indicators that can effectively characterize the differential operating state of the steam pipeline from the steam pipeline state characteristic index system, including the following sub-steps:

[0087] Step 3.1, sort out the characteristic index vectors of each section of the steam pipeline, form a separate data vector for each feature and store it.

[0088] Based on the characteristic index of heat dissipation capacity of steam pipeline (corresponding to QT n ) as an example, the corresponding data vector can be expressed as:

[0089] QT=[QT1,QT2,……,QT m ].

[0090] Step 3.2, export the 10 data vectors formed in Step 3.1, and perform variance analysis one by one.

[0091] Based on the characteristic index of heat dissipation capacity of steam pipeline (corresponding to QT n ) as an example, the statistical data variance of the heat dissipation capacity calculated based on the data vector QT is:

[0092]

[0093] Where: VAR QT It is the variance value corresponding to the characteristic index of the heat dissipation capacity of the steam pipeline. The smaller the value, the smaller the data change, and the less likely it is to reflect the difference characteristics between different steam pipeline sections.

[0094] Step 3.3: Predetermine the variance threshold based on expert experience, and screen out feature indicators with variance values ​​less than the variance threshold, thereby achieving an initial screening of feature indicators.

[0095] Step 3.4: Use the principal component analysis method to determine the weight factors of the remaining characteristic indicators after the initial screening, and achieve secondary screening of the characteristic indicators by retaining the steam pipeline operation status characteristic indicators with high weight factors.

[0096] First, the influence of the magnitude differences of indicators in each category on the calculation of weight factors is eliminated through standardization.

[0097] Based on the characteristic index of heat dissipation capacity of steam pipeline (corresponding to QT n ) as an example, the heat dissipation capacity index QT of the nth section steam pipeline n The standardized calculation formula is as follows:

[0098]

[0099] Where: QT std,n It is the heat dissipation capacity index value of the nth section of steam pipeline after standardized processing.

[0100] It should be noted that for the steam pipeline stability index (corresponding to LN n ), since it has been standardized during the calculation process, this step can be skipped.

[0101] Step 3.5: Based on the Person similarity, establish the correlation coefficient matrix between the characteristic indicators of each steam pipeline. Assuming that there are K characteristic indicators after the initial screening, the correlation coefficient matrix CR can be expressed as:

[0102]

[0103] In the formula, cr ij Represents the correlation coefficient matrix between the i-th feature index and the j-th feature index.

[0104] Step 3.6: Calculate the K eigenvalues ​​of the correlation coefficient matrix CR and the corresponding eigenvector in Where j = 1, 2,…, K.

[0105] And determine the p eigenvalues ​​that can make the cumulative contribution rate reach 80%. The p value can be calculated by the following formula:

[0106]

[0107] Step 3.7: Calculate the weight factor of the jth feature index by combining the first p feature vectors WF j , the calculation method is as follows:

[0108]

[0109] Step 3.8: Determine the number of key characteristic indicators based on the experience of on-site experts, and select the characteristic indicators with the largest weights as the key characteristic indicators of the steam pipeline operation status.

[0110] This step also includes: building a characteristic index selection module in the data station of the DCS system of the thermal power plant; the characteristic index selection module is used to execute step 3. Moreover, the characteristic index selection module takes the output of the characteristic index extraction module as input.

[0111] Step 4: Use the fuzzy entropy weight method to fuse the key characteristic indicators, and based on the fusion results, calculate the health of each section of the steam pipe and sort the health. The steam pipes with low health rankings are regarded as the weak links in the located steam pipe system; thus, the weak links in the steam pipe system are located through the health ranking results.

[0112] Specifically, when the fuzzy entropy weight method is used to fuse key feature indicators, the following steps are included:

[0113] Create a fuzzy membership matrix W with m rows and h columns; the element W in W ng (n=1,2,…,m;g=1,2,…,h;) takes the key characteristic index value after standardization of each section of steam pipeline;

[0114] Calculate the information entropy IE of each key feature indicator g :

[0115] In the formula, IE g Represents the information entropy corresponding to the g-th key feature indicator;

[0116] Based on the information entropy of the indicators and the preference weights of the operators of the thermal power plant, the proportional weight of each key characteristic indicator in the steam pipeline health evaluation is calculated as the fusion result;

[0117]

[0118] Where η g represents the fuzzy entropy weight of the g-th key feature index; g Represents the preference weight of the system operator for the g-th key characteristic indicator.

[0119] When calculating the health of each section of the steam pipeline, the health of each section of the steam pipeline is evaluated according to the proportional weight of each key characteristic indicator:

[0120]

[0121] In the formula, HI n It indicates the health evaluation result of the nth section of steam pipeline. The smaller the value is, the lower the operating status and health level of the steam pipeline is.

[0122] This step also includes: in the data station of the DCS system of the thermal power plant, building a result output module; the result output module is used to execute step 4; Figure 3 shown.

[0123] The present embodiment provides a method for locating weak links in the steam pipeline system of a thermal power plant based on data fusion, which can accurately and efficiently determine the health status of the steam pipeline and realize efficient and reasonable identification of weak links in the steam pipeline system, which is of great significance for ensuring the safe and stable operation of the steam pipeline system of the power plant, improving power generation efficiency and increasing economic benefits. In addition, this solution also forms a complete modular system framework of "multi-source data integration module-feature index extraction module-feature index selection module-result output module", which can be embedded and directly applied to the existing thermal power plant steam pipeline system monitoring platform, and can realize real-time update of weak steam pipelines through dynamic update of the multi-source data integration module, more effectively guide on-site steam pipeline system maintenance work, improve work efficiency, and have high practical application value.

[0124] The above is only an embodiment of the present invention. The common sense such as the known specific structure and characteristics in the scheme is not described in detail here. The ordinary technicians in the relevant field are aware of all the common technical knowledge in the technical field of the invention before the application date or priority date, can obtain all the existing technologies in the field, and have the ability to apply the conventional experimental means before that date. The ordinary technicians in the relevant field can improve and implement the scheme in combination with their own abilities under the enlightenment given by this application. Some typical known structures or known methods should not become obstacles for the ordinary technicians in the relevant field to implement this application. It should be pointed out that for the technicians in this field, without departing from the structure of the present invention, several deformations and improvements can be made, which should also be regarded as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the practicality of the patent.

Claims

1. A method for locating weak links in a steam pipeline system of a thermal power plant based on data fusion, characterized in that: The following steps are involved: Step 1: Analyze the multi-source data of the steam pipeline system of the thermal power plant, and build an operating status knowledge graph of the steam pipeline system of the thermal power plant based on the association analysis of multi-source data by combining the logical mapping relationship between the multi-source data and the multi-dimensional features of the steam pipeline; the data sources of the multi-source data include the DCS system of the thermal power plant, the inspection records of the on-site personnel of the thermal power plant, the equipment ledger of the steam pipeline, and the factory information of the steam pipeline; the multi-dimensional features of the steam pipeline include operating status features, structural features, reliability features, stability features, and thermal performance features; Step 2: Based on the multi-source data collected in step 1, a steam pipeline state characteristic index system is established in combination with the constructed operation state knowledge graph; the steam pipeline state characteristic index system includes steam pipeline failure rate, operation condition, stability index, and structural importance; Step 3, using the state characteristic index selection method based on VT-PCA, from the steam pipeline state characteristic index system, select the key characteristic index that can effectively characterize the differential operation state of the steam pipeline; Step 4: Use the fuzzy entropy weight method to fuse the key characteristic indicators, and based on the fusion results, calculate the health of each section of the steam pipeline and sort the health, so as to locate the weak links of the steam pipeline system through the health sorting results.

2. The method for locating weak links in a steam pipeline system of a thermal power plant based on data fusion according to claim 1 is characterized in that: The multi-source data includes a type of data; the type of data is the historical time series of pipe temperature values ​​and steam pressure values ​​of each section of the steam pipe extracted from the DCS system of the thermal power plant based on the pressure monitoring data and temperature monitoring data collected from the temperature sensors and pressure sensors in the steam pipe system of the thermal power plant; the type of data is associated with the operating status characteristics of the steam pipe.

3. The method for locating weak links in a steam pipeline system of a thermal power plant based on data fusion according to claim 1 is characterized in that: The multi-source data includes two types of data; the second type of data is the connection relationship between the inlet and outlet connection positions of each section of the steam pipeline determined according to the topological diagram of each steam pipeline system recorded in the DCS system of the thermal power plant; the second type of data is associated with the structural characteristics of the steam pipeline.

4. The method for locating weak links in a steam pipeline system of a thermal power plant based on data fusion according to claim 1 is characterized in that: The multi-source data includes three types of data; the three types of data are state duration matrices; the state duration matrix is ​​constructed according to the following steps: the wear and failure process of the steam pipeline section is divided into four states: normal integrity, crack initiation, crack propagation and burst leakage; according to the inspection records of on-site personnel and the commissioning time of the steam pipeline system, the historical duration of each section of the steam pipeline in the above four states is determined, and then the corresponding state duration matrix is ​​formed; the three types of data are associated with the reliability characteristics of the steam pipeline.

5. The method for locating weak links in a steam pipeline system of a thermal power plant based on data fusion according to claim 1 is characterized in that: The multi-source data includes four types of data; the four types of data are displacement indicators; the displacement indicators are obtained according to the following steps: analyzing the manual inspection data of the steam pipeline system, analyzing the displacement of key parts in the steam pipeline, and qualitatively describing the displacement of each section of the steam pipeline, and converting the qualitative description into a numerical quantitative indicator, namely, the displacement indicator; the four types of data are associated with the stability characteristics of the steam pipeline.

6. The method for locating weak links in a steam pipeline system of a thermal power plant based on data fusion according to claim 1 is characterized in that: The multi-source data includes five types of data; the five types of data are thermal performance parameters; the thermal performance parameters include the forced convection heat transfer thermal resistance in the steam pipe, the thermal conductivity thermal resistance parameters of the insulation layer, the convection heat transfer thermal resistance parameters between the outer wall of the pipe and the environment, and the historical average ambient temperature parameters of different steam pipe locations; the five types of data are associated with the thermal performance characteristics of the steam pipe.

7. The method for locating weak links in a steam pipeline system of a thermal power plant based on data fusion according to claim 1 is characterized in that: The characteristic indicators in the steam pipeline state characteristic indicator system include: CV n =[AT n ,VT n ,CT n ,AP n ,VP n ,CP n ,BS n ,λ n (t n ),LN n ,QT n ]; In the formula, CV n is the characteristic index vector corresponding to the nth section of steam pipeline; AT n VT n , CT n They are the temperature average characteristic index, temperature variance characteristic index and temperature trend characteristic index of the nth section of steam pipeline respectively; AP n VP n , CP n They are the pressure average characteristic index, pressure variance characteristic index and pressure trend characteristic index of the nth section steam pipeline respectively; BS n is the structural importance index of the nth steam pipeline after normalization and weighting; n (t n ) represents the nth steam pipeline at its current service time t n Failure rate at time; LN n It is the stability characteristic index of the nth section of steam pipeline after normalization; QT n is the average heat flux density of the nth section of the steam pipe.

8. The method for locating weak links in a steam pipeline system of a thermal power plant based on data fusion according to claim 1 is characterized in that: In step 4, when the fuzzy entropy weight method is used to fuse the key feature indicators, the following steps are included: Create a fuzzy membership matrix W with m rows and h columns; the element W in W ng (n=1,2,…,m;g=1,2,…,h;) takes the key characteristic index value after standardization of each section of steam pipeline; Calculate the information entropy IE of each key feature indicator g : In the formula, IE g Represents the information entropy corresponding to the g-th key feature indicator; Based on the information entropy of the indicators and the preference weights of the operators of the thermal power plant, the proportional weight of each key characteristic indicator in the steam pipeline health evaluation is calculated as the fusion result; Where η g represents the fuzzy entropy weight of the g-th key feature index; g Represents the preference weight of the system operator for the g-th key characteristic indicator.

9. The method for locating weak links in a steam pipeline system of a thermal power plant based on data fusion according to claim 8, characterized in that: When calculating the health of each section of the steam pipeline, the health of each section of the steam pipeline is evaluated according to the proportional weight of each key characteristic indicator: In the formula, HI n It indicates the health evaluation result of the nth section of steam pipeline. The smaller the value is, the lower the operating status and health level of the steam pipeline is.

10. The method for locating weak links in a steam pipeline system of a thermal power plant based on data fusion according to claim 1, characterized in that: Step 1 also includes: building a multi-source data integration module in a data station of a DCS system of a thermal power plant; the multi-source data integration module is used to execute step 1; Step 2 also includes: building a characteristic index extraction module in a data station of a DCS system of a thermal power plant; the characteristic index extraction module is used to execute step 2; Step 3 also includes: building a characteristic index selection module in a data station of a DCS system of a thermal power plant; the characteristic index selection module is used to execute step 3; Step 4 also includes: building a result output module in the data station of the DCS system of the thermal power plant; the result output module is used to execute step 4.

Images