Terminal meter fault feature association and health assessment method, system and device and storage medium

Abstract

The invention discloses a terminal meter fault feature association and health assessment method, system and device and a storage medium, and relates to the field of electric power metering, and the method comprises the steps: building a full-life-cycle dynamic model of a fault from initiation to failure based on the fault types and performance degradation conditions of a terminal and a meter; based on the full life cycle dynamic model, extracting and analyzing feature information of an equipment operation state, and generating a dynamic feature association result; designing a health state evaluation system and a rule base based on a dynamic feature association result; according to the method, the fault types of the power equipment terminal and the meter are classified, the characteristics of the power equipment terminal and the meter are extracted and analyzed, the correlation technology of the fault characteristics and the equipment health state is developed, a scientific and unified health state evaluation rule base is designed, the universality and practicability of the health state evaluation rule base are verified, and finally a fault early warning mechanism and a maintenance strategy are optimized. And the operation efficiency and reliability of the power system are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of electric power metering, and in particular to a terminal meter fault feature correlation and health assessment method, system, device and storage medium. BACKGROUND

[0002] In the current power system, the assessment of the health status of terminals and meters is very important, as it determines the stable operation of the power system and the accuracy of electricity charges. In actual work, the health status assessment of terminals and meters is usually implemented through regular inspection by maintenance personnel, but this method not only consumes human resources, but also may cause delayed troubleshooting or missed opportunities due to untimely and incomplete inspection.

[0003] Therefore, the industry has some technical solutions, such as using intelligent technology for automated inspection. This method collects relevant information about the operation of terminals and meters, and then uses machine learning and artificial intelligence technology to analyze the data, thereby assessing the operating status of terminals and meters. Although this technology can reduce the workload of manual inspection and improve the efficiency of inspection and maintenance, it ignores the influence of environmental factors, load conditions, etc. on the operating status of terminals and meters, and is also limited by the accuracy of feature selection and the complexity of the algorithm. In addition, some technologies use embedded sensors to continuously monitor the status of terminals and meters. This method can obtain real-time operating data of the equipment and reflect the operating status of the equipment. However, this method requires modification of the equipment, increasing the cost of equipment and maintenance, and if the sensor fails or is disturbed, it will affect the accuracy of the health status assessment. SUMMARY

[0004] In view of the above problems, the present application is proposed.

[0005] Therefore, the technical problem solved by the present application is how to achieve comprehensive intelligent assessment of the health status of terminals and meters and improve the accuracy and efficiency of the assessment.

[0006] To solve the above technical problems, the present application provides the following technical solutions:

[0007] In a first aspect, an embodiment of the present application provides a terminal meter fault feature correlation and health assessment method, comprising:

[0008] Based on the fault types and performance degradation status of terminals and meters, a full life cycle dynamic model of faults from inception to failure is established;

[0009] Based on the full life cycle dynamic model, feature information of the operating status of the equipment is extracted and analyzed to generate a dynamic feature correlation result;

[0010] Based on the dynamic feature correlation result, a health state evaluation system and a rule base are designed.

[0011] As a preferred scheme of the terminal meter fault feature correlation and health evaluation method, wherein:

[0012] The full life cycle dynamic model of the fault from the inception to the failure is established based on the terminal and meter fault types and performance degradation conditions.

[0013] The historical fault data including fault logs, sensor monitoring data and operation records are collected, the faults are classified according to fault modes and features, and a fault type library is constructed.

[0014] As a preferred scheme of the terminal meter fault feature correlation and health evaluation method, wherein:

[0015] The full life cycle dynamic model of the fault from the inception to the failure is established based on the terminal and meter fault types and performance degradation conditions.

[0016] The nonlinear mathematical model describing the running state of the equipment is established based on the fault type library, electrodynamics principles and equipment characteristics, and the hidden semi-Markov model is initialized by using the nonlinear mathematical model, and the initialization includes defining states, transition probabilities and observation probabilities.

[0017] The beneficial effects of the preferred technical scheme are that the nonlinear mathematical model is established in combination with the fault type library and the hidden semi-Markov model is initialized, which can more accurately reflect the running conditions of the terminal meter in different states and the transition relationship between the states, provide a scientific and reasonable model basis for the construction of the full life cycle dynamic model, and improve the simulation accuracy of the model on the fault evolution process.

[0018] As a preferred scheme of the terminal meter fault feature correlation and health evaluation method, wherein:

[0019] The full life cycle dynamic model of the fault from the inception to the failure is established based on the terminal and meter fault types and performance degradation conditions.

[0020] The performance degradation features are extracted by using simulation software through a digital simulation method to simulate the influence of different environmental factors and load conditions on the performance of the equipment, and the hidden semi-Markov model after initialization is used to model the fault evolution process to generate the full life cycle dynamic model.

[0021] The beneficial effects of the preferred technical solution are: by simulating the influence of different environments and load conditions on the performance of the equipment through digital simulation and extracting performance degradation features, the effects of various actual factors on the terminal meter fault evolution can be considered, and the full life cycle dynamic model is more close to the actual situation. By using the initialized hidden semi-Markov model for modeling, the dynamic transition law of the fault state can be accurately captured, and reliable model support is provided for subsequent fault feature correlation and health assessment.

[0022] As an preferred scheme of the terminal meter fault feature correlation and health assessment method, wherein:

[0023] The full life cycle dynamic model is used to extract and analyze the feature information of the equipment operating state, and generate a dynamic feature correlation result, which includes:

[0024] Multi-source data including sensor data, environmental data and operation data are collected to construct a high-dimensional feature matrix containing various state information; the high-dimensional feature matrix is processed by dimension reduction to extract key features reflecting the equipment operating state and define the state of the full life cycle dynamic model.

[0025] The beneficial effects of the preferred technical solution are: collecting multi-source data to construct a high-dimensional feature matrix can comprehensively cover various information in the equipment operation process, providing rich data support for accurate analysis of equipment state. The high-dimensional feature matrix is processed by dimension reduction to extract key features, which can remove redundant information, improve data processing efficiency, and accurately define the state of the full life cycle dynamic model, which helps to more clearly grasp the relationship between equipment operating state and features.

[0026] As an preferred scheme of the terminal meter fault feature correlation and health assessment method, wherein:

[0027] The full life cycle dynamic model is used to extract and analyze the feature information of the equipment operating state, and generate a dynamic feature correlation result, which includes:

[0028] The state transition and observation model in the full life cycle dynamic model is used for feature correlation analysis to establish the mapping relationship between the low-dimensional manifold of the feature matrix and the equipment operating state, and to obtain the dynamic relationship between the equipment operating state and the features; deep learning technology is used for feature adaptive representation to extract fault features, and the extracted fault features are input into the full life cycle dynamic model to generate a dynamic feature correlation result.

[0029] The beneficial effects of the preferred technical solutions are that: the state transition and observation model are used for feature correlation analysis and mapping relationship establishment, the internal relationship between the equipment operation state and the features can be deeply mined, and the dynamic relationship can be accurately obtained. The deep learning technology is used for feature self-adaptive representation and extraction of fault features input model, the adaptability and accuracy of the model to the fault features can be enhanced, the generated dynamic feature correlation result is more reliable, and a strong basis is provided for subsequent health assessment.

[0030] As a preferred scheme of the terminal meter fault feature correlation and health assessment method, wherein:

[0031] The health state assessment system and rule library design based on the dynamic feature correlation result comprises:

[0032] The terminal and meter state key feature fusion mechanism is introduced, the key feature indexes related to the health state are determined and integrated to form a comprehensive feature vector, and the full life cycle dynamic model is input for state assessment.

[0033] In the second aspect, the embodiment of the present application provides a terminal meter fault feature correlation and health assessment system, comprising:

[0034] The full life cycle dynamic modeling module is used for establishing a full life cycle dynamic model from the emergence to the failure of the fault based on the fault type and performance degradation condition of the terminal and meter.

[0035] The dynamic feature correlation module is used for extracting and analyzing the feature information of the equipment operation state based on the full life cycle dynamic model, and generating a dynamic feature correlation result.

[0036] The assessment module is used for designing a health state assessment system and a rule library based on the dynamic feature correlation result.

[0037] In the third aspect, the embodiment of the present application provides an electronic device, comprising:

[0038] A memory and a processor.

[0039] The memory is used for storing computer executable instructions, and the processor is used for executing the computer executable instructions, so that the one or more processors implement the terminal meter fault feature correlation and health assessment method according to any embodiment of the present application.

[0040] In the fourth aspect, the embodiment of the present application provides a computer readable storage medium, which stores computer executable instructions, and the computer executable instructions are executed by a processor to implement the terminal meter fault feature correlation and health assessment method.

[0041] The present application has the following beneficial effects: the health state evaluation method based on big data and digital model linkage of the present application can relatively automatically complete the health state evaluation of terminals and meters in combination with intelligent decision support, thereby reducing the consumption of human resources; compared with the traditional periodic inspection method, the present application comprehensively considers environmental factors and load conditions and other influencing factors when evaluating the health state of terminals and meters, and can more accurately reflect the real running state of the equipment; through model learning and optimization, the present application can timely predict and warn possible faults of terminals and meters, which is beneficial to timely eliminate hidden dangers and improve the stability and reliability of the power system; after collecting enough state information, the present application can quickly make an evaluation result according to the health state rule base, greatly improving the decision efficiency and being helpful to the stable operation of the power system; through accurate and timely health state evaluation, it can effectively avoid larger-scale maintenance caused by equipment state deterioration, thereby reducing the maintenance cost of terminals and meters. BRIEF DESCRIPTION OF DRAWINGS

[0042] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor.

[0043] Figure 1 is the overall flowchart of the terminal meter fault feature correlation and health evaluation method provided by the present application. DETAILED DESCRIPTION

[0044] In order to make the above-mentioned purposes, features and advantages of the present application more apparent and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the drawings of the specification. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should be within the protection scope of the present application.

[0045] Embodiment 1, refer to Figure 1 , the first embodiment of the present application provides a terminal meter fault feature correlation and health evaluation method, which comprises:

[0046] S1: based on the fault type and performance degradation state of the terminal and the meter, a full life cycle dynamic model from the inception to the failure of the fault is established;

[0047] S2: based on the full life cycle dynamic model, the feature information of the equipment running state is extracted and analyzed to generate a dynamic feature correlation result;

[0048] S3: Based on the dynamic feature correlation result, the health state evaluation system and rule base design are performed.

[0049] It should be noted that through steps S1-S3, the embodiment sorts out the common fault types of the meter and the terminal, establishes a nonlinear mathematical model based on electricity, studies the fault initiation and evolution law, establishes the mapping relationship between faults at different evolution stages, analyzes and reveals the failure mechanism of typical faults. Through digital simulation method, the influence law of environmental factors, load conditions and other factors on the meter and terminal is explored, the performance degradation law and fault evolution mechanism are explored; a high-dimensional feature matrix of terminal and meter state information is constructed, and a low-dimensional manifold reflecting the key feature information of the running state is extracted. The mapping relationship between the low-dimensional manifold of the feature matrix and the running state of the terminal and meter is revealed. A deep fault feature adaptive representation method for multi-source data is established, and a terminal and meter fault type and feature correlation technology is constructed; a terminal and meter state key feature fusion mechanism is proposed, and key feature indexes related to health state are established. A life cycle stage-based, multi-scale health state evaluation index system is constructed. Combined with domain knowledge and standards, based on the health state evolution mechanism, a scientific, perfect and unified health state evaluation rule base is constructed.

[0050] Embodiment 2, refer to Figure 1 For an embodiment of the present application, a terminal meter fault feature correlation and health evaluation method is provided based on the previous embodiment, comprising:

[0051] In the embodiment, the above step S1 based on the fault type and performance degradation status of the terminal and meter establishes a full life cycle dynamic model from fault initiation to failure, including:

[0052] Data collection and fault type sorting are performed;

[0053] Specifically, historical fault data including fault logs, sensor monitoring data and operation records are collected; the collected data are cleaned and classified through data mining technology and expert knowledge, and noise and outliers are removed; the faults are classified according to fault mode and feature, and a comprehensive fault type library is constructed.

[0054] Nonlinear mathematical model establishment and HSMM (Hidden Semi-Markov Models) initialization are performed, and a full life cycle dynamic model from fault initiation to failure is established based on the initialized HSMM;

[0055] It should be noted that the HMM (Hidden Markov Model) cannot truly reflect the duration of each degradation state in the degradation process, and it is necessary to improve the model description performance evolution law. The HSMM is an extension of the HMM. The HSMM has two characteristics superior to the HMM: first, the HSMM overcomes the limitations caused by the Markov chain assumption, has better analysis and modeling capabilities when solving real problems, and has more accurate state recognition and classification accuracy; second, the HSMM has the characteristics of reasonably describing the duration of each state, and is more suitable for performance degradation state recognition compared with the HMM.

[0056] The influence of different environmental factors and load conditions on the performance of the equipment is simulated by a digital simulation method, the simulation software (such as ANSYS, COMSOL) is used for simulation, and the performance degradation characteristics are extracted; the HSMM is used to model the fault evolution process, and the dynamic transfer law of the fault state is captured; the fault mapping relationship in different evolution stages is established, and the whole process from the inception to the evolution of the fault is described.

[0057] Specifically, based on the electrodynamic principle and the equipment characteristics, a nonlinear mathematical model for describing the running state of the equipment is established, and the nonlinear mathematical model is used to initialize the HSMM, including defining the state, the transition probability and the observation probability. The HSMM is an HMM considering that the state residence probability distribution is explicit (P i (d)) and is expressed as:

[0058] λ=(π,A,B,P i (d))

[0059] Wherein, π represents an initial probability distribution vector, π ∈ (π1, π2,..., π N ), wherein:

[0060] π i =P(q1=h i ),1≤i≤N

[0061] Wherein, q1 represents the state at the initial time 1.

[0062] A represents a state transition probability matrix, there are N states in the HSMM, A = {a ij} N×N , wherein:

[0063] a ij =P(q t+1 =s j |q t =s i ),1≤i,j≤N

[0064] B represents an observation value probability matrix, B = {b ij}N×N ,in:

[0065] b jk =P(O t =V k |q t =h j ),1≤j≤N,1≤k≤M

[0066] P i (d) represents the probability distribution of state residence time, i.e., state S i The probability of lasting d time units, where:

[0067] p i (d) = P(d|q t =i),1≤n≤N,1≤d≤D

[0068] Where D represents the maximum state residence time.

[0069] Since the initial state is normal, the initial state probability vector is:

[0070] π=[1,0,...,0]

[0071] The degradation state transition probability matrix is ​​expressed as:

[0072]

[0073] The state residence distribution C of HSMM is (P i (d)) N×D Probability distribution of available display state dwell time p i (d) means, where:

[0074] p i (d) = P(d|q t =i),1≤n≤N,1≤d≤D

[0075] Each state residence time D(h i ) of the mean μ(h i ) and variance σ 2 (h i ) can be obtained through the state residence distribution C.

[0076] Under the constraints In this case, maximize The residence time of each degenerate state is obtained, which is expressed as:

[0077] D(h i )=μ(h i )+ρσ 2 (h i )

[0078]

[0079] Further, it is obtained:

[0080] D(h i )=f(μ(h i ),σ 2 (h i ))

[0081] where μ(h i ) and σ 2 (h i ) are the mean and variance of the state residence time.

[0082] In another possible implementation, in the process of establishing a nonlinear mathematical model describing the running state of the device and initializing the HSMM using the nonlinear mathematical model, the states can be divided based on the simulation output of the nonlinear model.

[0083] For example, in the error model of the electric meter ∈(T)=a·e bT ,

[0084] Divide the state Z1 (normal): ∈≤0.5% (temperature T≤40℃);

[0085] State Z2 (mild degradation): 0.5%<∈≤1.5% (40℃<T≤60℃);

[0086] State Z3 (severe degradation): ∈>1.5% (T>60℃).

[0087] Transition probability (A matrix): the state transition frequency is counted by Monte Carlo simulation.

[0088] For example, in 100 times of temperature cycle simulation, Z1→Z2 transition occurs 15 times, then A 12 =0.15.

[0089] Observation probability (B matrix): the distribution is fitted according to the actual sensor data in the state.

[0090] For example, the current harmonic distortion rate in the Z2 state obeys N(μ=3%, σ=0.8%).

[0091] Residence time distribution: record the duration of each state in the simulation, and fit the Weibull distribution.

[0092] In the embodiment, the feature information of the running state of the device is extracted and analyzed based on the full life cycle dynamic model in the above step S2, and the dynamic feature correlation result includes:

[0093] Multi-source data including sensor data, environmental data and operation data are collected, and data fusion technology and database management system are used to fuse and store the data, and a high-dimensional feature matrix containing various state information is constructed.

[0094] The high-dimensional feature matrix is processed by dimension reduction through principal component analysis (PCA), linear discriminant analysis (LDA) and self-encoder, etc. Key features reflecting the running state of the equipment are extracted therefrom, and the low-dimensional features are used to define the state of the full life cycle dynamic model.

[0095] The state transition and observation model in the full life cycle dynamic model is used for feature correlation analysis to establish the mapping relationship between the low-dimensional manifold of the feature matrix and the running state of the equipment, and to capture the dynamic relationship between the running state of the equipment and the features. In order to further improve the representation ability of the fault features, deep learning technology (such as convolutional neural network, long short-term memory network) is used for adaptive representation of the features, and the extracted fault features are input into the full life cycle dynamic model to enhance the adaptability and accuracy of the model to the fault features, so as to generate dynamic feature correlation results.

[0096] In another possible implementation, when defining the state of the full life cycle dynamic model by using the low-dimensional features, the principal component vector PC1 after dimension reduction can be defined as the state boundary;

[0097] For example, PC1∈[-∞,-1.0]PC1∈[-∞,-1.0], it is determined as a fault state (Z4);

[0098] PC1∈(-1.0,0.5]PC1∈(-1.0,0.5], it is determined as a normal state (Z1);

[0099] PC1∈(0.5,2.0]PC1∈(0.5,2.0], it is determined as a slight degradation (Z2).

[0100] The state classification accuracy (≥90%) is calculated by confusion matrix.

[0101] In this embodiment, the health state evaluation system and rule base design based on the dynamic feature correlation results in step S3 include:

[0102] A terminal and meter state key feature fusion mechanism is introduced to determine key feature indicators related to the health state, select key features through feature selection algorithms (such as L1 regularization, recursive feature elimination) and expert scoring methods, and integrate them to form a comprehensive feature vector, which is input into the full life cycle dynamic model for state evaluation.

[0103] In combination with statistical analysis methods and fault tree analysis (FTA), a health state evaluation index system is designed, the health levels of different states are defined in the full life cycle dynamic model, and a full life cycle phased and multi-scale health state evaluation index system is established. Then, according to the domain knowledge and industry standards, a health state evaluation rule library is designed, and dynamic evaluation and verification are realized under the framework of the full life cycle dynamic model. Through knowledge engineering technology and expert systems, the evaluation rule library is continuously optimized.

[0104] The health state evaluation rule library is constructed: IoT (Internet of Things) devices and edge computing technology are deployed in the actual operating environment to realize real-time data collection and processing, online state evaluation and health state monitoring are realized by using the full life cycle dynamic model, and through feedback in actual operation, the evaluation system and rule library are continuously adjusted and optimized to ensure their practicality and reliability in different application scenarios.

[0105] In another possible implementation, the transition probability can be updated using the Bayesian update rule, which is represented as:

[0106]

[0107] where Δn ij is the number of transitions from i to j observed in the new data, and n is the total number of historical transitions.

[0108] The distribution parameters (such as α and β of the gamma distribution) can be refitted according to the new data.

[0109] When the evaluation error is greater than 10% for 3 consecutive times, the update is started, which can ensure that the model can adapt to changes in the actual operating environment.

[0110] In another possible implementation, when designing the health state evaluation index system, the score threshold can be defined as:

[0111] Healthy (100-80 points): all key features are within the normal range, and the HSMM state is Z1;

[0112] Warning (79-60 points): one or more features exceed the threshold but have not failed (such as Z2 state lasting for more than 24 hours);

[0113] Failure (<60 points): HSMM enters Z3 / Z4 or key features exceed the safety limit.

[0114] The corresponding maintenance strategy is:

[0115] Healthy state: quarterly inspection;

[0116] Warning state: weekly remote diagnosis and monthly on-site inspection;

[0117] Failure state: immediate shutdown for repair.

[0118] Embodiment 3, the above is a schematic scheme of the terminal meter fault feature correlation and health assessment method of the embodiment. It should be noted that the technical scheme of the terminal meter fault feature correlation and health assessment system belongs to the same concept as the technical scheme of the terminal meter fault feature correlation and health assessment method described above. The technical scheme of the terminal meter fault feature correlation and health assessment system in this embodiment is not described in detail. The details can be referred to the description of the technical scheme of the terminal meter fault feature correlation and health assessment method described above.

[0119] The embodiment also provides a terminal meter fault feature correlation and health assessment system, comprising:

[0120] A full life cycle dynamic modeling module is configured to establish a full life cycle dynamic model of fault from inception to failure based on the fault type and performance degradation status of the terminal and meter.

[0121] A dynamic feature correlation module is configured to extract and analyze feature information of the device operating state based on the full life cycle dynamic model, and generate a dynamic feature correlation result.

[0122] An evaluation module is configured to design a health state evaluation system and a rule library based on the dynamic feature correlation result.

[0123] The embodiment also provides an electronic device suitable for the terminal meter fault feature correlation and health assessment method, comprising:

[0124] A memory and a processor; the memory is configured to store computer executable instructions, and the processor is configured to execute the computer executable instructions to implement the terminal meter fault feature correlation and health assessment method proposed in the above embodiment.

[0125] The embodiment also provides a storage medium having a computer program stored thereon, which is executed by a processor to implement the terminal meter fault feature correlation and health assessment method proposed in the above embodiment.

[0126] The storage medium proposed in the embodiment and the terminal meter fault feature correlation and health assessment method proposed in the above embodiment belong to the same inventive concept. The technical details not described in detail in the embodiment can be referred to the above embodiment, and the embodiment has the same beneficial effects as the above embodiment.

[0127] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not limiting. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present application. They should all be covered in the scope of the claims of the present application.

Claims

1. A terminal meter fault feature correlation and health assessment method, characterized in that: include: Based on the fault types and performance degradation conditions of terminals and meters, a dynamic model of the entire life cycle from fault initiation to failure is established; Based on the full life cycle dynamic model, extract and analyze the characteristic information of the equipment operation status and generate dynamic feature correlation results; Based on the dynamic feature association results, the health status assessment system and rule base are designed.

2. A terminal meter fault feature correlation and health assessment method according to claim 1, characterized in that: The establishment of a full life cycle dynamic model from fault initiation to failure based on the fault types and performance degradation conditions of terminals and meters includes: Collect historical fault data including fault logs, sensor monitoring data and operation records, classify faults according to fault modes and characteristics, and build a fault type library.

3. A terminal meter fault feature correlation and health assessment method according to claim 2, characterized in that: The establishment of a full life cycle dynamic model from fault initiation to failure based on the fault types and performance degradation conditions of terminals and meters also includes: Combined with the fault type library, based on electrical principles and equipment characteristics, a nonlinear mathematical model describing the equipment operating status is established, and the nonlinear mathematical model is used to initialize the hidden semi-Markov model. The initialization includes defining the state, transition probability and observation probability.

4. A terminal meter fault feature correlation and health assessment method according to claim 3, characterized in that: The establishment of a full life cycle dynamic model from fault initiation to failure based on the fault types and performance degradation conditions of terminals and meters also includes: The impact of different environmental factors and load conditions on equipment performance is simulated through digital simulation methods, and performance degradation characteristics are extracted using simulation software. The fault evolution process is modeled using the initialized hidden semi-Markov model to generate a full life cycle dynamic model.

5. A terminal meter fault feature correlation and health assessment method according to claim 4, characterized in that: The extraction and analysis of characteristic information of the equipment operation status based on the full life cycle dynamic model to generate dynamic characteristic association results includes: Collect multi-source data including sensor data, environmental data and operation data, and construct a high-dimensional feature matrix containing various status information; perform dimensionality reduction processing on the high-dimensional feature matrix, extract key features reflecting the operating status of the equipment and define the status of the full life cycle dynamic model.

6. A terminal meter fault feature correlation and health assessment method according to claim 5, characterized in that: The extracting and analyzing the characteristic information of the equipment operation status based on the full life cycle dynamic model and generating the dynamic characteristic association result further includes: The state transition and observation model in the full life cycle dynamic model is used to perform feature association analysis, establish a mapping relationship between the low-dimensional manifold of the feature matrix and the equipment operating status, and obtain the dynamic relationship between the equipment operating status and the features; deep learning technology is used for feature adaptive representation to extract fault features, and the extracted fault features are input into the full life cycle dynamic model to generate dynamic feature association results.

7. A terminal meter fault feature correlation and health assessment method according to claim 6, characterized in that: The design of the health status assessment system and rule base based on the dynamic feature association results includes: A key feature fusion mechanism of terminal and meter status is introduced to determine the key feature indicators related to the health status and integrate them to form a comprehensive feature vector, which is input into the full life cycle dynamic model for status assessment.

8. A terminal meter fault feature correlation and health assessment system, applying the method according to any one of claims 1 to 7, characterized in that: include: The full life cycle dynamic modeling module is used to establish a full life cycle dynamic model from fault initiation to failure based on the fault types and performance degradation conditions of terminals and meters; Dynamic feature association module, used to extract and analyze the feature information of equipment operation status based on the full life cycle dynamic model and generate dynamic feature association results; The evaluation module is used to design the health status evaluation system and rule base based on the dynamic feature association results.

9. An electronic device, characterized in that: include: memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that It stores computer-executable instructions, which implement the steps of the method according to any one of claims 1 to 7 when executed by a processor.

Images