Failure analysis and system design method for conduction insulation tester

Abstract

The invention relates to the technical field of electrical equipment testing, and particularly discloses a conduction insulation tester failure analysis and system design method which comprises the following steps: firstly, establishing a fault tree model of tester failure, and performing failure data analysis; a combined model based on grey model prediction and BP neural network optimization series connection is established, and failure prediction is carried out; then obtaining predicted failure time by adopting a regression analysis method; and finally, optimizing the design and manufacturing process of the conduction insulation tester by using the obtained failure time and the analysis result. According to the failure analysis and system design method for the conduction and insulation tester, a modularized conduction and insulation test system is designed, through redundancy design and control of upper computer software, the system failure probability is reduced, the reliability is enhanced, meanwhile, the prediction precision and the system stability of the conduction and insulation tester can be improved, and the reliability of the conduction and insulation tester is improved. And the maintenance cost is reduced, and the method has important significance for improving the reliability of electrical test equipment.

Description

Technical Field

[0001] The invention relates to the technical field of electrical equipment testing, and in particular to a failure analysis and system design method for a continuity and insulation tester. Background Art

[0002] With the increasing complexity and sophistication of modern electrical systems, the cable network, as a key infrastructure, its conduction and insulation performance are crucial to the stability and reliability of the entire system. The cable network carries the task of transmitting energy and signals, and plays an irreplaceable role in both civil and military fields. However, the cable network is composed of a large number of wires, with a complex structure and interconnectedness. The stability and reliability of its performance are affected by various factors, such as material aging, environmental factors, mechanical damage, etc. Therefore, accurate and efficient conduction and insulation testing of the cable network has become a key link to ensure the normal operation of the electrical system and avoid catastrophic accidents.

[0003] Traditional continuity insulation test methods often rely on manual operation, which is not only inefficient, but also easily affected by human factors, making it difficult to ensure the accuracy and reliability of the test results. In addition, with the continuous development and upgrading of electrical systems, the scale and complexity of cable networks are also increasing, and traditional test methods can no longer meet the needs of modern production lines for efficiency and automation. Therefore, researching and developing a new type of continuity insulation test system with the characteristics of automation, intelligence, and efficiency has important practical significance and application value for improving the reliability and stability of electrical systems and ensuring the safe operation of equipment. Summary of the invention

[0004] The purpose of the present invention is to provide a failure analysis and system design method for a continuity insulation tester, which can achieve effective prediction and prevention of failure of the continuity insulation tester and improve the reliability and stability of the system through the comprehensive application of fault tree analysis, failure prediction and system design.

[0005] To achieve the above object, the present invention provides a failure analysis and system design method for a continuity insulation tester, comprising the following steps:

[0006] S1. Establish a fault tree model for tester failure and perform failure data analysis;

[0007] S2. Establish a combined model based on grey model prediction and BP neural network optimization series connection to perform failure prediction;

[0008] S3, using regression analysis to obtain the predicted failure time;

[0009] S4. Use the obtained failure time and analysis results to optimize the design and manufacturing process of the continuity insulation tester.

[0010] Preferably, step S1 specifically includes the following steps:

[0011] S11. By analyzing the logical relationship between the functional parts, determine the event set for fault tree analysis of the tester failure, use the fault tree analysis method, take the failure of the system conduction detection function and the failure of the insulation detection function as the top events, and establish the fault tree model of the tester failure system;

[0012] S12. Conduct qualitative and quantitative failure analysis on the continuity insulation tester. By analyzing its minimum cut set, failure probability, structural importance and probability importance, it is identified that the matrix board and the relays on the matrix board have the highest contribution to the system failure, and they are determined as the key components and weak links of the continuity insulation tester.

[0013] Preferably, step S1 further includes analyzing the component level of the tester.

[0014] Preferably, in step S12, the failure data is preprocessed using statistical methods and expert experience before performing failure analysis.

[0015] Preferably, in step S2, failure prediction is performed based on the weak links prone to failure determined in step S12, and the failure prediction includes determining prediction variables for characterizing the degradation trajectory, and inputting the prediction variables into a combined model based on the series connection of the grey model and the BP neural network model for prediction training.

[0016] Preferably, in step S2, the GM (1,1) grey model is used to predict the degradation trajectory of the relay contact resistance, and the BP neural network model is used to correct the prediction result. The specific process is:

[0017] For N 2 K-1 relays with sufficient contact resistance sample data, masking their Nth 1 +1~N 2 valid data, assuming that only the first N 1 In the case of sample data, the corresponding Nth 1 +1~N 2 The gray prediction result train_gm of the first step is used as the input data of the BP neural network, and the Nth step is actually observed. 1 +1~N 2 The data is recorded as train_real and used as the output data of the BP neural network to train the network.

[0018] Preferably, in step S3, a regression analysis method is used to extrapolate the life of the relay to obtain a predicted failure time.

[0019] Therefore, the present invention adopts the above-mentioned conduction insulation tester failure analysis and system design method, and the beneficial effects are as follows:

[0020] (1) The present invention uses the fault tree analysis method to analyze the overall composition and functional logic of the conductive insulation tester and establish a fault tree model of the tester failure. Through the fault tree model, the fault hierarchy and logical relationship of the conductive insulation tester can be clearly understood, providing a basis for subsequent failure prediction and system design.

[0021] (2) The present invention combines the failure data obtained from engineering experience to conduct qualitative and quantitative analysis on the fault tree model, and identifies the main failure modes and weak links of the system. Through in-depth analysis of failure data, the key factors and potential risks that lead to tester failure can be accurately identified, and early warning and preventive maintenance can be performed on key components to reduce failure risks, improve equipment reliability, and provide guidance for subsequent failure prediction and system design.

[0022] (3) The present invention targets weak links prone to failure, such as matrix board relays, by analyzing the failure mechanism of relays, determining the predictive variables used to characterize the degradation trajectory, proposing a combined model based on the series connection of the grey model and the BP neural network model, and using the regression analysis method to perform life extrapolation to obtain the predicted failure time. Through failure prediction technology, early warning and preventive maintenance of key components such as relays can be achieved, reducing the risk of failure.

[0023] (4) The present invention independently designs and builds a conduction insulation test system based on the failure prediction results. Through modular design, the system matrix boards are made redundant spare parts to reduce the probability of system failure. At the same time, the present invention provides ideas and methods for collecting and monitoring the contact resistance of the relay in the tester, reduces system failures caused by relay failures through early warning, and enhances system reliability. Through modular design and redundant backup strategies, the scalability and fault tolerance of the system can be improved, and the impact of a single failure on the entire system can be reduced.

[0024] (5) The present invention is designed to cooperate with the host computer software running in the test instrument equipment to realize system control and data processing functions. Through the host computer software, real-time monitoring and control of the test process, as well as collection, analysis and processing of test data can be realized; at the same time, the host computer software can also provide a user-friendly operation interface and powerful data analysis tools, which is convenient for users to configure and maintain the system, thereby improving the flexibility and ease of use of the system.

[0025] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1It is a functional logic diagram of a conduction insulation test system of an embodiment of a conduction insulation tester failure analysis and system design method of the present invention;

[0027] Figure 2 It is a grey neural network combination model of a failure analysis and system design method of a continuity insulation tester of the present invention;

[0028] Figure 3 It is a principle block diagram of a main control board of an embodiment of a conduction insulation tester failure analysis and system design method of the present invention;

[0029] Figure 4 It is a control logic diagram of the upper computer and the lower computer of an embodiment of a conduction insulation tester failure analysis and system design method of the present invention;

[0030] Figure 5 The invention discloses a general scheme of a conduction insulation test system according to an embodiment of a conduction insulation tester failure analysis and system design method. DETAILED DESCRIPTION

[0031] The technical solution of the present invention is further described below through the accompanying drawings and embodiments.

[0032] Unless otherwise defined, technical or scientific terms used in the present invention shall have the common meanings understood by one having ordinary skills in the field to which the present invention belongs.

[0033] The present invention is based on the research of the failure prediction and system design of the continuity insulation tester of the fault tree analysis, which is carried out to solve the above problems. Through the in-depth analysis and research of the failure mode of the cable network, the weak links and key components in the system can be identified, providing an important basis for the design and optimization of the test system. At the same time, through the failure prediction and life evaluation of key components, potential safety hazards can be discovered in time, and corresponding preventive measures can be taken to avoid the occurrence of failures. This continuity insulation test system based on fault tree analysis and failure prediction not only improves the test efficiency and accuracy, but also reduces the equipment failure rate and maintenance cost, providing a strong guarantee for the long-term stable operation of the electrical system.

[0034] In addition, the invention also has important theoretical value and innovative significance. Through in-depth research on the failure mechanism of cable networks, the reliability theory of electrical systems can be enriched and improved, providing useful references for research in related fields. At the same time, the conduction insulation test method based on fault tree analysis and failure prediction proposed by the invention is innovative and practical, and can provide new ideas and methods for the maintenance and management of electrical equipment, and promote the development of electrical systems to a higher level of reliability and safety.

[0035] As shown in the figure, a failure analysis and system design method for a continuity insulation tester includes the following steps:

[0036] S1. Establish a fault tree model for tester failure and perform failure data analysis, which specifically includes the following steps:

[0037] S11. By analyzing the logical relationship between each functional part, the event set for fault tree analysis of the tester failure is determined. On this basis, the fault tree analysis method is used, with the failure of the system conduction detection function and the failure of the insulation detection function as the top event, to analyze the overall composition and functional logic of the conduction insulation tester, and establish a fault tree model of the tester failure system. The fault tree model mainly analyzes the system components and functional logic, and summarizes the events included in the failure of the system conduction function into 1 top event, 19 bottom events and several intermediate events, and explains the event code and the corresponding specific meaning. Each bottom event is independent of each other, and there are only two states of "normal" or "failure", and there is no third fuzzy state between the two. Through the fault tree model, the fault hierarchy and logical relationship of the tester can be clearly understood, providing a basis for subsequent failure prediction and system design.

[0038] S12. Use statistical methods and expert experience to pre-process the failure data. The pre-processing includes:

[0039] Data cleaning: Remove erroneous data, duplicate data, and incomplete records from the data set. Data standardization: Convert data to a uniform scale for comparison and analysis. Data normalization: Scale data to a specified range, such as between 0 and 1. Missing value processing: For missing data, you can choose to fill missing values ​​(such as using the mean, median, or mode), or delete records containing missing values. Outlier detection and processing: Identify outliers in the data set and decide whether to delete these values ​​or make corrections. Data transformation: Transform the data, such as logarithmic transformation, square root transformation, etc., to improve the normal distribution characteristics of the data. Data encoding: Convert categorical variables to numerical form, such as using One-Hot Encoding or Label Encoding. Feature engineering: Create new features or modify existing features to improve the performance of the model. Data integration: Integrate data sets from different sources to obtain a more comprehensive data analysis perspective. Feature selection: Select the most informative features from multiple features to reduce the complexity of the model and the risk of overfitting. Data splitting: Divide the data into training and test sets to evaluate the performance of the model. Expert experience integration: Combine the knowledge and experience of experts to interpret and verify the data.

[0040] Then, we conducted qualitative and quantitative failure analysis on the insulation tester. By analyzing its minimum cut set, failure probability, structural importance, and probability importance, we identified that the matrix board and the relays on the matrix board had the highest contribution to the system failure, and identified them as the key components and weak links of the insulation tester. Through in-depth analysis of failure data, we can accurately identify the key factors and potential risks that lead to the failure of the tester, providing guidance for subsequent failure prediction and system design.

[0041] The fault analysis in step S1 also includes analyzing the component level of the tester to establish a more accurate fault tree model.

[0042] S2. Establish a combined model based on grey model prediction and BP neural network optimization series connection to perform failure prediction;

[0043] The learning process of BP neural network consists of two processes: forward propagation and back propagation. In the forward propagation process, the input information is processed layer by layer from the input layer through the hidden layer and transmitted to the output layer. The state of each layer of neurons only affects the state of the next layer of neurons. If the output result obtained in the output layer does not meet the expectations, the back propagation process of the error begins, and the error signal is returned along the original connection channel, and the weights of the neurons in each layer are adjusted to minimize the error signal.

[0044] Failure prediction is performed based on the weak links prone to failure determined in step S12, and failure prediction is performed by analyzing the failure mechanism of the relay, including determining prediction variables for characterizing the degradation trajectory, and inputting the prediction variables into a combined model based on the series connection of the grey model and the BP neural network model for prediction training.

[0045] The GM (1,1) grey model is used to predict the degradation trajectory of the relay contact resistance, and the BP neural network model is used to correct the prediction results. The specific process is as follows:

[0046] For N 2 K-1 relays with sufficient contact resistance sample data, masking their Nth 1 +1~N 2 valid data, assuming that only the first N 1 In the case of sample data, the corresponding Nth 1 +1~N 2 The gray prediction result train_gm of the first step is used as the input data of the BP neural network, and the Nth step is actually observed. 1 +1~N 2The data is recorded as train_real and used as the output data of the BP neural network to train the network. At this time, each of the (K-1) relays can provide (N2-N1) input and output samples, which can meet the requirements of the BP neural network training process for larger samples.

[0047] S3. Regression analysis is used to extrapolate the life of relays to obtain predicted failure time, which can achieve early warning and preventive maintenance of key components such as relays and reduce failure risks.

[0048] S4. Use the obtained failure time and analysis results to optimize the design and manufacturing process of the continuity insulation tester.

[0049] like Figure 1 As shown: the normal and effective execution of the system functions depends on the functional coordination between the main control board, matrix board, power board and baseboard, and the process shown responds step by step. If any link in the process fails, the system as a whole will be unable to complete the established functions and will change from a normal working state to a failed state. Among them, the failure of any functional component in the main control board, including the controller, multiplexer, instrument amplifier, analog-to-digital converter, serial communication module, operational amplifier and transistor in the excitation source, etc., will cause the main control board to fail. For the matrix board, if any of its functional components CPLD or Darlington driver chip or relay in the switch array fails, it will also cause the matrix board to fail. For the power board, only if the AC-DC conversion module and the DC voltage conversion module can work normally, the power board will not fail.

[0050] like Figure 2 As shown in the figure: First, the BP neural network is trained using the relay group information. For the (K-1) relays with N2 sufficient contact resistance sample data, the valid data of the N1+1~N2th are masked, and it is assumed that only the first N1 sample data are known. The gray prediction results train_gm corresponding to the N1+1~N2th steps are obtained through the GM(1,1) model as the input data of the BP neural network, and the actually observed data of the N1+1~N2th steps are recorded as train_real as the output data of the BP neural network to train the network. At this time, the (K-1) relays can each provide (N2-N1) input and output samples, which can meet the demand for larger samples in the BP neural network training process.

[0051] like Figure 3 As shown: The main control board integrates the functions of the excitation part, the sensing part and the control part of the conduction insulation test system. It is connected to the power board through the base board, provides excitation to the object under test through the matrix board, realizes signal sensing, and is responsible for completing the communication with the host computer and other functional boards.

[0052] like Figure 4 As shown: After the tester is powered on, it first performs hardware initialization, checks the status of each matrix board, checks and manages the available hardware resources, sets GPIO and resets related relays to complete hardware initialization; when the main control board receives the task information sent by the host computer, it parses the UDP data message according to the instruction format to complete the conduction and insulation test.

[0053] Therefore, the present invention adopts the above-mentioned conduction insulation tester failure analysis and system design method to identify the weak links and key components in the system, providing an important basis for the design and optimization of the test system. At the same time, through the failure prediction and life evaluation of key components, potential safety hazards can be discovered in time, and corresponding preventive measures can be taken to avoid the occurrence of failures, which not only improves the test efficiency and accuracy, but also reduces the equipment failure rate and maintenance cost, providing a strong guarantee for the long-term stable operation of the electrical system.

[0054] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solution of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solution to deviate from the spirit and scope of the technical solution of the present invention.

Claims

1. A method for failure analysis and system design of a continuity insulation tester, characterized in that: The following steps are involved: S1. Establish a fault tree model for tester failure and perform failure data analysis; S2. Establish a combined model based on grey model prediction and BP neural network optimization series connection to perform failure prediction; S3, using regression analysis to obtain the predicted failure time; S4. Use the obtained failure time and analysis results to optimize the design and manufacturing process of the continuity insulation tester.

2. A method for failure analysis and system design of a continuity insulation tester according to claim 1, characterized in that: Step S1 specifically includes the following steps: S11. By analyzing the logical relationship between the functional parts, determine the event set for fault tree analysis of the tester failure, use the fault tree analysis method, take the failure of the system conduction detection function and the failure of the insulation detection function as the top events, and establish the fault tree model of the tester failure system; S12. Conduct qualitative and quantitative failure analysis on the continuity insulation tester. By analyzing its minimum cut set, failure probability, structural importance and probability importance, it is identified that the matrix board and the relays on the matrix board have the highest contribution to the system failure, and they are determined as the key components and weak links of the continuity insulation tester.

3. A method for failure analysis and system design of a continuity insulation tester according to claim 2, characterized in that: Step S1 also includes analyzing the component level of the tester.

4. A method for failure analysis and system design of a continuity insulation tester according to claim 3, characterized in that: In step S12, the failure data is preprocessed using statistical methods and expert experience before performing failure analysis.

5. A method for failure analysis and system design of a continuity insulation tester according to claim 4, characterized in that: In step S2, failure prediction is performed based on the weak links prone to failure determined in step S12. Failure prediction includes determining prediction variables for characterizing degradation trajectories, and inputting the prediction variables into a combined model based on a series connection of a grey model and a BP neural network model for prediction training.

6. A method for failure analysis and system design of a continuity insulation tester according to claim 5, characterized in that: In step S2, the GM (1,1) grey model is used to predict the degradation trajectory of the relay contact resistance, and the BP neural network model is used to correct the prediction result. The specific process is as follows: For K-1 relays with N2 sufficient contact resistance sample data, mask their N1+1~N2th valid data, and assume that only the first N1 sample data are known. Through the GM(1,1) grey model, obtain the grey prediction result train_gm corresponding to the N1+1~N2th steps. The prediction result train_gm is used as the input data of the BP neural network, and the actually observed N1+1~N2th data is recorded as train_real as the output data of the BP neural network to train the network.

7. A method for failure analysis and system design of a continuity insulation tester according to claim 6, characterized in that: In step S3, the life of the relay is extrapolated using a regression analysis method to obtain a predicted failure time.

Images