Power distribution box internal component aging diagnosis and maintenance decision-making method and system

Abstract

The invention discloses a power distribution box internal component aging diagnosis maintenance decision method and system, and relates to the technical field of component aging diagnosis maintenance, and the method comprises the steps: carrying out the random disturbance of an initial component decomposition value based on a computer random number, and obtaining a plurality of different component decomposition values; decomposing the operation data by applying a plurality of different component decomposition values based on EMD, and further performing element aging diagnosis; acquiring equipment aging identification data and anomaly detection sensitive data in an abnormal data detection process based on EMD; constructing a component decomposition value evaluation model based on a bp neural network; and establishing a component decomposition value-effect curve by using the output of the component decomposition value evaluation model and the corresponding component decomposition value, and performing curve analysis to obtain an improved component decomposition value to be applied to element aging diagnosis. The problem that setting of component decomposition values lacks a unified standard, and accuracy of aging diagnosis and maintenance decision is affected is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of component aging diagnosis and maintenance, and more specifically, to a method and system for making decisions on aging diagnosis and maintenance of components inside a power distribution box. Background Art

[0002] In the existing methods for diagnosing and repairing components in power distribution boxes, by optimizing the collection and storage methods of distribution box operation data, the power load data can be split into components and abnormality detected, and key data can be screened and stored, effectively reducing storage space occupation while retaining key information on abnormal load changes; these abnormal data can be used to identify possible component aging characteristics and provide data support for subsequent diagnosis; In addition, the technology of automatically generating system models through the layout information of the distribution box can combine the structure and layout characteristics of the distribution box to conduct multi-dimensional analysis of the operating status of the components and extract aging characteristic data such as current, voltage fluctuations, temperature rise changes, etc. These characteristic data can be input into the risk assessment model based on machine learning or neural networks to generate component aging diagnosis reports and evaluate the degree of aging and risk level; At the same time, the maintenance decision support method based on aging risk assessment prioritizes components by risk level, formulates emergency maintenance plans for high-risk components, and combines historical data and diagnostic results to provide maintenance suggestions such as component replacement, cleaning or parameter adjustment, further extending the service life of the distribution box and improving the operating stability of the system.

[0003] For example, the invention patent with announcement number: CN116755641B announces a method for optimizing the collection and storage of distribution box operation data, the method comprising: obtaining a power load sequence; setting component splitting control conditions according to the power load sequence, and obtaining multiple IMF component sequences according to the component splitting control conditions; calculating the similarity of each IMF component sequence, and obtaining a similarity component sequence according to the similarity; calculating the contribution degree of each similarity component sequence, and obtaining multiple strong contribution component sequences and multiple weak contribution component sequences according to the contribution degree and the similarity; obtaining a power consumption abnormality judgment condition according to the strong contribution component sequence and the weak contribution component sequence, and performing storage control on the power load sequence according to the power consumption abnormality judgment condition, thereby achieving the goal of reducing storage space while retaining important data.

[0004] For example, the invention patent with announcement number: CN116305685B announces a method, system, equipment and medium for drawing a low-voltage power distribution vertical system diagram. The method for drawing a low-voltage power distribution vertical system diagram includes: obtaining the layout information of the low-voltage distribution box input by the user; forming a data group of a layered distribution box according to the layout information of the low-voltage distribution box; drawing a low-voltage distribution model in the view according to the data group of the layered distribution box; generating a DWG drawing according to the low-voltage distribution model. The present invention can automatically draw a low-voltage power distribution vertical system diagram according to the layout information of the low-voltage distribution box, without consuming manpower for drawing, and can avoid omissions in manual drawing, which can effectively improve the work efficiency of designers and save time and effort.

[0005] The above disclosed technical solutions have at least the following technical problems: In the power distribution box, components such as circuit breakers, contactors, capacitors, etc. will age with time and load changes. During the aging process, the current, voltage, temperature and other parameters of these components may fluctuate or change slightly. These slight fluctuations usually need to be discovered through the analysis of historical data, and the role of component decomposition values ​​is particularly important in this process.

[0006] Low component decomposition values ​​allow the system to more sensitively detect minor anomalies that occur during the aging process of equipment. For example, when the contacts of a contactor gradually wear out, current fluctuations may become subtle. Through the low threshold decomposition method, these minor fluctuations can be captured and extracted into independent components, thereby diagnosing the aging of the equipment in advance. High-component decomposition values ​​are suitable for removing these weak fluctuations and retaining only more obvious abnormal signals, such as large current fluctuations or voltage fluctuations, which can help avoid too many false alarms in environments with small load fluctuations. In this way, the system can focus more on more serious fault signals, thereby improving the accuracy of fault judgment.

[0007] In the prior art, there is no unified standard for setting component decomposition values, and different threshold selections may lead to a balance problem between diagnostic sensitivity and false alarm rate, affecting the accuracy of aging diagnosis and maintenance decisions. In view of the above problems, the present invention proposes a solution. Summary of the invention

[0008] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a method and system for aging diagnosis and maintenance decision-making of internal components of an electric power distribution box, which solves the problem of lack of unified standards for setting component decomposition values ​​through the effects of different component decomposition values ​​in abnormal detection. Different threshold selections may lead to the balance problem between diagnostic sensitivity and false alarm rate, affecting the accuracy of aging diagnosis and maintenance decisions.

[0009] To achieve the above object, the present invention provides the following technical solutions: The invention discloses a maintenance decision method and system for aging diagnosis of components inside a power distribution box, comprising the following steps: obtaining operation data of components inside a power distribution box and preprocessing the operation data; performing data analysis on the distribution characteristics of the operation data to obtain an initial component decomposition value; performing random perturbations on the initial component decomposition value based on computer random numbers to obtain a number of different component decomposition values; applying a number of different component decomposition values ​​to decompose the operation data based on EMD to obtain a number of groups of corresponding IMFs, judging the components that deviate from the normal operating state for each group of IMFs based on similarity analysis, and then performing component aging diagnosis; obtaining equipment aging recognition data and abnormal detection sensitive data in the abnormal data detection process based on EMD; constructing a component decomposition value evaluation model based on the equipment aging recognition data and the abnormal detection sensitive data based on a bp neural network; establishing a component decomposition value-effect curve based on the output of the component decomposition value evaluation model and the corresponding component decomposition value, and performing curve analysis on the component decomposition value-effect curve to obtain an improved component decomposition value for component aging diagnosis.

[0010] In a preferred embodiment, the data analysis of the distribution characteristics of the operating data is performed to obtain the initial component decomposition value, specifically: a time series graph is drawn based on the operating data to analyze the trend of the operating data over time; based on the calculation of basic statistics of the operating data, the basic statistics include mean, variance, standard deviation, maximum value, minimum value, skewness, and kurtosis; the basic statistics are converted from time domain data to frequency domain data through fast Fourier transform, the spectrum characteristics are analyzed, and it is determined whether there is periodic fluctuation in the operating data; if there is periodic fluctuation, the result of FFT is used as the initial component decomposition value; if there is no periodic fluctuation, the initial component decomposition value is set according to historical experience.

[0011] In a preferred embodiment, the initial component decomposition value is randomly perturbed based on computer random numbers to obtain several different component decomposition values, specifically: step 1, using a computer random number generator to generate several random numbers; step 2, adding the generated random numbers to the initial component decomposition values ​​respectively; step 3, repeating step 2 several times to generate several different component decomposition values.

[0012] In a preferred embodiment, the components that deviate from the normal operating state are determined based on similarity analysis for each group of IMFs, and then component aging diagnosis is performed, specifically: a reference IMF component of the normal operating state is obtained; for each group of IMF components to be analyzed, a deviation pattern is evaluated by calculating the similarity with the reference IMF component, and the deviation pattern includes frequency offset, amplitude change, and waveform distortion; and an aging diagnosis decision is made according to the deviation pattern.

[0013] In a preferred embodiment, the equipment aging identification data includes the accuracy fluctuation coefficient of equipment aging pattern identification and the cost-effectiveness impact coefficient of maintenance decision; the specific method for obtaining the accuracy fluctuation coefficient of equipment aging pattern identification is as follows: obtaining the accuracy of aging pattern identification corresponding to the same time interval within a preset time period when performing component aging diagnosis; and marking the accuracy of aging pattern identification corresponding to the same time interval within a preset time period when performing component aging diagnosis; calculating the aging pattern identification standard deviation and the aging pattern identification average value; calculating the aging pattern identification variation coefficient based on the aging pattern identification standard deviation and the aging pattern identification average value; calculating the accuracy fluctuation coefficient of equipment aging pattern identification based on the aging pattern identification variation coefficient.

[0014] In a preferred embodiment, the specific method for obtaining the maintenance decision cost-benefit impact coefficient is as follows: obtain the load change of the part aging diagnosis maintenance decision, the equipment operating status, and the equipment aging status, and construct an aging status evaluation model through diagnostic software; cluster the aging status evaluation model; perform feature extraction on the processed aging status evaluation model, and calculate the maintenance priority evaluation coefficient based on the equipment aging trend, aging rate and failure mode; simulate the priority of the part aging diagnosis maintenance decision through the aging status evaluation model, and establish a maintenance timeliness model to obtain the maintenance timeliness evaluation coefficient; calculate the maintenance decision cost-benefit impact coefficient in combination with the maintenance priority evaluation coefficient.

[0015] In a preferred embodiment, the abnormal detection sensitivity data includes an abnormal detection sensitivity influence coefficient; the specific method for obtaining the abnormal detection sensitivity influence coefficient is as follows: obtain the abnormal signal response intensity at several time points when the equipment is running; the sensitivity standard deviation and sensitivity average of the abnormal signal response intensity at the same time interval when the equipment is running; calculate the coefficient of variation of the abnormal signal response intensity at the time point based on the sensitivity standard deviation and the sensitivity average; calculate the abnormal detection sensitivity influence coefficient based on the coefficient of variation of the abnormal signal response intensity at the time point based on a preset abnormal detection sensitivity influence coefficient calculation formula.

[0016] In a preferred embodiment, the output of the component decomposition value evaluation model and the corresponding component decomposition value are used to establish a component decomposition value-effect curve, specifically: the component decomposition value evaluation coefficients of the output under different component decomposition values ​​are obtained from the component decomposition value evaluation model; the X-axis is set as the component decomposition value, and the Y-axis is represented as the component decomposition value evaluation coefficient to establish a rectangular coordinate system; the component decomposition value evaluation coefficients corresponding to each group of component decomposition values ​​are marked on the rectangular coordinate system to form marked points, and adjacent marked points are connected to obtain the component decomposition value-effect curve.

[0017] In a preferred embodiment, the component decomposition value-effect curve is subjected to curve analysis to obtain an improved component decomposition value for application in component aging diagnosis, specifically: several peak values ​​of the component decomposition value-effect curve are obtained, several peak values ​​are integrated respectively, and the component decomposition value with the minimum integral result is taken as the improved component decomposition value, and applied to component aging diagnosis.

[0018] Technical effects and advantages of the method and system for aging diagnosis and maintenance decision-making of internal components of power distribution boxes of the present invention: 1. The present invention can effectively identify the aging mode of the equipment and optimize the maintenance decision by comprehensively analyzing and processing the operating data of the power distribution box components and combining the empirical mode decomposition (EMD) technology. First, by preprocessing the current, voltage, power, temperature and other data, extracting the key operating characteristics, and then combining the fast Fourier transform (FFT) to judge the periodic fluctuations, accurately identify the operating trend and fluctuation characteristics of the equipment. Then, by introducing random disturbances to generate multiple component decomposition values, multiple decompositions are performed using EMD to accurately extract the abnormal signals of equipment aging, and the deviation from the normal state is judged based on similarity analysis. Further, by calculating the accuracy fluctuation coefficient and the maintenance decision cost-effectiveness influence coefficient, the balance between diagnostic sensitivity and false alarm rate is optimized, the accuracy of aging pattern recognition is improved, misdiagnosis or missed diagnosis is reduced, and a scientific basis is provided for maintenance decisions. This method can improve the accuracy of equipment aging diagnosis, optimize maintenance timing and strategies, reduce equipment downtime and maintenance costs, and improve the operating efficiency and reliability of the distribution system.

[0019] 2. The present invention achieves high-precision diagnosis of aging of power distribution box components by constructing a component decomposition value evaluation model based on BP neural network and combining equipment aging pattern recognition data with abnormal detection sensitivity data. Specifically, firstly, by establishing a component decomposition value evaluation model of accuracy fluctuation coefficient, maintenance decision cost-effectiveness influence coefficient and abnormal detection sensitivity influence coefficient, a component decomposition value evaluation coefficient is generated, and combined with the component decomposition value-effect curve for analysis, the application of component decomposition value is optimized, thereby improving the sensitivity of aging pattern recognition and diagnostic accuracy. Through the curve analysis of the component decomposition value-effect curve, the component decomposition value is further improved, misdiagnosis or missed diagnosis is reduced, and the false alarm rate and missed alarm rate are reduced. The method can accurately identify the aging status of components in a shorter time, optimize maintenance decisions and timing, improve the reliability of the distribution system, reduce equipment downtime and maintenance costs, and thus significantly improve the operating efficiency and safety of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 It is a structural schematic diagram of the method for diagnosing and maintaining the aging of components inside a power distribution box according to the present invention. DETAILED DESCRIPTION

[0021] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0022] Embodiment 1, Figure 1 The present invention provides a method for diagnosing and repairing the aging of components inside a power distribution box, comprising the following steps: S1, obtaining the operating data of the internal components of the power distribution box and preprocessing the operating data.

[0023] The obtaining of the operating data of the internal components of the power distribution box and preprocessing of the operating data are specifically as follows: The operation data includes current, voltage, power, and temperature, forming an operation load sequence; The preprocessing includes removing abnormal data, smoothing and standardization.

[0024] It should be noted that the above steps have mature technologies and will not be repeated here.

[0025] S2, performing data analysis on the distribution characteristics of the operating data to obtain initial component decomposition values.

[0026] The data analysis of the distribution characteristics of the operating data is performed to obtain the initial component decomposition value, which is specifically: Draw a time series graph based on the operating data and analyze the trend of operating data over time; Based on the calculation of basic statistics of the running data, the basic statistics include mean, variance, standard deviation, maximum value, minimum value, skewness, and kurtosis; The basic statistics are converted from time domain data to frequency domain data through fast Fourier transform, and the spectrum characteristics are analyzed to determine whether there are periodic fluctuations in the operating data; If there are periodic fluctuations, the FFT result is used as the initial component decomposition value; If there are no cyclical fluctuations, the initial component decomposition values ​​are set based on historical experience.

[0027] It should be noted that by comprehensively using time series graphs, basic statistics and fast Fourier transform (FFT) to analyze the time domain and frequency domain characteristics of operating data, the operating status of the internal components of the power distribution box can be accurately judged. By analyzing the trend, volatility, periodicity and other characteristics of the data, the appropriate initial component decomposition value is automatically determined, thereby providing an effective decomposition basis for the subsequent empirical mode decomposition (EMD). The technical effect of this method is that it can flexibly adjust the decomposition parameters according to the different characteristics of the actual operating data, thereby improving the accuracy and adaptability of the decomposition. Through the periodic judgment of FFT, data with periodic fluctuations can be accurately decomposed, while for non-periodic data, reasonable settings rely on historical experience, ensuring the effectiveness of the method under different operating conditions, and thus providing reliable data support for the aging diagnosis of power distribution box components. The existing technology has mature methods, which will not be repeated here.

[0028] S3, randomly perturbing the initial component decomposition value based on computer random numbers to obtain a number of different component decomposition values.

[0029] The initial component decomposition value is randomly disturbed based on the computer random number to obtain several different component decomposition values, specifically: Step 1, using a computer random number generator to generate a number of random numbers; Step 2, adding the generated random numbers to the initial component decomposition values ​​respectively; Step 3, repeat step 2 several times to generate several different component decomposition values.

[0030] S4, based on EMD, several different component decomposition values ​​are applied to decompose the operation data to obtain several groups of corresponding IMFs. For each group of IMFs, the components that deviate from the normal operating state are determined based on similarity analysis, and then component aging diagnosis is performed.

[0031] The component that deviates from the normal operating state is determined based on similarity analysis for each group of IMFs, and then component aging diagnosis is performed, specifically: Obtain reference IMF components for normal operation; For each group of IMF components to be analyzed, the deviation pattern is evaluated by calculating the similarity between the components and the reference IMF components, and the deviation pattern includes frequency offset, amplitude change and waveform distortion; Making aging diagnostic decisions based on deviation patterns.

[0032] It should be noted that frequency deviation may indicate an electrical fault in the equipment, such as poor contact or damage to electrical components; amplitude change may reflect abnormal equipment load or signs of component aging, such as aging of capacitors, inductors, and switching components; waveform distortion may indicate increased electrical noise or damage to electrical components; the similarity between the calculated and reference IMF components can be determined through a variety of methods such as cross-correlation analysis, cosine similarity, mean square error, etc. The existing technology has mature technologies and will not be repeated here.

[0033] S5, obtaining equipment aging identification data and abnormality detection sensitivity data in the abnormal data detection process based on EMD.

[0034] The equipment aging identification data includes the accuracy fluctuation coefficient of equipment aging pattern identification and the maintenance decision cost-effectiveness influence coefficient; the abnormality detection sensitivity data includes the abnormality detection sensitivity influence coefficient.

[0035] The accuracy fluctuation coefficient of equipment aging pattern recognition is a key parameter used to measure the accuracy stability of equipment aging pattern recognition results. This coefficient reflects whether the accuracy of equipment aging pattern recognition fluctuates under different component decomposition thresholds, especially in the detection process of early aging. Specifically, a high threshold may cause weak aging signals to be ignored, thereby affecting the recognition accuracy of early aging, and may even lead to misdiagnosis or delayed diagnosis; while a low threshold can increase the sensitivity to subtle aging signals, but may introduce false signals, leading to premature or frequent alarms. By evaluating this coefficient, the stability and accuracy of equipment aging pattern recognition under different threshold settings can be quantified, helping to optimize the aging detection process and ensure that the diagnosis results are both accurate and timely, thereby providing reliable data support for maintenance decisions.

[0036] Analyzing the accuracy fluctuation coefficient of equipment aging pattern recognition is helpful for judging the effect of part aging diagnosis, thereby solving the problem of lack of unified standards for setting component decomposition values. Different threshold selections may lead to the balance problem between diagnostic sensitivity and false alarm rate, affecting the accuracy of aging diagnosis and maintenance decisions. It has the following advantages: Optimize the balance between diagnostic sensitivity and false alarm rate: The accuracy fluctuation coefficient can help analyze the impact of different thresholds on diagnostic sensitivity and false alarm rate, ensuring that too many false alarms are avoided at high sensitivity, or reducing the risk of delayed diagnosis at high accuracy. By controlling the fluctuation coefficient, a balance between sensitivity and false alarm rate can be achieved, improving the accuracy of diagnostic results.

[0037] Provide a basis for threshold optimization: This coefficient provides an objective evaluation basis for the selection of component decomposition thresholds, and can quantify the impact of different thresholds on the diagnostic effect, thereby helping to optimize the threshold setting and ensuring that early signs of equipment aging can be more accurately identified in actual applications, reducing missed diagnoses or misdiagnoses caused by improper threshold selection.

[0038] Improve the accuracy of aging pattern recognition: By quantifying the fluctuation coefficient, the recognition accuracy of different equipment aging patterns can be analyzed to find the most suitable combination of component decomposition values ​​and thresholds, making the recognition of equipment aging patterns more accurate, especially in early fault recognition, reducing equipment damage or safety hazards caused by untimely diagnosis.

[0039] Enhance the scientific nature of maintenance decisions: By accurately evaluating the stability and accuracy of equipment aging pattern recognition, it can provide a reliable basis for maintenance decisions and ensure that maintenance plans can be adjusted according to the actual aging status of the equipment, thereby reducing maintenance costs, extending equipment life, and improving the overall operating efficiency of the distribution box.

[0040] Improve the adaptability and reliability of the system: The analysis of the accuracy fluctuation coefficient helps to reveal the diagnostic adaptability of the equipment under different operating conditions. In particular, when facing different equipment types and operating states, it can optimize the diagnostic parameters, make the system more robust, adapt to more practical scenarios, and avoid unstable system performance due to improper threshold setting.

[0041] In summary, the accuracy fluctuation coefficient of equipment aging pattern recognition not only provides a more scientific basis for the setting of component decomposition values, but also effectively solves the balance problem between diagnostic sensitivity and false alarm rate, ensuring the accuracy, timeliness and cost-effectiveness of equipment aging diagnosis and maintenance decisions.

[0042] The specific method for obtaining the accuracy fluctuation coefficient of the equipment aging pattern recognition is as follows: Obtaining the accuracy of aging pattern recognition corresponding to the same time interval within a preset time period when performing component aging diagnosis; and marking the accuracy of aging pattern recognition corresponding to the same time interval within a preset time period when performing component aging diagnosis; Calculate the standard deviation of aging pattern recognition and the mean value of aging pattern recognition; The coefficient of variation of aging pattern recognition is calculated based on the aging pattern recognition standard deviation and the aging pattern recognition mean; The accuracy fluctuation coefficient of equipment aging pattern recognition is calculated based on the aging pattern recognition variation coefficient.

[0043] The specific calculation formula of the aging pattern recognition standard deviation is as follows:

[0044] The specific calculation formula of the aging pattern recognition average value is as follows:

[0045] The specific calculation formula of the accuracy fluctuation coefficient of the equipment aging pattern recognition is as follows:

[0046] In the formula, is the standard deviation for aging pattern identification, is the mean value for aging pattern identification, is the number of pattern recognitions during component aging diagnosis, It is the label for pattern recognition in component aging diagnosis. Correct identification ratio for component aging diagnosis, is the number of aging modes, is the accuracy fluctuation coefficient of equipment aging pattern recognition.

[0047] The maintenance decision cost-effectiveness impact coefficient is a key parameter used to measure the balance between maintenance decision effectiveness and maintenance efficiency. This coefficient reflects the impact of different component decomposition thresholds on the accuracy and efficiency of maintenance decisions, especially in diagnosing early aging and implementing maintenance timing. Specifically, a high threshold may reduce false alarms and abnormal signals, thereby improving the accuracy of maintenance decisions, but may lead to early aging of some equipment not being discovered in time, thereby delaying maintenance opportunities, increasing equipment downtime and maintenance costs; while a low threshold can detect potential problems earlier, but may lead to premature warnings and frequent maintenance decisions, affecting maintenance efficiency and increasing unnecessary maintenance costs. By evaluating this coefficient, the accuracy and cost-effectiveness of maintenance decisions under different threshold settings can be quantified, helping to optimize maintenance strategies and ensure that maintenance decisions can both detect faults in a timely manner and reasonably control maintenance costs, thereby improving the economy and reliability of equipment operation.

[0048] Analyzing the cost-effectiveness impact coefficient of maintenance decision-making is helpful for judging the effect of part aging diagnosis, thereby solving the problem of lack of unified standards for setting component decomposition values. Different threshold selections may lead to the problem of balancing the diagnostic sensitivity and false alarm rate, affecting the accuracy of aging diagnosis and maintenance decisions. It has the following advantages: Optimize the accuracy and timeliness of maintenance decisions: The maintenance decision cost-effectiveness impact coefficient helps quantify the diagnostic accuracy and maintenance timeliness under different thresholds, ensuring that while improving diagnostic sensitivity, premature or frequent maintenance decisions caused by false alarms are avoided. By balancing diagnostic sensitivity and false alarm rate, the accuracy of maintenance decisions can be improved, ensuring that maintenance work can be carried out in a timely and effective manner.

[0049] Reduce maintenance costs and downtime: A high threshold may lead to missed diagnosis of early aging, delay maintenance, and increase equipment downtime and maintenance costs. Although a low threshold can catch potential problems in advance, it may increase maintenance frequency and unnecessary downtime. By evaluating the cost-effectiveness impact coefficient of maintenance decisions, the maintenance cost and downtime under different thresholds can be accurately evaluated, helping to select the most appropriate threshold, optimize maintenance strategies, and reduce overall maintenance costs.

[0050] Enhance the economy and rationality of maintenance decisions: By analyzing the cost and benefit of maintenance decisions under different threshold settings, it can help determine the timing of the best maintenance decision, ensure timely preventive maintenance before the equipment aging reaches a serious level, and avoid excessive or delayed maintenance. This coefficient helps balance the economy and effectiveness of maintenance, thereby achieving higher asset utilization.

[0051] Improve the optimization capability of maintenance plans: The maintenance decision cost-benefit impact coefficient provides an evaluation standard based on cost and benefit, helping decision makers make the most reasonable maintenance decisions when faced with different diagnostic sensitivities and false alarm rates. By evaluating the maintenance impact under different thresholds, maintenance plans can be continuously optimized to ensure that major failures can be avoided and maintenance resources can be reasonably allocated.

[0052] Enhance the adaptability and flexibility of the system: Through the comprehensive analysis of this coefficient, potential risks and opportunities in the equipment aging diagnosis process can be identified, so that the system can adjust the threshold settings according to the actual operating environment and equipment conditions, thereby improving the adaptability and flexibility of diagnosis and maintenance decisions.

[0053] In summary, the maintenance decision cost-effectiveness impact coefficient can not only solve the standard problem of setting component decomposition values, but also effectively balance the diagnostic sensitivity and false alarm rate under different thresholds, improve the accuracy of aging diagnosis and the economy of maintenance decisions, help optimize maintenance strategies, reduce maintenance costs, reduce equipment downtime, and improve the overall operating efficiency of equipment. The specific method for obtaining the maintenance decision cost-benefit influence coefficient is as follows: Obtain the load changes, equipment operation status, and equipment aging status of parts aging diagnosis and maintenance decisions, and build an aging status assessment model through diagnostic software; Perform clustering on the aging status assessment model; Extract features from the processed aging status assessment model, and calculate the maintenance priority assessment coefficient based on the equipment aging trend, aging rate and failure mode; The priority of maintenance decision-making for part aging diagnosis is simulated through the aging status assessment model, and a maintenance timeliness model is established to obtain the maintenance timeliness assessment coefficient; the maintenance decision cost-effectiveness impact coefficient is calculated in combination with the maintenance priority assessment coefficient.

[0054] The specific calculation formula of the maintenance priority evaluation coefficient is as follows:

[0055] The specific calculation formula of the maintenance timeliness evaluation coefficient is as follows:

[0056] The specific calculation formula of the maintenance decision cost-benefit impact coefficient is as follows:

[0057] In the formula, is the maintenance priority assessment coefficient, is the maintenance timeliness evaluation coefficient, is the cost-effectiveness coefficient of maintenance decision, is the deviation value, Cost range for repair decisions for part degradation diagnosis, The scope of repairability for diagnosing and repairing parts aging, For device responsiveness, is the aging rate, is the aging mode propagation model, is the maintenance resource consumption rate, is the equipment health level assessment value, This is the maintenance time adjustment item.

[0058] The anomaly detection sensitivity impact coefficient is used to measure the impact of the component decomposition threshold on the balance between anomaly detection sensitivity and false alarm rate. This coefficient reflects the system's sensitivity to equipment aging signals and abnormal fluctuations under different threshold settings. A high threshold helps reduce false alarms and avoid misjudging normal fluctuations as abnormalities, but may ignore small early aging signals, resulting in delayed diagnosis; while a low threshold can improve the detection capability of small anomalies and capture early aging signals in a timely manner, but it is easy to cause a high false alarm rate and affect the reliability of the system. By evaluating the anomaly detection sensitivity impact coefficient, we can more accurately balance sensitivity and false alarm rate, ensure diagnostic accuracy, and reduce unnecessary warnings and false alarms, thereby optimizing the maintenance decision-making process and improving the efficiency and economy of equipment management.

[0059] Analyzing the influence coefficient of abnormal detection sensitivity is helpful for judging the effect of part aging diagnosis, thereby solving the problem of lack of unified standard for setting component decomposition values. Different threshold selections may lead to the balance problem between diagnostic sensitivity and false alarm rate, affecting the accuracy of aging diagnosis and maintenance decision-making. It has the following advantages: Optimize the balance between diagnostic sensitivity and false alarm rate: The anomaly detection sensitivity impact coefficient can help decision makers identify the trade-off between diagnostic sensitivity and false alarm rate at different thresholds. Although a high threshold can reduce false alarms, it may miss early aging signals, while a low threshold can detect potential problems in advance but may cause too many false alarms. By quantifying this coefficient, the component decomposition threshold can be adjusted to make it more suitable for the actual operating status of the device, thereby improving the accuracy and timeliness of diagnosis.

[0060] Improve early fault diagnosis capabilities: By analyzing the sensitivity impact coefficient of abnormal detection, it can help identify the appropriate sensitivity threshold to ensure that the system can capture the early aging signals of the equipment in a timely manner. High sensitivity can detect problems as early as possible, thereby reducing the incidence of equipment failures, extending the service life of the equipment, and reducing maintenance costs. By optimizing the threshold setting, the diagnostic system can effectively improve the early fault diagnosis capabilities.

[0061] Reduce the risk of false alarms and improve the reliability of maintenance decisions: Too high a false alarm rate may lead to too frequent maintenance decisions, waste resources and reduce maintenance efficiency. By evaluating the impact coefficient of abnormal detection sensitivity, it is possible to avoid unnecessary false alarms while improving detection sensitivity. This makes maintenance decisions more accurate, avoids excessive maintenance or ignores potential risks, and thus improves the economy and efficiency of maintenance decisions.

[0062] Improve the accuracy and reliability of diagnosis: Analyzing the sensitivity influence coefficient of abnormal detection helps ensure that the setting of component decomposition values ​​can match the actual operating status of the equipment. By adjusting the sensitivity under different thresholds, the system can accurately identify the aging signals of the equipment while avoiding misdiagnosis and missed diagnosis, thereby improving the reliability and accuracy of the diagnosis results.

[0063] Improve maintenance strategies and reduce downtime: Optimize threshold selection and balance sensitivity and false alarm rate to more accurately identify potential risks of equipment failure and reduce unnecessary downtime and maintenance. Reasonable threshold settings can ensure that equipment problems are discovered and maintained as early as possible, thereby reducing downtime, avoiding long-term equipment shutdowns, and improving overall production efficiency.

[0064] In summary, the anomaly detection sensitivity influence coefficient can effectively help solve the problem of lack of unified standards for setting component decomposition values, balance the relationship between diagnostic sensitivity and false alarm rate, improve the accuracy of equipment aging diagnosis, optimize maintenance decisions, and reduce the risk of false alarms and missed diagnosis, thereby improving equipment maintenance efficiency and reducing operating costs.

[0065] The specific method for obtaining the abnormal detection sensitivity influence coefficient is as follows: Obtain the abnormal signal response strength at several time points when the device is running; The sensitivity standard deviation and sensitivity mean of the abnormal signal response intensity at the same time interval when the equipment is running; According to the sensitivity standard deviation and the sensitivity mean, the coefficient of variation of the abnormal signal response intensity at the time point is calculated; The abnormal signal response intensity variation coefficient at a time point is used to calculate the abnormal detection sensitivity influence coefficient based on a preset abnormal detection sensitivity influence coefficient calculation formula.

[0066] The specific calculation formula for the coefficient of variation of the abnormal signal response intensity at the time point is as follows:

[0067] The specific calculation formula of the abnormal detection sensitivity influence coefficient is as follows:

[0068] In the formula, is the coefficient of variation of the abnormal signal response intensity at the time point, is the abnormal signal response intensity at the jth time point, is the total number of moments collected when the device is running, and j is the moment number; is the influence coefficient of anomaly detection sensitivity.

[0069] This implementation can effectively identify the aging mode of equipment and optimize maintenance decisions by comprehensively analyzing and processing the operating data of power distribution box components and combining empirical mode decomposition (EMD) technology. First, by preprocessing data such as current, voltage, power, and temperature, key operating characteristics are extracted, and then combined with fast Fourier transform (FFT) to judge periodic fluctuations, the operating trend and fluctuation characteristics of the equipment are accurately identified. Then, by introducing random disturbances to generate multiple component decomposition values, multiple decompositions are performed using EMD to accurately extract abnormal signals of equipment aging, and deviations from normal states are judged based on similarity analysis. Further, by calculating the accuracy fluctuation coefficient and the maintenance decision cost-effectiveness impact coefficient, the balance between diagnostic sensitivity and false alarm rate is optimized, the accuracy of aging pattern recognition is improved, misdiagnosis or missed diagnosis is reduced, and a scientific basis for maintenance decisions is provided. This method can improve the accuracy of equipment aging diagnosis, optimize maintenance timing and strategies, reduce equipment downtime and maintenance costs, and improve the operating efficiency and reliability of the distribution system.

[0070] Embodiment 2, S6, constructs a component decomposition value evaluation model based on the bp neural network using the equipment aging identification data and the abnormality detection sensitive data.

[0071] The equipment aging identification data and abnormal detection sensitive data are based on the bp neural network to build a component decomposition value evaluation model, specifically: The obtained accuracy fluctuation coefficient of equipment aging pattern recognition, maintenance decision cost-benefit influence coefficient and abnormal detection sensitivity influence coefficient are used to establish a component decomposition value evaluation model to generate a component decomposition value evaluation coefficient; The calculation formula of the component decomposition value evaluation coefficient is as follows:

[0072] In the formula, is the component decomposition value evaluation coefficient, The preset proportionality coefficient for the accuracy fluctuation coefficient of equipment aging pattern recognition, The preset proportionality coefficient for the maintenance decision cost-effectiveness impact coefficient, is the preset proportional coefficient of the abnormal detection sensitivity influence coefficient, is the accuracy fluctuation coefficient of equipment aging pattern recognition, is the cost-effectiveness coefficient of maintenance decision, is the influence coefficient of anomaly detection sensitivity.

[0073] S7, establishing a component decomposition value-effect curve using the output of the component decomposition value evaluation model and the corresponding component decomposition value, and performing curve analysis on the component decomposition value-effect curve to obtain an improved component decomposition value for component aging diagnosis.

[0074] The output of the component decomposition value evaluation model and the corresponding component decomposition value are used to establish a component decomposition value-effect curve, specifically: Obtaining component decomposition value evaluation coefficients of outputs under different component decomposition values ​​from a component decomposition value evaluation model; The X-axis is set as the component decomposition value, and the Y-axis is represented as the component decomposition value evaluation coefficient, and a rectangular coordinate system is established; The component decomposition value evaluation coefficient corresponding to each group of component decomposition values ​​is marked on the rectangular coordinate system to form marking points, and adjacent marking points are connected to obtain the component decomposition value-effect curve.

[0075] The component decomposition value-effect curve is analyzed to obtain an improved component decomposition value for component aging diagnosis, specifically: Several peaks of the component decomposition value-effect curve are obtained, and the several peaks are integrated respectively. The minimum component decomposition value of the integration result is used as the improved component decomposition value, and is applied to component aging diagnosis.

[0076] This embodiment achieves high-precision diagnosis of aging of power distribution box components by constructing a component decomposition value evaluation model based on BP neural network and combining equipment aging pattern recognition data with abnormal detection sensitivity data. Specifically, firstly, by establishing a component decomposition value evaluation model of accuracy fluctuation coefficient, maintenance decision cost-effectiveness influence coefficient and abnormal detection sensitivity influence coefficient, a component decomposition value evaluation coefficient is generated, and combined with the component decomposition value-effect curve for analysis, the application of component decomposition value is optimized, thereby improving the sensitivity of aging pattern recognition and diagnostic accuracy. Through the curve analysis of the component decomposition value-effect curve, the component decomposition value is further improved, misdiagnosis or missed diagnosis is reduced, and the false alarm rate and missed alarm rate are reduced. This method can accurately identify the aging status of components in a shorter time, optimize maintenance decisions and timing, improve the reliability of the distribution system, reduce equipment downtime and maintenance costs, and thus significantly improve the operating efficiency and safety of the power system.

[0077] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0078] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product.

[0079] Those of ordinary skill in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design failure mode of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0080] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0081] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0082] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for aging diagnosis and maintenance decision-making of internal components of power distribution boxes, characterized in that: The steps include: Obtain the operating data of the internal components of the power distribution box and pre-process the operating data; Perform data analysis on the distribution characteristics of the operating data to obtain initial component decomposition values; The initial component decomposition value is randomly disturbed based on a computer random number to obtain a number of different component decomposition values; Based on EMD, several different component decomposition values ​​are applied to decompose the operation data to obtain several groups of corresponding IMFs. For each group of IMFs, the components that deviate from the normal operating state are determined based on similarity analysis, and then component aging diagnosis is performed; Obtain equipment aging identification data and abnormality detection sensitivity data in the abnormal data detection process based on EMD; The equipment aging identification data and abnormal detection sensitive data are based on the BP neural network to build a component decomposition value evaluation model; The output of the component decomposition value evaluation model and the corresponding component decomposition value are used to establish a component decomposition value-effect curve, and the component decomposition value-effect curve is analyzed to obtain the improved component decomposition value for component aging diagnosis.

2. The method for aging diagnosis and maintenance decision-making of internal components of a power distribution box according to claim 1 is characterized in that: The data analysis of the distribution characteristics of the operating data is performed to obtain the initial component decomposition value, which is specifically: Draw a time series graph based on the operating data and analyze the trend of operating data over time; Based on the calculation of basic statistics of the running data, the basic statistics include mean, variance, standard deviation, maximum value, minimum value, skewness, and kurtosis; The basic statistics are converted from time domain data to frequency domain data through fast Fourier transform, and the spectrum characteristics are analyzed to determine whether there are periodic fluctuations in the operating data; If there are periodic fluctuations, the FFT result is used as the initial component decomposition value; If there are no cyclical fluctuations, the initial component decomposition values ​​are set based on historical experience.

3. The method for aging diagnosis and maintenance decision-making of internal components of a power distribution box according to claim 2 is characterized in that: The initial component decomposition value is randomly disturbed based on the computer random number to obtain several different component decomposition values, specifically: Step 1, using a computer random number generator to generate a number of random numbers; Step 2, adding the generated random numbers to the initial component decomposition values ​​respectively; Step 3, repeat step 2 several times to generate several different component decomposition values.

4. The method for aging diagnosis and maintenance decision-making of internal components of a power distribution box according to claim 3 is characterized in that: The component that deviates from the normal operating state is determined based on similarity analysis for each group of IMFs, and then component aging diagnosis is performed, specifically: Obtain reference IMF components for normal operation; For each group of IMF components to be analyzed, the deviation pattern is evaluated by calculating the similarity between the components and the reference IMF components, and the deviation pattern includes frequency offset, amplitude change and waveform distortion; Making aging diagnostic decisions based on deviation patterns.

5. The method for aging diagnosis and maintenance decision-making of internal components of a power distribution box according to claim 4 is characterized in that: The equipment aging identification data includes the accuracy fluctuation coefficient of equipment aging pattern identification and the maintenance decision cost-benefit impact coefficient; the specific method for obtaining the accuracy fluctuation coefficient of equipment aging pattern identification is as follows: Obtaining the accuracy of aging pattern recognition corresponding to the same time interval within a preset time period when performing component aging diagnosis; and marking the accuracy of aging pattern recognition corresponding to the same time interval within a preset time period when performing component aging diagnosis; Calculate the standard deviation of aging pattern recognition and the mean value of aging pattern recognition; The coefficient of variation of aging pattern recognition is calculated based on the aging pattern recognition standard deviation and the aging pattern recognition mean; The accuracy fluctuation coefficient of equipment aging pattern recognition is calculated based on the aging pattern recognition variation coefficient.

6. The method for aging diagnosis and maintenance decision-making of internal components of a power distribution box according to claim 5, characterized in that: The specific method for obtaining the maintenance decision cost-benefit influence coefficient is as follows: Obtain the load changes, equipment operation status, and equipment aging status of parts aging diagnosis and maintenance decisions, and build an aging status assessment model through diagnostic software; Perform clustering on the aging status assessment model; Extract features from the processed aging status assessment model, and calculate the maintenance priority assessment coefficient based on the equipment aging trend, aging rate and failure mode; The priority of maintenance decision-making for part aging diagnosis is simulated through the aging status assessment model, and a maintenance timeliness model is established to obtain the maintenance timeliness assessment coefficient; the maintenance decision cost-effectiveness impact coefficient is calculated in combination with the maintenance priority assessment coefficient.

7. The method for aging diagnosis and maintenance decision-making of internal components of a power distribution box according to claim 6, characterized in that: The abnormality detection sensitivity data includes an abnormality detection sensitivity influence coefficient; the specific method for obtaining the abnormality detection sensitivity influence coefficient is as follows: Obtain the abnormal signal response strength at several time points when the device is running; The sensitivity standard deviation and sensitivity mean of the abnormal signal response intensity at the same time interval when the equipment is running; According to the sensitivity standard deviation and the sensitivity mean, the coefficient of variation of the abnormal signal response intensity at the time point is calculated; The abnormal signal response intensity variation coefficient at a time point is used to calculate the abnormal detection sensitivity influence coefficient based on a preset abnormal detection sensitivity influence coefficient calculation formula.

8. The method for aging diagnosis and maintenance decision-making of internal components of a power distribution box according to claim 7, characterized in that: The output of the component decomposition value evaluation model and the corresponding component decomposition value are used to establish a component decomposition value-effect curve, specifically: Obtaining component decomposition value evaluation coefficients of outputs under different component decomposition values ​​from a component decomposition value evaluation model; The X-axis is set as the component decomposition value, and the Y-axis is represented as the component decomposition value evaluation coefficient, and a rectangular coordinate system is established; The component decomposition value evaluation coefficient corresponding to each group of component decomposition values ​​is marked on the rectangular coordinate system to form marking points, and adjacent marking points are connected to obtain the component decomposition value-effect curve.

9. The method for aging diagnosis and maintenance decision-making of internal components of a power distribution box according to claim 8, characterized in that: The component decomposition value-effect curve is analyzed to obtain an improved component decomposition value for component aging diagnosis, specifically: Several peaks of the component decomposition value-effect curve are obtained, and the several peaks are integrated respectively. The component decomposition value with the minimum component decomposition value of the integration result is taken as the improved component decomposition value, and applied to the component aging diagnosis.

10. A system using the method for aging diagnosis and maintenance decision-making of internal components of a power distribution box according to any one of claims 1 to 9, characterized in that It includes a data acquisition module, an initial component decomposition value setting module, a to-be-screened component decomposition value acquisition module, a component aging diagnosis module, a data analysis module, a component decomposition value evaluation module, and an improvement module; A data acquisition module is used to acquire the operating data of the internal components of the power distribution box and pre-process the operating data; An initial component decomposition value setting module is used to perform data analysis on the distribution characteristics of the operating data to obtain an initial component decomposition value; The to-be-screened component decomposition value acquisition module is used to randomly perturb the initial component decomposition value based on a computer random number to obtain a number of different component decomposition values; The component aging diagnosis module is used to decompose the operation data by applying several different component decomposition values ​​based on EMD to obtain several groups of corresponding IMFs. For each group of IMFs, the components that deviate from the normal operating state are determined based on similarity analysis, and then the component aging diagnosis is performed; A data analysis module is used to obtain equipment aging identification data and abnormality detection sensitivity data in the abnormal data detection process based on EMD; The component decomposition value evaluation module is used to construct a component decomposition value evaluation model based on the BP neural network based on the equipment aging identification data and abnormal detection sensitive data; The improvement module is used to establish a component decomposition value-effect curve using the output of the component decomposition value evaluation model and the corresponding component decomposition value, and to perform curve analysis on the component decomposition value-effect curve to obtain an improved component decomposition value for application in component aging diagnosis.

Images