A method, device and system for online fault diagnosis of electrical equipment

Abstract

The present invention relates to the technical field of electrical equipment fault diagnosis, and specifically to an electrical equipment fault online diagnosis method, device and system, the method comprising: obtaining a suspected fault period and its first extreme point according to voltage time series data and temperature time series data, obtaining a front-end temperature control force index of the suspected fault period before the first extreme point according to usage degree data and temperature time series data, and obtaining a rear-end temperature control force index of the suspected fault period after the first extreme point according to temperature time series data, and then obtaining a first K value, assigning the first K value to each data in the suspected fault period, and assigning a second K value of a preset fixed value to each data in the non-suspected fault period; using a LOF algorithm to perform abnormality detection on the data in the suspected fault period and the non-suspected fault period, and then judging whether the electrical equipment is abnormal. Using the present invention, the detection of electrical equipment can be more accurate.

Description

Technical Field

[0001] The present invention relates to the technical field of electrical equipment fault diagnosis, and in particular to an electrical equipment fault online diagnosis method, device and system. Background Art

[0002] With the continuous advancement of industrial automation and intelligent manufacturing technology, the reliability and stability of electrical equipment have become the core elements to ensure the continuity and safety of the production process. However, due to equipment aging, wear, failure and changes in the external environment, electrical equipment faces a variety of potential abnormal risks during operation. At present, although the power system can track key indicators such as voltage and temperature of electrical equipment in real time through data acquisition modules, how to effectively use these data to accurately identify and locate equipment abnormalities is still a major problem.

[0003] Most existing monitoring methods rely on simple threshold settings or expert systems based on preset rules. These methods often fail to accurately reflect the true health status of the equipment when dealing with complex and changing operating environments and equipment status. At the same time, the intensity of equipment use and fluctuations in environmental conditions also have a significant impact on equipment operation, and these factors are often ignored in traditional anomaly detection methods.

[0004] When using the LOF (Local Outlier Factor) algorithm for anomaly detection, it is crucial to choose an appropriate K value (i.e., the number of neighboring points considered when calculating the local outlier factor). A smaller K value makes the algorithm highly sensitive to local density changes, which may cause noise points to be misjudged as outliers; while a larger K value makes the algorithm focus more on global density and may miss local anomalies. Therefore, it is necessary to choose an appropriate K value based on the specific characteristics of the data set to achieve accurate detection of outliers. Summary of the invention

[0005] In order to solve the technical problem that the fault diagnosis of electrical equipment is not accurate enough, the purpose of the present invention is to provide an online fault diagnosis method for electrical equipment. The technical solution adopted is as follows:

[0006] Obtaining usage data, voltage time series data, and temperature time series data of the electrical equipment to be tested;

[0007] Based on the voltage time series data and the temperature time series data, obtaining an abnormal performance value of each reference data point; based on the abnormal performance value of the reference data point, screening the reference data point to determine a suspected fault period;

[0008] Obtain a first extreme point of the suspected fault period, obtain a front-end temperature control force index of the suspected fault period before the first extreme point according to the usage degree data and the temperature time series data, and obtain a rear-end temperature control force index of the suspected fault period after the first extreme point according to the temperature time series data;

[0009] Obtaining a first K value for the suspected fault period according to the front-end temperature controllability index and the rear-end temperature controllability index, assigning the first K value to each data in the suspected fault period, and assigning a second K value of a preset fixed value to each data in the non-suspected fault period;

[0010] The LOF algorithm is used to perform abnormality detection on the data in the suspected fault period and the non-suspected fault period, and whether the electrical equipment to be detected is abnormal is determined according to the abnormality detection result.

[0011] Furthermore, obtaining the abnormal performance value of each reference data point includes:

[0012] Adding the values ​​of the voltage time series data and the temperature time series data corresponding to each other in time series in sequence to obtain each reference data point;

[0013] Obtaining a value of the i-th reference data point and a variance of the reference time series, wherein the reference time series is composed of the reference data points, and the value of i ranges from 1 to the number of the reference data points;

[0014] The value of the i-th reference data point divided by the variance of the reference time series is equal to the abnormal performance value of the i-th reference data point;

[0015] The abnormal performance value acquisition process is repeated to obtain the abnormal performance value of each reference data point.

[0016] Furthermore, the process of obtaining the initial abnormal performance value includes:

[0017] Fitting the temperature time series data and the voltage time series data constituting the target period respectively to obtain a temperature fitting straight line and a voltage fitting straight line, and obtaining an absolute value of a temperature slope of the temperature fitting straight line and an absolute value of a voltage slope of the voltage fitting straight line;

[0018] A first absolute value of a difference between the absolute value of the temperature slope and the absolute value of the voltage slope is calculated, and an initial abnormal performance value is obtained according to the first absolute value and an average of the abnormal performance values ​​of each target data point in the target time period.

[0019] Further, obtaining an initial abnormal performance value according to the first absolute value and the average of the abnormal performance values ​​of each target data point in the target time period includes:

[0020] Obtaining a first mean of the abnormal performance values ​​of each target data point in the target period to which the pth target data point belongs, and dividing the first mean by a second mean of the abnormal performance values ​​of each target data point in the target period to which any other target data point belongs to obtain an abnormal performance ratio, wherein the value range of p is 1 to the number of the target data points;

[0021] Repeat the abnormal performance ratio acquisition process to obtain the abnormal performance ratio corresponding to the remaining target data points;

[0022] Adding the abnormal performance ratios in sequence to obtain a sum of the abnormal performance ratios, and multiplying the sum of the abnormal performance ratios by the first absolute value to obtain the initial abnormal performance value of the p-th target data point;

[0023] The process of obtaining the initial abnormal performance value is repeated to obtain the initial abnormal performance value of each target data point.

[0024] Furthermore, the process of obtaining the final abnormal performance value includes:

[0025] The final abnormal performance value of the target period to which the p-th target data point belongs is obtained by multiplying the initial abnormal performance value corresponding to the p-th target data point, the duration of the target period to which the p-th target data point belongs, and the abnormal performance value of the p-th target data point in sequence;

[0026] The process of obtaining the final abnormal performance value is repeated to obtain the final abnormal performance value corresponding to each target time period.

[0027] Furthermore, the process of obtaining the latter temperature controllability index includes:

[0028] Obtaining a temperature mean value of each of the suspected fault time periods, and clustering each of the suspected fault time periods according to the temperature mean value to obtain at least one cluster;

[0029] Obtaining the usage degree value of the electrical equipment to be detected, the first initial temperature control force index of the s-th suspected fault period, and the second initial temperature control force index of the remaining suspected fault periods belonging to the same cluster as the s-th suspected fault period, wherein the value range of s is 1 to the number of the suspected fault periods after the first extreme point, and calculating the usage degree value according to the usage degree data;

[0030] Calculating an average of the absolute values ​​of the differences between the first initial temperature controllability index and the second initial temperature controllability index;

[0031] After multiplying the first initial temperature controllability index by the average value, the index is divided by the usage degree value to obtain a final temperature controllability index of the latter part of the s-th suspected fault period;

[0032] The process of obtaining the rear-stage final temperature controllability index is repeated to obtain the rear-stage final temperature controllability index of each of the suspected fault time periods.

[0033] Furthermore, the process of obtaining the initial temperature controllability index includes:

[0034] Obtain the temperature values ​​corresponding to the extreme points in the s-th suspected fault period, and obtain the second absolute value of the difference between the temperature value of the h-th extreme point and the temperature value of the first extreme point, wherein the value range of h is 1 to the number of extreme points in the s-th suspected fault period;

[0035] The second absolute value is multiplied by the final abnormal performance value of the suspected fault period to obtain the temperature performance value of the hth extreme point;

[0036] Repeat the process of obtaining the temperature performance value to obtain the temperature performance value of each extreme point;

[0037] Fitting the data in the suspected fault period after the first extreme point and obtaining a fitting slope;

[0038] Obtaining the mean of the abnormal performance values ​​of the extreme points;

[0039] The initial temperature control force index is obtained by multiplying the mean value of the temperature performance value, the fitting slope and the mean value of the abnormal performance value of the extreme point in sequence and then performing normalization processing.

[0040] Furthermore, the process of obtaining the front-end temperature controllability index includes:

[0041] Obtaining the temperature rise rate of each data point in the suspected fault period before the first extreme value point;

[0042] The standard deviation of the heating rate is obtained, and the standard deviation is normalized to obtain the front-end temperature control force index.

[0043] The embodiment of the present invention further provides an online fault diagnosis device for electrical equipment, the device comprising:

[0044] A data acquisition module is used to acquire usage data, voltage time series data and temperature time series data of the electrical equipment to be tested;

[0045] A suspected fault period module is used to obtain an abnormal performance value of each reference data point based on the voltage time series data and the temperature time series data; based on the abnormal performance value of the reference data point, the reference data point is screened to determine a suspected fault period;

[0046] A temperature control force index module is used to obtain a first extreme point of the suspected fault period, obtain a front-end temperature control force index of the suspected fault period before the first extreme point according to the usage degree data and the temperature time series data, and obtain a rear-end temperature control force index of the suspected fault period after the first extreme point according to the temperature time series data;

[0047] A K value module, used to obtain a first K value of the suspected fault period according to the front-end temperature control force index and the rear-end temperature control force index, assign the first K value to each data in the suspected fault period, and assign a second K value of a preset fixed value to each data in the non-suspected fault period;

[0048] The abnormality detection module is used to use the LOF algorithm to perform abnormality detection on the data in the suspected fault period and the non-suspected fault period, and determine whether the electrical equipment to be detected is abnormal based on the abnormality detection result.

[0049] An embodiment of the present invention further provides an online fault diagnosis system for electrical equipment, the system comprising the above-mentioned online fault diagnosis device for electrical equipment.

[0050] The present invention has the following beneficial effects:

[0051] First, obtain the usage data, voltage timing data and temperature timing data of the electrical equipment to be detected. This is the data basis for subsequent data analysis. Secondly, obtain the abnormal performance value of each reference data point, obtain the suspected fault period, and through in-depth analysis of the suspected fault period, you can better understand the operating status of the electrical equipment, and promptly discover and solve potential problems. Furthermore, the front-end temperature control force index and the rear-end temperature control force index are both for obtaining the first K value of the suspected fault period. In addition, the first K value of the suspected fault period is obtained according to the front-end temperature control force index and the rear-end temperature control force index, and the first K value of the suspected fault period is assigned to each data in the suspected fault period, and the second K value of the preset fixed value is assigned to each data in the non-suspected fault period. The data obtained from the electrical equipment are assigned corresponding K values. Finally, the LOF algorithm is used to perform abnormal detection on the data in the suspected fault period and the non-suspected fault period, and then determine whether the electrical equipment to be detected is abnormal. The LOF algorithm is used to detect whether the electrical equipment to be detected is abnormal. The present invention applies an adaptive K value (i.e. assigning different K values ​​to different time periods) to the LOF algorithm, which can quantify the degree of abnormality of each data point. Therefore, the abnormality detection of electrical equipment is more accurate, which can help maintenance engineers quickly locate the fault point. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0053] Figure 1 A flowchart of an online fault diagnosis method for electrical equipment provided by the first embodiment of the present invention;

[0054] Figure 2 A flowchart of a process for obtaining an abnormal performance value provided by the second embodiment of the present invention;

[0055] Figure 3 A flowchart of a process for obtaining an initial abnormal performance value provided by the third embodiment of the present invention;

[0056] Figure 4 Another flow chart of the process of obtaining the initial abnormal performance value provided by the fourth embodiment of the present invention;

[0057] Figure 5 A flowchart of a process for obtaining a final abnormal performance value provided in a fifth embodiment of the present invention;

[0058] Figure 6 A flowchart of a process for obtaining a rear-stage temperature control force index provided by a sixth embodiment of the present invention;

[0059] Figure 7 A flowchart of a process for obtaining an initial temperature controllability index provided by a seventh embodiment of the present invention;

[0060] Figure 8 A flow chart of a process for obtaining a front-end temperature controllability index provided by an eighth embodiment of the present invention;

[0061] Fig. 9 This is a schematic diagram of an online fault diagnosis device for electrical equipment provided by a ninth embodiment of the present invention. DETAILED DESCRIPTION

[0062] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the method, device and system for online fault diagnosis of electrical equipment proposed by the present invention, its specific implementation, structure, features and effects in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0063] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0064] A specific scheme of an online fault diagnosis method for electrical equipment provided by the present invention is described in detail below with reference to the accompanying drawings.

[0065] See also Figure 1 , which shows a flow chart of an online fault diagnosis method for electrical equipment provided by an embodiment of the present invention, the method comprising:

[0066] S101. Obtain usage data, voltage time series data, and temperature time series data of the electrical equipment to be detected.

[0067] Obtain the voltage and temperature (device itself) time series data of the electrical equipment to be tested. Use the data acquisition module (in the power system, the data acquisition module can be connected to the control circuit of equipment such as transformers, generators or motors, and monitor the operating status of these electrical equipment in real time through voltage sensors, temperature sensors, etc., and upload the data to the monitoring system or background server) to monitor and record the voltage and temperature parameters of the electrical equipment in real time.

[0068] Voltage timing data: Voltage timing data monitoring helps ensure that electrical equipment operates within the normal operating range. Excessively high or low voltage may affect the performance and life of the equipment, or even cause equipment failure.

[0069] Temperature time series data: Electrical equipment generates heat during operation. If the heat dissipation is poor, excessive temperature may cause the equipment to overheat, leading to failure or safety accidents.

[0070] Usage data: collects the usage time and age of electrical equipment, which can be obtained through equipment nameplates and equipment files.

[0071] The collected raw data are cleaned and normalized to convert them into a unified format for subsequent processing and analysis.

[0072] S102. Based on the voltage time series data and the temperature time series data, obtain the abnormal performance value of each reference data point; based on the abnormal performance value of the reference data point, screen the reference data point to determine the suspected fault period.

[0073] The values ​​corresponding to the voltage time series data and the temperature time series data in time series are sequentially added to obtain each reference data point, and the abnormal performance value of each reference data point is obtained. The reference data point whose abnormal performance value is greater than a preset abnormal performance threshold is used as a target data point, and each of the target data points that are continuous in time constitutes a target time period. The initial abnormal performance value of the target data point is obtained, and the final abnormal performance value of each target time period is calculated according to the initial abnormal performance value, the abnormal performance value and the duration of the target time period. The target time period corresponding to the normalized value of the final abnormal performance value greater than the preset final abnormal performance threshold is marked as a suspected fault period.

[0074] In the same area, most electrical equipment is of the same type and specification, and their operating characteristics and data performance are often similar. In order to obtain the suspected fault period, the suspected fault period can be identified by segmenting the voltage time series data and the temperature time series data. During the suspected fault period, the degree of abnormality of the suspected fault period can be determined by further analysis of the use of the electrical equipment, temperature changes, and voltage changes.

[0075] At this point, the suspected fault periods of electrical equipment can be effectively identified and located. Through in-depth analysis of these periods, the operating status of electrical equipment can be better understood, and potential problems can be discovered and resolved in a timely manner.

[0076] The preset abnormal performance threshold can be set independently, and is preferably 1. The longer the duration of the target period is, the longer the abnormal state lasts, which may further indicate that the severity of the problem and its impact range are wider.

[0077] The preset final abnormal performance threshold can be set independently, and is preferably 0.7.

[0078] The process of obtaining the abnormal performance value will be described in detail in the second embodiment and will not be repeated here.

[0079] The process of obtaining the initial abnormal performance value will be described in detail in the third and fourth embodiments, and will not be repeated here.

[0080] The process of obtaining the final abnormal performance value will be described in detail in the fifth embodiment and will not be repeated here.

[0081] S103. Obtain the first extreme point of the suspected fault period, obtain the front temperature control force index of the suspected fault period before the first extreme point according to the usage level data and the temperature time series data, and obtain the rear temperature control force index of the suspected fault period after the first extreme point according to the temperature time series data.

[0082] The temperature of electrical equipment tends to gradually increase with the increase of usage time, eventually reaching a relatively stable state and fluctuating in this state. Abnormal temperature performance is usually caused by faults, which may be caused by factors such as internal faults of electrical equipment, load changes, unstable power supply or external interference. However, some abnormal temperature performance is caused by poor heat dissipation due to equipment aging. Aging is usually manifested as slow changes in temperature with a certain regularity, while faults are manifested as chaotic and drastic changes in temperature.

[0083] Therefore, in order to accurately determine the cause of the suspected fault period, it is necessary to further analyze the stable control ability of the temperature during these periods. This helps to distinguish the impact of faults and aging on equipment performance. The stable control ability of temperature can be reflected in the reaction of abnormal temperature, so it is necessary to further analyze the suspected fault period to characterize the stable control ability of temperature. The first extreme point of the entire suspected fault period often represents the performance of abnormal temperature in electrical equipment. Therefore, the suspected fault period is divided into two parts based on the first extreme point.

[0084] The total duration of all suspected fault periods can be used as an indicator to evaluate the health status of electrical equipment. The duration from the first maximum value to the last target data point in the target period can reflect the change in the control ability of the electrical equipment over a period of time. In the sth suspected fault period, the duration from the first extreme value point to the last target data point is obtained. , The larger the value is, the longer the electrical equipment has experienced an unstable or abnormal state during the sth suspected fault period. This may mean that the electrical equipment has not been able to return to normal for a long time, or its performance has degraded significantly. Therefore, the ability to stably control the temperature is worse.

[0085] When the heat dissipation capacity of electrical equipment is good, the surface temperature remains basically consistent. When the surface temperature of electrical equipment exceeds the ambient temperature, natural convection heat dissipation will be enhanced, which may cause the temperature rise rate to slow down. Therefore, the first extreme point in the entire suspected fault sequence of the electrical equipment represents the temperature value at which the electrical equipment initially reaches the state where heat dissipation is required.

[0086] The process of obtaining the latter temperature controllability index will be described in detail in the sixth and seventh embodiments and will not be repeated here.

[0087] The process of obtaining the front-end temperature controllability index will be described in detail in the eighth embodiment and will not be repeated here.

[0088] S104. Obtain the first K value of the suspected fault period according to the front-end temperature control force index and the rear-end temperature control force index, assign the first K value to each data in the suspected fault period, and assign a second K value of a preset fixed value to each data in the non-suspected fault period.

[0089] The first K value can be given by ) to obtain, wherein the is a normalization function, such as a range normalization function, , Q represents the front-end temperature control index, and v represents the back-end temperature control index. W reflects the temperature control ability of the electrical equipment during the suspected fault period. When W is small, it indicates that the temperature control of the electrical equipment is relatively stable. On the contrary, it indicates that the electrical equipment may have unstable temperature control or fault.

[0090] The second K value may be set to 1.

[0091] Measuring the abnormal performance of electrical equipment in a certain period of time with voltage is the main judgment indicator for monitoring. Time periods with a higher degree of abnormality mean that the density of data points in these time periods is significantly different from that of surrounding data points. Therefore, a smaller K value is required to capture this density change more finely. On the contrary, for normal or less abnormal time periods, a larger K value can be used to reduce the calculation complexity and maintain a certain detection accuracy.

[0092] S105. Use the LOF algorithm to perform anomaly detection on the data in the suspected fault period and the non-suspected fault period, and determine whether the electrical equipment to be detected has an abnormality based on the anomaly detection result.

[0093] The LOF algorithm is an existing technology, which can detect whether the data in the suspected fault period and the data in the non-suspected fault period are abnormal. If the data is abnormal, the electrical equipment to be detected is abnormal, and a maintenance engineer is required to further investigate the cause of the abnormality.

[0094] Figure 2 The flowchart of the process of obtaining the abnormal performance value provided by the second embodiment of the present invention, obtaining the abnormal performance value of each reference data point includes:

[0095] S201. Obtain the value of the i-th reference data point and the variance of the reference time series, wherein the reference time series is composed of the reference data points, and the value range of i is 1 to the number of the reference data points.

[0096] Since the device voltage is an important factor affecting the device temperature, the voltage time series data and the temperature time series data are added correspondingly in time series, and the temperature time series data scale is enlarged to obtain a reference time series. It is understandable that in order to avoid the influence of the dimension, before the voltage time series data and the temperature time series data are added, they need to be standardized to remove the influence of the dimension. The method of standardization is a well-known technology and will not be introduced in detail here.

[0097] The value of the i-th reference data point can be To express it, the variance of the reference time series can be expressed as To express.

[0098] S202. The value of the i-th reference data point divided by the variance of the reference time series equals the abnormal performance value of the i-th reference data point.

[0099] If the value of a reference data point is significantly outside the overall normal fluctuation range, this may indicate that the point is affected by an equipment anomaly.

[0100] The abnormal performance value of the i-th reference data point can be expressed as:

[0101] ;

[0102] Among them, the represents the abnormal performance value of the i-th reference data point.

[0103] If the abnormal performance value of a reference data point is significantly higher or lower than the abnormal performance values ​​of other reference data points, this indicates that the reference data point may be abnormal. If the abnormal performance value of a reference data point is much higher than 1, it means that the abnormal performance value of this reference data point is much greater than the average fluctuation level of the entire reference time series, which may be caused by equipment failure or other abnormal conditions.

[0104] S203. Repeat the abnormal performance value acquisition process to obtain the abnormal performance value of each reference data point.

[0105] The reference time series obtained above is an overall fluctuation of the temperature change of the electrical equipment. Short-term fluctuations are usually caused by external factors, such as instantaneous overload, voltage sag or transient interference. These fluctuations may be temporary and do not necessarily mean that there are long-term problems with the equipment. However, deviations under sustained periods of time mean that the anomaly may be caused by problems with the electrical equipment itself, such as failure, wear or performance degradation. Therefore, it is necessary to further analyze the temporal continuity of the reference data points corresponding to the larger abnormal performance values ​​and the change characteristics of the abnormal performance values ​​to screen out suspected fault periods.

[0106] The more abnormal the temperature performance of electrical equipment is during the same period, the greater the final abnormal performance value. Temperature abnormality is often caused by voltage abnormality, because voltage fluctuations may cause current instability, which in turn causes temperature changes inside the equipment. Therefore, the more severe the voltage fluctuation, the lower the stability and safety of the equipment operation, and the more significant the temperature abnormality will be.

[0107] Figure 3 This is a flow chart of a process for obtaining an initial abnormal performance value provided by a third embodiment of the present invention. The process for obtaining an initial abnormal performance value includes:

[0108] S301. Fit the temperature time series data and the voltage time series data constituting the target time period respectively to obtain a temperature fitting straight line and a voltage fitting straight line, and obtain the temperature slope absolute value of the temperature fitting straight line and the voltage slope absolute value of the voltage fitting straight line.

[0109] The fitting method may be the least squares method or the like.

[0110] S302. Calculate a first absolute value of the difference between the absolute value of the temperature slope and the absolute value of the voltage slope, and obtain an initial abnormal performance value according to the first absolute value and the average of the abnormal performance values ​​of each target data point in the target time period.

[0111] The first absolute value is the absolute value of the difference between the temperature slope absolute value and the voltage slope absolute value. The larger the first absolute value is, the more likely it is that there may be some systemic problem in the electrical equipment, causing abnormal performance.

[0112] Figure 4 Another flow chart of the process of obtaining the initial abnormal performance value provided by the fourth embodiment of the present invention, obtaining the initial abnormal performance value according to the first absolute value and the average of the abnormal performance values ​​of each of the target data points in the target period includes:

[0113] S401. Obtain a first mean of the abnormal performance values ​​of each target data point in the target time period to which the p-th target data point belongs, and divide the first mean by a second mean of the abnormal performance values ​​of each target data point in the target time period to which any remaining target data point belongs to obtain an abnormal performance ratio, wherein the value range of p is 1 to the number of target data points.

[0114] Said represents the first mean of the abnormal performance values ​​of the target period to which the p-th target data point belongs, The second mean of the abnormal performance values ​​of the target time period to which the yth target data point belongs is represented. The y is not equal to the p, and the value range of the y is 1 to Y-1, where the Y represents the number of the target data points.

[0115] S402. Repeat the abnormal performance ratio acquisition process to obtain the abnormal performance ratio corresponding to the remaining target data points.

[0116] The number of abnormal performance ratios is Y-1.

[0117] S403. Add the abnormal performance ratios in sequence to obtain the sum of the abnormal performance ratios, and multiply the sum of the abnormal performance ratios by the first absolute value to obtain the initial abnormal performance value of the p-th target data point.

[0118] The sum of the abnormal performance ratios can be expressed as: .

[0119] The initial abnormal performance value corresponding to the p-th target data point can be expressed as:

[0120] ;

[0121] Among them, h represents the initial abnormal performance value corresponding to the p-th target data point, and E represents the absolute value of the difference, that is, the first absolute value.

[0122] The larger the h is, the more significant the abnormal behavior of the electrical equipment in the target time period to which the p-th target data point belongs is.

[0123] S404. Repeat the process of obtaining the initial abnormal performance value to obtain the initial abnormal performance value of each target data point.

[0124] Figure 5 This is a flowchart of a process for obtaining a final abnormal performance value provided by a fifth embodiment of the present invention. The process for obtaining a final abnormal performance value includes:

[0125] S501. Multiply the initial abnormal performance value corresponding to the pth target data point, the duration of the target period to which the pth target data point belongs, and the abnormal performance value of the pth target data point in sequence to obtain the final abnormal performance value of the target period to which the pth target data point belongs.

[0126] The final abnormal performance value corresponding to the p-th target data point can be expressed as:

[0127] ;

[0128] Wherein, c represents the final abnormal performance value corresponding to the p-th target data point, represents the abnormal performance value of the p-th target data point, represents the initial abnormal performance value corresponding to the p-th target data point, Indicates the duration of the target period to which the p-th target data point belongs.

[0129] The larger the c is, the greater the possibility that the electrical device to be detected has an abnormality itself.

[0130] S502. Repeat the process of obtaining the final abnormal performance value to obtain the final abnormal performance value corresponding to each target time period.

[0131] Figure 6 This is a flow chart of a process for obtaining a rear-stage temperature controllability index provided by a sixth embodiment of the present invention. The process for obtaining a rear-stage temperature controllability index includes:

[0132] S601. Obtain a temperature mean value for each of the suspected fault time periods, and cluster each of the suspected fault time periods according to the temperature mean value to obtain at least one cluster.

[0133] The clustering method is K-Means algorithm. Here, the purpose is to further analyze the control performance of electrical equipment at the same temperature, that is, the difference between the sth suspected fault period and other suspected fault periods in the cluster to which the sth suspected fault period belongs. At the same temperature, the control effect of aging is similar, while the control effect of fault is chaotic.

[0134] S602. Obtain the usage level value of the electrical equipment to be detected, the first initial temperature control force index of the sth suspected fault period, and the second initial temperature control force index of the remaining suspected fault periods belonging to the same cluster as the sth suspected fault period, wherein the value range of s is 1 to the number of the suspected fault periods after the first extreme point, and the usage level value is calculated based on the usage level data.

[0135] The process of obtaining the initial temperature controllability index will be described in detail in the seventh embodiment and will not be repeated here.

[0136] S603. Calculate an average of the absolute values ​​of the differences between the first initial temperature controllability index and the second initial temperature controllability index.

[0137] The average value of the absolute values ​​of the differences can be expressed as: , wherein the represents the first initial temperature control force index of the sth suspected fault period, represents a second initial temperature controllability index of the rth suspected fault period, wherein R represents the number of the suspected fault period. The smaller the value, the more similar the temperature control capabilities are, that is, the more the abnormality is caused by equipment aging.

[0138] S604. Multiply the first initial temperature controllability index by the average value, and then divide the index by the usage level value to obtain a final temperature controllability index for the latter part of the sth suspected fault period.

[0139] The final temperature controllability index of the latter part of the sth suspected fault period can be expressed as:

[0140] ;

[0141] Among them, the represents the final temperature control force index of the latter part of the s-th suspected fault period, represents the first initial temperature control force index of the sth suspected fault period, and f represents the usage degree value.

[0142] The process of obtaining the usage degree value includes:

[0143] ;

[0144] Among them, the represents the usage time of the electrical equipment to be detected, G represents the service life of the electrical equipment to be detected, and f represents the usage degree value. The closer the value is to 1, the longer the usage time of the electrical equipment is, and the higher the degree of wear and aging is.

[0145] S605. Repeat the process of obtaining the latter-stage final temperature controllability index to obtain the latter-stage final temperature controllability index of each of the suspected fault time periods.

[0146] Figure 7 This is a flow chart of a process for obtaining an initial temperature controllability index provided by a seventh embodiment of the present invention. The process for obtaining an initial temperature controllability index includes:

[0147] S701. Obtain the temperature value corresponding to each extreme point in the sth suspected fault period, and obtain the second absolute value of the difference between the temperature value of the hth extreme point and the temperature value of the first extreme point, wherein the value range of h is 1 to the number of extreme points in the sth suspected fault section.

[0148] The absolute value of the difference between the temperature value of the hth extreme point and the temperature value of the first extreme point can be represented by u.

[0149] If the temperature value of the extreme point in the suspected fault period is not much different from the temperature value of the first extreme point, it indicates that the temperature control of the electrical equipment during the suspected fault period is relatively stable and there is no drastic fluctuation. On the contrary, if the temperature value is greatly different, it indicates that the electrical equipment may have unstable temperature control or failure during the suspected fault period.

[0150] S702. Multiply the second absolute value by the final abnormal performance value of the suspected fault period to obtain the temperature performance value of the hth extreme point.

[0151] The temperature performance value can be expressed as:

[0152] ;

[0153] Among them, m represents the temperature performance value of the hth extreme point, c represents the final abnormal performance value corresponding to the suspected fault period, and u represents the second absolute value.

[0154] S703. Repeat the temperature performance value acquisition process to obtain the temperature performance value of each extreme point.

[0155] The larger the mean value of the temperature performance value is, the more significant the temperature abnormality of the electrical equipment during the sth suspected fault period is.

[0156] S704. Fit the data in the suspected fault period after the first extreme point and obtain a fitting slope.

[0157] The fitted slope can be expressed as To indicate. The larger the value is, the faster the temperature rise rate of the electrical equipment during the suspected fault period is, which further indicates that the electrical equipment may have unstable temperature control or failure during the suspected fault period.

[0158] S705: Obtain the mean value of the abnormal performance value of the extreme point.

[0159] The mean of the abnormal performance values ​​of the extreme points can be To express.

[0160] S706. The average value of the temperature performance value, the fitting slope, and the average value of the abnormal performance value of the extreme point are multiplied in sequence and then normalized to obtain the initial temperature control force index.

[0161] The initial temperature control index can be expressed as:

[0162] ;

[0163] Wherein, T represents the initial temperature control force index, represents the mean of the abnormal performance values ​​of the extreme points, represents the mean value of the temperature performance value, is a normalized function, which may be a range normalized function. The larger the T is, the more serious the temperature control problem of the electrical equipment during the suspected fault period is.

[0164] Figure 8 This is a flow chart of a process for obtaining a front-end temperature controllability index provided by an eighth embodiment of the present invention. The process for obtaining a front-end temperature controllability index includes:

[0165] S801. Obtain the temperature rise rate of each data point in the suspected fault period before the first extreme point.

[0166] The heating rate can be expressed as:

[0167] ;

[0168] Wherein, the g represents the The heating rate corresponding to the target data point, Indicates the sth suspected fault period The temperature value of the target data point, Indicates the sth suspected fault period The temperature value of the target data point.

[0169] The heating rate corresponding to each target data point can be obtained by repeating the process of obtaining the heating rate.

[0170] S802. Obtain the standard deviation of the heating rate, and normalize the standard deviation to obtain the front-end temperature control force index.

[0171] The larger the front-end temperature control force index is, the greater the difference in heating rate at different time points is. This may be caused by unstable operation due to aging, wear or failure of internal components of the electrical equipment, and the greater the possibility of electrical equipment abnormality.

[0172] Fig. 9 A schematic diagram of an online fault diagnosis device for electrical equipment provided by a ninth embodiment of the present invention, the device comprising:

[0173] The data acquisition module 901 is used to acquire the usage data, voltage time series data and temperature time series data of the electrical equipment to be detected;

[0174] A suspected fault period module 902 is used to obtain an abnormal performance value of each reference data point based on the voltage time series data and the temperature time series data; and screen the reference data points to determine a suspected fault period based on the abnormal performance value of the reference data point;

[0175] The temperature control force index module 903 is used to obtain the first extreme point of the suspected fault period, obtain the temperature control force index of the front section of the suspected fault period before the first extreme point according to the usage degree data and the temperature time series data, and obtain the temperature control force index of the rear section of the suspected fault period after the first extreme point according to the temperature time series data;

[0176] A K value module 904 is used to obtain a first K value of the suspected fault period according to the front-end temperature control force index and the rear-end temperature control force index, assign the first K value to each data in the suspected fault period, and assign a second K value of a preset fixed value to each data in the non-suspected fault period;

[0177] The abnormality detection module 905 is used to use the LOF algorithm to perform abnormality detection on the data in the suspected fault period and the non-suspected fault period, and determine whether the electrical equipment to be detected has an abnormality according to the abnormality detection result.

[0178] The technical features and technical effects of an online fault diagnosis device for electrical equipment proposed in an embodiment of the present invention are the same as those of the method proposed in an embodiment of the present invention, and are not described in detail herein.

[0179] An embodiment of the present invention further provides an online fault diagnosis system for electrical equipment, the system comprising the above-mentioned online fault diagnosis device for electrical equipment.

[0180] The present invention has the following beneficial effects:

[0181] First, obtain the usage data, voltage timing data and temperature timing data of the electrical equipment to be tested. This is the data basis for subsequent data analysis. Secondly, add the corresponding values ​​of the voltage timing data and the temperature timing data in time sequence to obtain each reference data point, obtain the abnormal performance value of each reference data point, take the reference data point whose abnormal performance value is greater than the preset abnormal performance threshold as the target data point, and each target data point that is continuous in time constitutes a target period, obtain the initial abnormal performance value of the target data point, calculate the final abnormal performance value of each target period according to the initial abnormal performance value, the abnormal performance value and the duration of the target period, and mark the target period corresponding to the normalized value of the final abnormal performance value greater than the preset final abnormal performance threshold as a suspected fault period. Obtaining the suspected fault period and through in-depth analysis of the suspected fault period can better understand the operating status of the electrical equipment, and timely discover and solve potential problems. Furthermore, the first extreme point of the suspected fault period is obtained, and the front temperature control force index of the suspected fault period before the first extreme point is obtained according to the usage degree data and the temperature time series data, and the rear temperature control force index of the suspected fault period after the first extreme point is obtained according to the temperature time series data. The front temperature control force index and the rear temperature control force index are both for obtaining the first K value of the suspected fault period. In addition, the first K value of the suspected fault period is obtained according to the front temperature control force index and the rear temperature control force index, and the first K value is assigned to each data in the suspected fault period, and the second K value with a preset fixed value is assigned to each data in the non-suspected fault period. The data obtained from the electrical equipment are assigned corresponding K values. Finally, the LOF algorithm is used to perform abnormality detection on the data in the suspected fault period and the non-suspected fault period, and whether the electrical equipment to be detected is abnormal is determined according to the abnormality detection result. The LOF algorithm is a prior art and is used to detect whether the electrical equipment to be detected is abnormal. The present invention applies an adaptive K value (i.e. assigning different K values ​​to different time periods) to the LOF algorithm, which can quantify the degree of abnormality of each data point. Therefore, the abnormality detection of electrical equipment is more accurate, which can help maintenance engineers quickly locate the fault point.

[0182] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0183] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A method for online fault diagnosis of electrical equipment, characterized in that: The method comprises: Obtaining usage data, voltage time series data, and temperature time series data of the electrical equipment to be tested; Based on the voltage time series data and the temperature time series data, obtaining an abnormal performance value of each reference data point; based on the abnormal performance value of the reference data point, screening the reference data point to determine a suspected fault period; Obtain a first extreme point of the suspected fault period, obtain a front-end temperature control force index of the suspected fault period before the first extreme point according to the usage degree data and the temperature time series data, and obtain a rear-end temperature control force index of the suspected fault period after the first extreme point according to the temperature time series data; Obtaining a first K value for the suspected fault period according to the front-end temperature controllability index and the rear-end temperature controllability index, assigning the first K value to each data in the suspected fault period, and assigning a second K value of a preset fixed value to each data in the non-suspected fault period; The LOF algorithm is used to perform abnormality detection on the data in the suspected fault period and the non-suspected fault period, and whether the electrical equipment to be detected is abnormal is determined according to the abnormality detection result.

2. The method for online fault diagnosis of electrical equipment according to claim 1, characterized in that: Obtaining the abnormal performance value of each reference data point includes: Adding the values ​​of the voltage time series data and the temperature time series data corresponding to each other in time series in sequence to obtain each reference data point; Obtaining the value of the i-th reference data point and the variance of the reference time series, wherein the reference time series is composed of the reference data points, and the value range of i is 1 to the number of the reference data points; The value of the i-th reference data point divided by the variance of the reference time series is equal to the abnormal performance value of the i-th reference data point; The abnormal performance value acquisition process is repeated to obtain the abnormal performance value of each reference data point.

3. The method for online fault diagnosis of electrical equipment according to claim 1, characterized in that: The process of obtaining the initial abnormal performance value includes: Fitting the temperature time series data and the voltage time series data constituting the target period respectively to obtain a temperature fitting straight line and a voltage fitting straight line, and obtaining an absolute value of a temperature slope of the temperature fitting straight line and an absolute value of a voltage slope of the voltage fitting straight line; A first absolute value of a difference between the absolute value of the temperature slope and the absolute value of the voltage slope is calculated, and an initial abnormal performance value is obtained according to the first absolute value and an average of the abnormal performance values ​​of each target data point in the target time period.

4. The method for online fault diagnosis of electrical equipment according to claim 3, characterized in that: Obtaining an initial abnormal performance value according to the first absolute value and the average of the abnormal performance values ​​of each of the target data points in the target period includes: Obtaining a first mean of the abnormal performance values ​​of each target data point in the target period to which the pth target data point belongs, and dividing the first mean by a second mean of the abnormal performance values ​​of each target data point in the target period to which any other target data point belongs to obtain an abnormal performance ratio, wherein the value range of p is 1 to the number of the target data points; Repeat the abnormal performance ratio acquisition process to obtain the abnormal performance ratio corresponding to the remaining target data points; Adding the abnormal performance ratios in sequence to obtain a sum of the abnormal performance ratios, and multiplying the sum of the abnormal performance ratios by the first absolute value to obtain the initial abnormal performance value of the p-th target data point; The process of obtaining the initial abnormal performance value is repeated to obtain the initial abnormal performance value of each target data point.

5. The method for online fault diagnosis of electrical equipment according to claim 4, characterized in that: The process of obtaining the final abnormal performance value includes: The final abnormal performance value of the target period to which the p-th target data point belongs is obtained by multiplying the initial abnormal performance value corresponding to the p-th target data point, the duration of the target period to which the p-th target data point belongs, and the abnormal performance value of the p-th target data point in sequence; The process of obtaining the final abnormal performance value is repeated to obtain the final abnormal performance value corresponding to each target time period.

6. The method for online fault diagnosis of electrical equipment according to claim 5, characterized in that: The process of obtaining the latter temperature controllability index includes: Obtaining a temperature mean value of each of the suspected fault time periods, and clustering each of the suspected fault time periods according to the temperature mean value to obtain at least one cluster; Obtaining the usage degree value of the electrical equipment to be detected, the first initial temperature control force index of the s-th suspected fault period, and the second initial temperature control force index of the remaining suspected fault periods belonging to the same cluster as the s-th suspected fault period, wherein the value range of s is 1 to the number of the suspected fault periods after the first extreme point, and calculating the usage degree value according to the usage degree data; Calculating an average of the absolute values ​​of the differences between the first initial temperature controllability index and the second initial temperature controllability index; After multiplying the first initial temperature controllability index by the average value, the index is divided by the usage degree value to obtain a final temperature controllability index of the latter part of the s-th suspected fault period; The process of obtaining the rear-stage final temperature controllability index is repeated to obtain the rear-stage final temperature controllability index of each of the suspected fault time periods.

7. The method for online fault diagnosis of electrical equipment according to claim 6, characterized in that: The process of obtaining the initial temperature control force index includes: Obtain the temperature values ​​corresponding to the extreme points in the s-th suspected fault period, and obtain the second absolute value of the difference between the temperature value of the h-th extreme point and the temperature value of the first extreme point, wherein the value range of h is 1 to the number of extreme points in the s-th suspected fault period; The second absolute value is multiplied by the final abnormal performance value of the suspected fault period to obtain the temperature performance value of the hth extreme point; Repeat the process of obtaining the temperature performance value to obtain the temperature performance value of each extreme point; Fitting the data in the suspected fault period after the first extreme point and obtaining a fitting slope; Obtaining the mean of the abnormal performance values ​​of the extreme points; The initial temperature control force index is obtained by multiplying the mean value of the temperature performance value, the fitting slope and the mean value of the abnormal performance value of the extreme point in sequence and then performing normalization processing.

8. The method for online fault diagnosis of electrical equipment according to claim 1, characterized in that: The process of obtaining the front-end temperature control force index includes: Obtaining the temperature rise rate of each data point in the suspected fault period before the first extreme value point; The standard deviation of the heating rate is obtained, and the standard deviation is normalized to obtain the front-end temperature control force index.

9. An online fault diagnosis device for electrical equipment, characterized in that: The device comprises: A data acquisition module is used to acquire usage data, voltage time series data and temperature time series data of the electrical equipment to be tested; A suspected fault period module is used to obtain an abnormal performance value of each reference data point based on the voltage time series data and the temperature time series data; based on the abnormal performance value of the reference data point, the reference data point is screened to determine a suspected fault period; A temperature control force index module is used to obtain a first extreme point of the suspected fault period, obtain a front-end temperature control force index of the suspected fault period before the first extreme point according to the usage degree data and the temperature time series data, and obtain a rear-end temperature control force index of the suspected fault period after the first extreme point according to the temperature time series data; A K value module, used to obtain a first K value of the suspected fault period according to the front-end temperature control force index and the rear-end temperature control force index, assign the first K value to each data in the suspected fault period, and assign a second K value of a preset fixed value to each data in the non-suspected fault period; The abnormality detection module is used to use the LOF algorithm to perform abnormality detection on the data in the suspected fault period and the non-suspected fault period, and determine whether the electrical equipment to be detected is abnormal based on the abnormality detection result.

10. An online fault diagnosis system for electrical equipment, characterized in that: The system comprises the online fault diagnosis device for electrical equipment as claimed in claim 9.

Images