Power transformation equipment fault prediction method based on big data

Abstract

The invention relates to the technical field of data processing, in particular to a power transformation equipment fault prediction method based on big data, and the method comprises the steps: obtaining current data, load data and dissolved gas concentration of power transformation equipment in a plurality of periods; determining the abnormal degree of the current period; clustering the load data and the dissolved gas concentration based on an association relationship between the load data and the dissolved gas concentration to obtain a plurality of clusters; correcting the abnormal degree of each period according to the load data and the dissolved gas concentration in each cluster to obtain the corrected final abnormal degree of each period, and selecting the period of which the final abnormal degree is greater than a first threshold value as an abnormal fluctuation period; determining the possibility of a fault occurring in the abnormal fluctuation period; and predicting whether the power transformation equipment has a potential fault according to the possibility of the fault in the abnormal fluctuation period. Therefore, the power transformation equipment fault prediction accuracy and reliability are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a method for predicting failures of substation equipment based on big data. Background Art

[0002] In the process of building the power system, transformers in the power grid, as key equipment, play an irreplaceable role. Transformation equipment such as large main transformers and medium and low voltage distribution transformers are widely distributed in various power supply and distribution scenarios, especially in complex distribution networks with high penetration of renewable energy. The stability and reliability of their operation are directly related to the safety of the entire power system.

[0003] In some scenarios, as the operation time of substation equipment accumulates, the internal insulation layer and insulating oil will inevitably be damaged and decomposed, which can easily cause partial discharge inside the transformer. Partial discharge will not only further aggravate the damage of the insulating material, but may also cause transformer failure, affecting the continuity and stability of power supply. At present, the current of substation equipment is often monitored to predict the potential initial failure of substation equipment. However, there are many abnormalities in the potential initial failure of substation equipment. It is difficult to effectively monitor and accurately predict the initial potential failure of substation equipment based on current monitoring alone, resulting in low accuracy and reliability of substation equipment failure prediction. Summary of the invention

[0004] In order to solve the technical problem of low accuracy and reliability in fault prediction of substation equipment, the purpose of the present invention is to provide a method for fault prediction of substation equipment based on big data. The technical solution adopted is as follows: In a first aspect, an embodiment of the present invention provides a method for predicting faults of substation equipment based on big data, including: obtaining current data, load data and dissolved gas concentration of the substation equipment in multiple cycles; determining the degree of abnormality of the current cycle according to the amplitude of the maximum point of the current data in the current cycle, the first number of the maximum points of the current data in the current cycle and the degree of abnormality of the previous cycle adjacent to the current cycle; clustering the load data and the dissolved gas concentration based on the correlation between the load data and the dissolved gas concentration to obtain multiple clustering clusters; correcting the degree of abnormality of each cycle according to the load data and the dissolved gas concentration in each clustering cluster to obtain the final corrected degree of abnormality of each cycle, and selecting the cycle with the final degree of abnormality greater than the first threshold as the abnormal fluctuation cycle; determining the possibility of failure in the abnormal fluctuation cycle according to the amplitude of each reference cycle of the abnormal fluctuation cycle and the time interval between adjacent reference cycles; predicting whether the substation equipment has potential faults according to the possibility of failure in the abnormal fluctuation cycle.

[0005] Optionally, determining the abnormality level of the current cycle based on the amplitude of the maximum points of the current data in the current cycle, the first number of the maximum points of the current data in the current cycle, and the abnormality level of the previous cycle adjacent to the current cycle includes: calculating a first difference between the first number and a predetermined value, and a second difference between the predetermined value and the abnormality level of the adjacent previous cycle; calculating a third difference between the amplitude of each maximum point of the current data in the current cycle and the average of the amplitudes of all maximum points of the current data in the adjacent previous cycle, and superimposing each third difference to obtain a first superimposed value; calculating a first product between the first difference and the second difference and a second product between the first superimposed value and the abnormality level of the adjacent previous cycle; and determining a first sum between the first product and the second product as the abnormality level of the current cycle.

[0006] Optionally, the abnormality degree of each period is corrected according to the load data and dissolved gas concentration in each cluster to obtain the corrected final abnormality degree of each period, including: sorting the load data and dissolved gas concentration in each cluster in chronological order and then segmenting them to obtain multiple data segments; determining the temporal continuity of the cluster according to the second number of data segments in the cluster, the third number of data points in the cluster, the fourth number of data points in the data segment, and the change between every two adjacent data points in the data segment; determining the temporal distribution degree of the cluster according to the time interval between every two adjacent data segments in the cluster and the second number; correcting the abnormality degree of each period according to the temporal continuity and temporal distribution degree of the cluster to obtain the corrected final abnormality degree of each period.

[0007] Optionally, the load data and dissolved gas concentration in each cluster are sorted in chronological order and then segmented to obtain multiple data segments, including: determining the time interval between the load data and dissolved gas concentration at the current time point in the cluster and the load data and dissolved gas concentration at the adjacent previous time point; when the time interval is greater than a second threshold, the load data and dissolved gas concentration at the current time point are used as the starting point of a new segment of data points; when the time interval is not greater than the second threshold, the load data and dissolved gas concentration at the current time point are classified into the data segment where the previous adjacent load data is located.

[0008] Optionally, determining the temporal continuity of the clusters based on the second number of data segments in the clusters, the third number of data points in the clusters, the fourth number of data points in the data segments, and the amount of change between every two adjacent data points in the data segments includes: calculating a first ratio between the second number and the third number and a second ratio between the fourth number and the third number; calculating an average change in the amount of change between every two adjacent data points in the data segment and a variance of the amount of change between every two adjacent data points in the data segment; calculating a third product between the second ratio, the inverse of the average change and the inverse of the variance, and superimposing the third products to obtain a second superimposed value; and normalizing the fourth product between the second superimposed value and the first ratio to obtain the temporal continuity of the clusters.

[0009] Optionally, determining the time distribution degree of the cluster cluster based on the time interval between every two adjacent data segments in the cluster cluster and the second quantity includes: superimposing the time interval between every two adjacent data segments in the cluster cluster to obtain a third superposition value, and calculating the inverse of the third difference between the second quantity and a predetermined value; and determining the fifth product between the third superposition value and the inverse of the third difference as the time distribution degree of the cluster cluster.

[0010] Optionally, the abnormality degree of each period is corrected according to the temporal continuity and time distribution degree of the clustering clusters to obtain the final abnormality degree after correction of each period, including: calculating the third ratio between the time distribution degree of the clustering clusters where each period is located and the temporal continuity of the clustering clusters where each period is located, and the sixth product between each third ratio and the abnormality degree of each period; normalizing each sixth product to obtain the final abnormality degree after correction of each period.

[0011] Optionally, determining the possibility of a fault in the abnormal fluctuation cycle based on the amplitudes of each reference cycle of the abnormal fluctuation cycle and the time interval between adjacent reference cycles includes: taking the period within the most adjacent time period of the abnormal fluctuation cycle as the reference period of the abnormal fluctuation cycle; calculating the fourth difference between the fifth number of reference cycles and a predetermined value, and the fifth difference between the amplitudes of two adjacent reference cycles; calculating the fourth ratio between the fifth difference and the amplitude of the previous reference cycle of two adjacent reference cycles, and the second sum between the fourth ratio and the inverse of the time interval between adjacent reference cycles; superimposing each second sum to obtain a fourth superimposed value; and determining the inverse of the fourth superimposed value and the fourth difference as the possibility of a fault in the abnormal fluctuation cycle.

[0012] Optionally, predicting whether the substation equipment has a potential fault based on the possibility of a fault occurring during an abnormal fluctuation cycle includes: determining that the substation equipment has a potential fault when the possibility of a fault occurring during a most recent abnormal fluctuation cycle of the substation equipment is greater than a third threshold.

[0013] Optionally, the third threshold is a third difference between an average of the probability of failure of multiple substations in the most recent abnormal fluctuation cycle in historical data and a preset value.

[0014] The present invention has the following beneficial effects: first, the current data, load data and dissolved gas concentration of the substation equipment in multiple cycles are obtained; then, the abnormality degree of the current cycle is determined according to the amplitude of the maximum point of the current data in the current cycle, the first number of the maximum points of the current data in the current cycle and the abnormality degree of the previous cycle adjacent to the current cycle; and the load data and the dissolved gas concentration are clustered based on the correlation between the load data and the dissolved gas concentration to obtain multiple clusters; secondly, the abnormality degree of each cycle is corrected according to the load data and the dissolved gas concentration in each of the clusters to obtain the final abnormality degree after correction of each of the cycles; and the cycle whose final abnormality degree is greater than the first threshold is selected as the abnormal fluctuation cycle; finally, according to the amplitude of each reference cycle of the abnormal fluctuation cycle and the time interval between adjacent reference cycles, the possibility of failure of the abnormal fluctuation cycle is determined; and according to the possibility of failure of the abnormal fluctuation cycle, it is predicted whether the substation equipment has potential failure.

[0015] In this way, the embodiment of the present invention can monitor the current data, load data and dissolved gas concentration generated by the substation equipment during operation in real time, and determine the abnormal fluctuation period in combination with the above multi-dimensional data. It has the advantages of first analyzing the abnormal degree of each period through the current data, and then correcting the abnormal degree of each period using the load data and the dissolved gas concentration to obtain the final abnormal degree. Finally, the possibility of failure in the abnormal fluctuation period is determined, and it is used to predict whether there is a potential failure in the substation equipment. In this way, through the analysis of multi-dimensional data, the potential failures in the substation equipment can be effectively monitored and accurately predicted in the early stage, which improves the accuracy and reliability of the substation equipment failure prediction, avoids the expansion of equipment damage, and improves the reliability of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] 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.

[0017] Figure 1 A flowchart of a method for predicting failure of substation equipment based on big data provided by an embodiment of the present invention; Figure 2A schematic diagram of a current variation curve of an oil-immersed transformer under normal conditions provided by an embodiment of the present invention; Figure 3 A schematic diagram of a current variation curve of an oil-immersed transformer under abnormal conditions is provided in accordance with an embodiment of the present invention. DETAILED DESCRIPTION

[0018] 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 specific implementation method, structure, features and effects of a substation fault prediction method based on big data proposed by the present invention 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.

[0019] 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.

[0020] The following is a detailed description of a method for predicting faults of substation equipment based on big data provided by the present invention in conjunction with the accompanying drawings.

[0021] Embodiment 1: See also Figure 1 , which shows a flowchart of a method for predicting faults of substation equipment based on big data provided by an embodiment of the present invention, including: S101, obtaining current data, load data and dissolved gas concentration of a substation in multiple cycles.

[0022] Specifically, the types of substation equipment include but are not limited to transformers, circuit breakers, disconnectors, etc. The types of transformers include but are not limited to oil-immersed transformers, dry-type transformers, autotransformers, and isolation transformers. In the embodiment of the present invention, when predicting the fault of the substation equipment, taking the oil-immersed transformer as an example, its common fault is the damage of the internal equipment caused by abnormal discharge. Therefore, when analyzing, it is necessary to analyze and collect the common current data and load data in the transformer, and analyze the concentration of the gas of the insulating oil decomposed therein during the discharge in the insulating oil. Therefore, the embodiment of the present invention needs to install a current sensor and a load sensor at the current input and output ports of each oil-immersed transformer, and install a dissolved gas analyzer in the insulating oil inside the oil-immersed transformer, and continuously collect the real-time data such as the current data, load data and dissolved gas concentration on the oil-immersed transformer in real time. Among them, the current data can be collected in real time, and the collection interval of the dissolved gas concentration and load data can be 1s. In addition, the embodiment of the present invention also standardizes the collected data to facilitate subsequent calculations.

[0023] S102, determining the abnormality degree of the current cycle according to the amplitude of the maximum point of the current data in the current cycle, the first number of the maximum points of the current data in the current cycle, and the abnormality degree of the previous cycle adjacent to the current cycle.

[0024] Specifically, the change of the current in the oil-immersed transformer is a smooth sine wave showing periodic changes under normal circumstances. When the load of the oil-immersed transformer changes at the moment of abnormal discharge, the waveform of the current in the oil-immersed transformer will fluctuate. Due to the influence of the high-frequency current or load change generated during abnormal discharge, the peak value in the fluctuation image will be larger than that in normal circumstances in the same current cycle as normal, and the number of peak values ​​is also greater, showing a high-frequency pulse state. Therefore, the embodiment of the present invention judges the abnormal moment according to the difference between the amplitude and frequency of the current in a fixed cycle and that in normal circumstances. However, due to the interference of the load, it is impossible to accurately judge whether it is caused by abnormal discharge.

[0025] And because the current generated by abnormal discharge will decompose the insulating layer and insulating oil in the local area of ​​the oil-immersed transformer, the total amount of various gases dissolved in the insulating oil will increase. Therefore, according to whether the solubility and total amount increase at each abnormal moment, the interference caused by the load at the abnormal moment can be screened out, and by analyzing the screened abnormal moments, it can be determined whether the frequency of occurrence increases, and compared with historical data to obtain the final abnormal degree of the oil-immersed transformer.

[0026] Furthermore, in the oil-immersed transformer, the alternating current of different voltage levels is converted by the electromagnetic induction principle, and the current changes accordingly. Figure 2 and Figure 3 The current variation of oil-immersed transformer is explained. Figure 2 A schematic diagram of a current variation curve of an oil-immersed transformer under normal conditions is provided in accordance with an embodiment of the present invention. Figure 3 A schematic diagram of a current variation curve of an oil-immersed transformer under abnormal conditions is provided in accordance with an embodiment of the present invention. Figure 2 As shown in the figure, the amplitude, frequency and waveform of the oil-immersed transformer current are relatively stable and change slowly. However, if partial discharge occurs in the oil-immersed transformer, high-frequency pulses and irregular fluctuations will be introduced, such as Figure 3 As shown, it will cause the current to change dramatically instantly, the waveform becomes chaotic, and the frequency may fluctuate abnormally. In addition to partial discharge, changes in grid load will also cause current amplitude, waveform or frequency to fluctuate. By analyzing the current data in each time period in the oil-immersed transformer, the abnormal fluctuation point can be identified.

[0027] Furthermore, since the current in the oil-immersed transformer is a superposition of multiple current fluctuations, the period of its basic current is fixed. However, when abnormal discharge occurs in the oil-immersed transformer or the load in its power grid changes suddenly, the various high-frequency chaotic currents caused by the abnormal discharge or load change will cause the basic current to change after superposition. However, since the duration of abnormal discharge is very short, usually between a few milliseconds and several hundred milliseconds, and lasts for multiple cycles, the degree of abnormality in the current cycle can be judged based on the number of maximum points in each cycle of the current in the oil-immersed transformer, the maximum value of the maximum points in the cycle, and the number and amplitude of the maximum points in the previous cycle.

[0028] Further, as an optional embodiment of the present invention, according to the amplitude of the maximum point of the current data in the current cycle, the first number of the maximum points of the current data in the current cycle and the abnormality degree of the previous cycle adjacent to the current cycle, determining the abnormality degree of the current cycle includes: calculating a first difference between the first number and a predetermined value, and a second difference between the predetermined value and the abnormality degree of the adjacent previous cycle; calculating a third difference between the amplitude of each maximum point of the current data in the current cycle and the average of the amplitude of all maximum points of the current data in the adjacent previous cycle, and superimposing each third difference to obtain a first superimposed value; calculating a first product between the first difference and the second difference and a second product between the first superimposed value and the abnormality degree of the adjacent previous cycle; determining a first sum between the first product and the second product as the abnormality degree of the current cycle.

[0029] Specifically, the embodiment of the present invention obtains the amplitude of the maximum point of the current data in each cycle and the number of maximum points in the cycle respectively. For each cycle, the difference between the amplitude of the maximum point and the number of maximum points in the previous cycle can be calculated. If the previous cycle is a normal cycle, and the amplitude of the maximum point in the current cycle suddenly increases compared with it, the number of maximum points also increases, and the degree of abnormality is greater. If the previous cycle is an abnormal cycle, and the number and amplitude of the maximum points in the current cycle are not much different from them, the current cycle is also an abnormal cycle, and the degree of abnormality is greater. If the current cycle is greatly different from it, then the current cycle may be a normal cycle. According to the amplitude and number of the maximum points of each cycle and the previous cycle, the degree of abnormality of each cycle can be obtained. Among them, the embodiment of the present invention specifically uses the following formula to calculate the degree of abnormality of the current cycle: In the above formula, For the The degree of abnormality of the cycle. For the The first number of maximum points in a cycle. The larger the number, the more chaotic the frequency in the cycle is, that is, the greater the abnormality of the cycle is. For the cycle, that is, The abnormality level of the previous cycle adjacent to the cycle. For the The third difference between the amplitude of the jth maximum point in a cycle and the mean of the amplitudes of all maximum points in the previous cycle.

[0030] It is worth noting that when it is in the first cycle, its abnormality has no direct relationship with the abnormality of the previous cycle, because there is no previous cycle and it is impossible to compare the amplitude. As the abnormality of the first cycle, and To correct the degree of abnormality in the current cycle, the greater the abnormality of the previous cycle, the more the similarity between the maximum point in the current cycle and it is considered, because abnormal discharge will cause a rapid increase in current, and the smaller the abnormality, the more it is necessary to determine whether there is an abnormality by the number of extreme points.

[0031] S103: clustering the load data and the dissolved gas concentration based on the correlation between the load data and the dissolved gas concentration to obtain a plurality of clusters.

[0032] Specifically, since the equipment connected to the oil-immersed transformer is relatively complex, its load changes are relatively frequent and complex. When its load changes, the frequency and amplitude of the current in the oil-immersed transformer will also change. Some of them are similar to abnormal discharges, which will interfere with the detection of abnormal discharges. The moment caused by abnormal discharges cannot be obtained only through the fluctuation of the current in the oil-immersed transformer, so it is necessary to exclude the abnormal changes caused by load changes. Since the interior of the oil-immersed transformer is a closed space, the total amount of gas produced by the decomposition of the insulating oil under abnormal discharge is constant during two abnormal discharges. Therefore, in the two abnormal discharges, its solubility at the same temperature at different times is similar, and the only factors that cause temperature changes in the oil-immersed transformer are load and abnormal discharge.

[0033] Furthermore, when the oil-immersed transformer is working normally, the change in load will cause the equipment to heat up, thus causing its temperature to rise. Since the area of ​​heat release is large, the temperature rise of the insulating oil is proportional to the increase in load, which causes the solubility in the insulating oil to change accordingly with the change in load. When abnormal discharge occurs inside the oil-immersed transformer, it will only cause high heat and discharge in a small local area of ​​the oil-immersed transformer, and its impact on the overall temperature of the oil-immersed transformer is almost negligible. Therefore, the solubility in the insulating oil should still change with the change in load, but because the abnormal discharge will cause the insulating oil and the insulating layer in the local area to decompose and ionize, the total amount of various gases (dissolved and undissolved) inside will increase. Therefore, due to the increase in the pressure of the gas in the closed oil-immersed transformer at the corresponding temperature, the corresponding relationship between solubility and temperature will change compared with before discharge, that is, the load is proportional to the change in solubility.

[0034] Therefore, according to the analysis of the above embodiment, it can be known that there is a correlation between the change of solubility and load between two abnormal discharges, and there are differences in this correlation on both sides of each abnormal discharge. Therefore, the embodiment of the present invention uses the correlation relationship between the load data and the dissolved gas concentration at each moment as a parameter to perform clustering through the K-means clustering algorithm to obtain several clusters representing similar data sets. Among them, the clustering space of the K-means clustering algorithm is constructed by load and solubility. When clustering, the data with similar load data and dissolved gas concentration are clustered into a normal cluster. For some scattered data in which the load suddenly increases due to irregular abnormal discharge but the solubility does not change much, since the load data of these data points deviate significantly from other data points, they will be clustered into an independent abnormal cluster.

[0035] S104, correcting the abnormality degree of each cycle according to the load data and dissolved gas concentration in each cluster to obtain the final abnormality degree of each cycle after correction, and selecting the cycle with the final abnormality degree greater than the first threshold as the abnormal fluctuation cycle.

[0036] Specifically, for each cluster cluster obtained by clustering in the above embodiment, including a normal cluster cluster and an abnormal cluster cluster caused by abnormal discharge, in an oil-immersed transformer, abnormal discharge is a low-probability event. Since each abnormal discharge will cause the total amount of various gases in the oil-immersed transformer to change, resulting in different solubility on the same load in different abnormal discharge time periods (between two abnormal discharges), therefore, when clustering, all the time points of abnormal discharge will be clustered into one cluster cluster, and the data in the same load interval in the same abnormal discharge time period will be clustered into one cluster cluster, and since the solubility of the gas under the same load in different abnormal discharge time periods is different, there are several cluster clusters with different abnormal discharge time periods in the same load interval. For each cluster cluster, since the load in the oil-immersed transformer is constantly changing, the data in the cluster cluster is also multiple segments of data under the same load, and the data segments are not connected.

[0037] Further, as an optional embodiment of the present invention, the abnormality degree of each period is corrected according to the load data and dissolved gas concentration in each cluster cluster to obtain the final abnormality degree after correction of each period, including: sorting the load data and dissolved gas concentration in each cluster cluster in chronological order and then segmenting them to obtain multiple data segments; determining the temporal continuity of the cluster cluster according to the second number of data segments in the cluster cluster, the third number of data points in the cluster cluster, the fourth number of data points in the data segment, and the change between every two adjacent data points in the data segment; determining the temporal distribution degree of the cluster cluster according to the time interval between every two adjacent data segments in the cluster cluster and the second number; correcting the abnormality degree of each period according to the temporal continuity and temporal distribution degree of the cluster cluster to obtain the final abnormality degree after correction of each period.

[0038] Specifically, the embodiment of the present invention arranges the elements in each of the obtained clusters in chronological order. Since each cluster is not completely composed of a continuous data segment, but is composed of several continuous data segments, and in the abnormal discharge cluster, it is composed of several discrete abnormal discharge time points, the embodiment of the present invention needs to divide the data segments in each cluster first, but since the load in the oil-immersed transformer will fluctuate, when the load changes and then recovers again in a short time, it will not cause a large change in the temperature of the oil-immersed transformer, so these fluctuations need to be smoothed. As an optional embodiment of the present invention, the load data and dissolved gas concentration in each cluster are sorted in chronological order and then segmented to obtain multiple data segments, including: determining the time interval between the load data and dissolved gas concentration at the current time point in the cluster and the load data and dissolved gas concentration at the adjacent previous time point; when the time interval is greater than a second threshold, the load data and dissolved gas concentration at the current time point are used as the starting point of a new segment of data points; when the time interval is not greater than the second threshold, the load data and dissolved gas concentration at the current time point are classified into the data segment where the previous adjacent load data is located.

[0039] Specifically, the second threshold can be determined according to actual conditions. In the embodiment of the present invention, the second threshold is For each data point in the cluster, that is, the load data and dissolved gas concentration at each time point, determine whether the time interval between it and the previous data point is greater than If it is greater than , it is used as the starting point of a new data segment. If it is less than , it is classified into the data segment where the previous data point is located, thereby obtaining several data segments in each cluster. It is worth noting that an isolated data point can also be regarded as a data segment.

[0040] Furthermore, for each cluster, the closer the number of data segments is to the number of data points in the cluster, the lower the continuity between the data in the cluster, that is, the more discrete the data is, the more likely it is that multiple abnormal discharges occurred in the oil-immersed transformer, and the more likely the discontinuity of the data in the cluster is caused by abnormal discharge. For each data segment, the smaller the difference in the time series of the data within it, the more stable the change of the data segment is, and the higher its continuity is. Therefore, the temporal continuity of each cluster can be obtained based on the data segments in each cluster. As an optional embodiment of the present invention, determining the temporal continuity of the clustering clusters based on the second number of data segments in the clustering clusters, the third number of data points in the clustering clusters, the fourth number of data points in the data segments, and the amount of change between every two adjacent data points in the data segments includes: calculating a first ratio between the second number and the third number and a second ratio between the fourth number and the third number; calculating an average change in the amount of change between every two adjacent data points in the data segment and a variance of the amount of change between every two adjacent data points in the data segment; calculating a third product between the second ratio, the inverse of the average change and the inverse of the variance, and superimposing the third products to obtain a second superimposed value; and normalizing the fourth product between the second superimposed value and the first ratio to obtain the temporal continuity of the clustering clusters.

[0041] Specifically, the following formula is used in the embodiment of the present invention to calculate the temporal continuity of clusters: In the above formula, For the The temporal continuity of clusters. For the The second number of data segments in the clusters. For the The third number of data points in the cluster. For the The first cluster The fourth number of data points in the data segment. For the The first cluster The mean of the changes between every two adjacent data points in a data segment, that is, the average change. For the The first The variance of the change between every two adjacent data points in a data segment. Function is used to Perform normalization.

[0042] in, The relationship between the number of data segments and the total number of data points is measured. The more data segments there are, the worse the continuity of the data is, which reflects that the data fluctuation in this time period in the oil-immersed transformer is longer and the data is broken. The more data breaks there are in the cluster, the lower the temporal continuity of the cluster. It can reflect the stability and scale of each data segment in the cluster, that is, the stability of the load change in the oil-immersed transformer during this time period. The number of data points in the data segment is used to indicate whether there are large abnormal fluctuations in the operating state of the oil-immersed transformer during this period. The larger the data segment, the more likely the time series in the data segment is to be stable and has a higher continuity. and The mean and variance of the changes between adjacent data points in a data segment are measured respectively, which together reflect the stationarity of the data segment. The smaller the mean and the smaller the variance, the more stable the change of the data segment is and the higher its continuity is.

[0043] Further, since the data in the cluster cluster is the clustering between each data segment when the total amount of various gases in the oil-immersed transformer remains unchanged, its data only includes the data between two abnormal discharges. The time interval formed by all the data segments in each normal cluster cluster is relatively short, which is only the length between two abnormal discharges, while in the cluster cluster caused by abnormal discharge, it contains the data when the oil-immersed transformer has abnormal discharge at each time node, so the longer the time interval in the cluster cluster, the longer the time interval, so the time distribution degree of each cluster cluster can be determined according to the time interval between each data segment in each cluster cluster. As an optional embodiment of the present invention, according to the time interval between each two adjacent data segments in the cluster cluster and the second quantity, determining the time distribution degree of the cluster cluster includes: superimposing the time interval between each two adjacent data segments in the cluster cluster to obtain a third superposition value, and calculating the inverse of the third difference between the second quantity and the predetermined value; determining the fifth product between the third superposition value and the inverse of the third difference as the time distribution degree of the cluster cluster.

[0044] Specifically, the predetermined value in the embodiment of the present invention may be 1. The embodiment of the present invention specifically uses the following formula to calculate the time distribution degree of the cluster: In the above formula, For the The temporal distribution of clusters. For the The second number of data segments in the clusters. For the The first The data segment and The time interval between one data segment and the next adjacent data segment. The time distribution of the cluster is described by the average value of the time intervals of adjacent data segments in the cluster. The smaller the value, the shorter the time interval between the data segments, the more concentrated the time distribution, and the cluster is normal. On the contrary, if the value is large, it means that the time interval is long, and there may be abnormal discharge.

[0045] Furthermore, after obtaining the temporal continuity of each cluster and the degree of temporal distribution within the cluster, they can be used as weights to correct the degree of abnormality of each cycle, thereby obtaining the final degree of abnormality of each cycle. As an optional embodiment of the present invention, the degree of abnormality of each cycle is corrected according to the temporal continuity and temporal distribution of the cluster, and the final degree of abnormality after correction of each cycle is obtained, including: calculating the third ratio between the degree of temporal distribution of the cluster where each cycle is located and the temporal continuity of the cluster where each cycle is located, and the sixth product between each third ratio and the degree of abnormality of each cycle; normalizing each sixth product to obtain the final degree of abnormality after correction of each cycle.

[0046] Specifically, the embodiment of the present invention specifically uses the following formula to calculate the final abnormality degree: In the above formula, For the The final abnormality level after correction for a period of time. For the The degree of abnormality of the cycle. For the The temporal continuity of the clusters in which the cycles are located. For the The time distribution degree corresponding to the clustering cluster where the period is located. Function is used to Normalization is performed. Come to the The abnormal degree of each cycle is corrected. If If the value is larger, it means that the time distribution in the cluster is wider and there is a longer time interval, which means that it contains multiple abnormal discharges in the transformer, which will increase the weight of the abnormality at that moment, indicating that the area with a longer time distribution is more likely to have an abnormality. If the value is smaller, it means that the time continuity within the cluster is poor and the data fluctuates greatly, indicating that the cluster may have abnormal transformer discharge, which will amplify the abnormality of the cycle, indicating that the cycle with large fluctuations and instability is more likely to be an abnormal cycle.

[0047] Furthermore, the first threshold value can be determined according to actual conditions, and in the embodiment of the present invention, the value is 0.8. The period in which the final abnormal degree after correction is greater than 0.8 is recorded as the abnormal fluctuation period.

[0048] S105, determining the possibility of a fault occurring in the abnormal fluctuation period according to the amplitude of each reference period of the abnormal fluctuation period and the time interval between adjacent reference periods.

[0049] Specifically, when partial discharge occurs in an oil-immersed transformer, the insulation material inside the oil-immersed transformer will be further damaged due to local high heat and ionization. In addition, as the oil-immersed transformer is used, the vibration and temperature changes during operation of the oil-immersed transformer will also damage some parts of the oil-immersed transformer. These reasons make partial discharge more likely to occur in the oil-immersed transformer. Therefore, as the oil-immersed transformer is used, the frequency and amplitude of partial discharge in the oil-immersed transformer will increase. Therefore, the abnormal fluctuation cycle caused by abnormal discharge in the historical data of the oil-immersed transformer can be analyzed.

[0050] Further, as an optional embodiment of the present invention, determining the possibility of a fault in an abnormal fluctuation cycle based on the amplitudes of each reference cycle of the abnormal fluctuation cycle and the time interval between adjacent reference cycles includes: taking the period within the most adjacent time period of the abnormal fluctuation cycle as the reference period of the abnormal fluctuation cycle; calculating the fourth difference between the fifth number of reference cycles and a predetermined value, and the fifth difference between the amplitudes of two adjacent reference cycles; calculating the fourth ratio between the fifth difference and the amplitude of the previous reference cycle of two adjacent reference cycles, and the second sum between the fourth ratio and the inverse of the time interval between adjacent reference cycles; superimposing each second sum to obtain a fourth superimposed value; and determining the inverse of the fourth superimposed value and the fourth difference as the possibility of a fault in the abnormal fluctuation cycle.

[0051] Specifically, the embodiment of the present invention uses the period in the shortest time of the abnormal fluctuation period as the reference period of the abnormal fluctuation period. According to the time point and amplitude of the current data of the reference period of each abnormal fluctuation period in the time series, the possibility of failure of each abnormal period is obtained. Among them, the predetermined value can be 1, and the embodiment of the present invention specifically uses the following formula to calculate the possibility: In the above formula, For the The possibility of failure during an abnormal fluctuation cycle. For the The fifth number of reference periods of abnormal fluctuation periods. and Respectively The abnormal fluctuation cycle The amplitude of the current data of the first reference cycle and the The amplitude of the current data of a reference cycle. For the The abnormal fluctuation cycle The time interval between a reference cycle and its previous reference cycle. To indicate the degree of amplitude progression, the larger the value, the more significant the amplitude progression, and the greater the possibility of abnormal fault. To indicate the degree of time progression, the larger the value, the shorter the time interval, the higher the frequency of abnormal fluctuations, and the greater the possibility that it is an abnormal failure.

[0052] S106, predicting whether there is a potential fault in the substation equipment based on the possibility of a fault occurring during the abnormal fluctuation period.

[0053] Specifically, the embodiment of the present invention can compare the possibility of failure in the abnormal fluctuation cycle with the third threshold value to determine whether the substation equipment has a potential failure. As an optional embodiment of the present invention, predicting whether the substation equipment has a potential failure based on the possibility of failure in the abnormal fluctuation cycle includes: determining that the substation equipment has a potential failure when the possibility of failure in the most recent abnormal fluctuation cycle of the substation equipment is greater than the third threshold value.

[0054] Specifically, for each oil-immersed transformer, its status is detected in real time. When it is monitored that the possibility of failure in its most recent abnormal fluctuation cycle is greater than the third threshold when the oil-immersed transformer needs maintenance, it means that the risk of equipment failure is high and maintenance is required. The relevant personnel will be notified to carry out maintenance.

[0055] Furthermore, the third threshold value can be determined according to actual conditions. When determining the third threshold value in the embodiment of the present invention, the third threshold value is the third difference between the average of the probability of failure of multiple substations in the most recent abnormal fluctuation cycle in historical data and the preset value.

[0056] Specifically, the preset value can be taken according to the actual situation, and the value of the embodiment of the present invention is 0.2. After obtaining the possibility of failure in each abnormal fluctuation cycle, the average of each possibility is calculated according to the possibility of failure in the most recent abnormal fluctuation cycle before the failure in the historical data of multiple oil-immersed transformers. Since it is impossible to wait until the equipment reaches the limit before maintenance, the value of the average minus 0.2 is used as the third threshold when the oil-immersed transformer needs to be repaired.

[0057] The embodiment of the present invention can monitor the current data, load data and dissolved gas concentration generated by the substation equipment during operation in real time, and determine the abnormal fluctuation period in combination with the above multi-dimensional data. It has the advantages of first analyzing the abnormal degree of each period through the current data, and then correcting the abnormal degree of each period using the load data and the dissolved gas concentration to obtain the final abnormal degree. Finally, the possibility of failure in the abnormal fluctuation period is determined, and it is used to predict whether there is a potential failure in the substation equipment. In this way, through the analysis of multi-dimensional data, the potential failures in the substation equipment can be effectively monitored and accurately predicted in the early stage, which improves the accuracy and reliability of the prediction of substation equipment failures, avoids the expansion of equipment damage, and improves the reliability of the power system.

[0058] 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.

[0059] 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 predicting failure of substation equipment based on big data, characterized in that: The power substation equipment fault prediction method based on big data includes: Obtain current data, load data and dissolved gas concentration of substation equipment in multiple cycles; Determine the abnormality degree of the current cycle according to the amplitude of the maximum value point of the current data in the current cycle, the first number of the maximum value points of the current data in the current cycle, and the abnormality degree of the previous cycle adjacent to the current cycle; Clustering the load data and the dissolved gas concentration based on the correlation between the load data and the dissolved gas concentration to obtain a plurality of clusters; Correcting the abnormality degree of each cycle according to the load data and dissolved gas concentration in each of the clusters to obtain the final abnormality degree of each cycle after correction, and selecting the cycle whose final abnormality degree is greater than the first threshold as the abnormal fluctuation cycle; Determining the possibility of a fault occurring in the abnormal fluctuation period according to the amplitude of each reference period of the abnormal fluctuation period and the time interval between adjacent reference periods; Whether the power transformation equipment has a potential fault is predicted based on the possibility of a fault occurring during the abnormal fluctuation period.

2. The method for predicting failure of substation equipment based on big data according to claim 1 is characterized in that: Determining the abnormality degree of the current cycle according to the amplitude of the maximum value point of the current data in the current cycle, the first number of the maximum value points of the current data in the current cycle, and the abnormality degree of the previous cycle adjacent to the current cycle includes: Calculating a first difference between the first quantity and a predetermined value, and a second difference between the predetermined value and the abnormality degree of the adjacent previous cycle; Calculating a third difference between the amplitude of each of the maximum value points of the current data in the current cycle and the average of the amplitudes of all the maximum value points of the current data in the adjacent previous cycle, and superimposing each of the third differences to obtain a first superimposed value; Calculating a first product between the first difference and the second difference and a second product between the first superposition value and the abnormality degree of the adjacent previous cycle; A first sum value between the first product and the second product is determined as the abnormality degree of the current cycle.

3. The method for predicting failure of substation equipment based on big data according to claim 1, characterized in that: The abnormality degree of each cycle is corrected according to the load data and dissolved gas concentration in each cluster to obtain the final abnormality degree of each cycle after correction, which includes: Sorting the load data and dissolved gas concentration in each of the clusters in chronological order and then segmenting them to obtain a plurality of data segments; Determining the temporal continuity of the clusters according to a second number of the data segments in the clusters, a third number of data points in the clusters, a fourth number of data points in the data segments, and a change between every two adjacent data points in the data segments; Determining a time distribution degree of the cluster according to a time interval between every two adjacent data segments in the cluster and the second quantity; The abnormality degree of each period is corrected according to the temporal continuity of the clusters and the temporal distribution degree, so as to obtain the corrected final abnormality degree of each period.

4. The method for predicting failure of substation equipment based on big data according to claim 3 is characterized in that: The load data and dissolved gas concentration in each cluster are sorted in chronological order and then segmented to obtain multiple data segments including: Determine the time interval between the load data and dissolved gas concentration at a current time point and the load data and dissolved gas concentration at an adjacent previous time point in the cluster; When the time interval is greater than the second threshold, the load data and dissolved gas concentration at the current time point are used as the starting point of a new segment of data points; when the time interval is not greater than the second threshold, the load data and dissolved gas concentration at the current time point are classified into the data segment where the previous adjacent load data is located.

5. The method for predicting failure of substation equipment based on big data according to claim 3 is characterized in that: Determining the temporal continuity of the clusters according to the second number of the data segments in the clusters, the third number of the data points in the clusters, the fourth number of the data points in the data segments, and the amount of change between every two adjacent data points in the data segments comprises: calculating a first ratio between the second quantity and the third quantity and a second ratio between the fourth quantity and the third quantity; Calculate the average change in the change between every two adjacent data points in the data segment and the variance of the change between every two adjacent data points in the data segment; Calculating a third product of the second ratio, the inverse of the average change, and the inverse of the variance, and superimposing the third products to obtain a second superimposed value; A fourth product of the second superposition value and the first ratio is normalized to obtain the temporal continuity of the clusters.

6. The method for predicting failure of substation equipment based on big data according to claim 3 is characterized in that: Determining the time distribution degree of the cluster according to the time interval between every two adjacent data segments in the cluster and the second quantity includes: superimposing the time intervals between every two adjacent data segments in the cluster to obtain a third superposition value, and calculating the reciprocal of a third difference between the second number and a predetermined value; A fifth product of the third superposition value and the inverse of the third difference value is determined as the time distribution degree of the cluster.

7. The method for predicting failure of substation equipment based on big data according to claim 3 is characterized in that: The abnormality degree of each period is corrected according to the temporal continuity of the clusters and the temporal distribution degree to obtain the corrected final abnormality degree of each period, including: Calculating a third ratio between the time distribution degree of the clusters where each period is located and the temporal continuity of the clusters where each period is located, and a sixth product between each of the third ratios and the abnormality degree of each period; Each of the sixth products is normalized to obtain a final abnormality degree after correction of each of the cycles.

8. The method for predicting failure of substation equipment based on big data according to any one of claims 1 to 7, characterized in that: Determining the possibility of a fault occurring in the abnormal fluctuation period according to the amplitude of each reference period of the abnormal fluctuation period and the time interval between adjacent reference periods includes: Taking the period in the nearest adjacent time period of the abnormal fluctuation period as the reference period of the abnormal fluctuation period; calculating a fourth difference between a fifth number of the reference cycles and a predetermined value, and a fifth difference between the amplitudes of two adjacent reference cycles; Calculating a fourth ratio between the fifth difference and the amplitude of a previous reference cycle of the two adjacent reference cycles, and a second sum of the fourth ratio and the inverse of the time interval between the adjacent reference cycles; Superimposing each of the second sum values ​​to obtain a fourth superimposed value; The reciprocal of the fourth superposition value and the fourth difference value is determined as the possibility of a fault occurring in the abnormal fluctuation period.

9. The method for predicting failure of substation equipment based on big data according to claim 1, characterized in that: The predicting whether the substation equipment has a potential fault according to the possibility of a fault occurring in the abnormal fluctuation period includes: When the possibility of a failure in the most recent abnormal fluctuation cycle of the power transformation equipment is greater than a third threshold, it is determined that a potential failure exists in the power transformation equipment.

10. The method for predicting failure of substation equipment based on big data according to claim 9, characterized in that: The third threshold is a third difference between the average of the probability of failure of multiple substations in the most recent abnormal fluctuation cycle in historical data and a preset value.

Images