A high voltage voltage transformer performance monitoring system and method based on artificial intelligence

By building an online performance monitoring system based on artificial intelligence on high-voltage voltage transformers, using sensors to collect data and machine learning analysis, real-time monitoring and intelligent early warning of high-voltage voltage transformers are solved, which cannot be effectively monitored and diagnosed in the existing technology. The problem has improved the reliability of the equipment and the safety of the locomotive.

CN118818407BActive Publication Date: 2024-12-03CHANGZHOU SHUNCHUANG ELECTRICAL TECH
View PDF 9 Cites 0 Cited by

Patent Information

Application Number
CN202411305281.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-19
Publication Date
2024-12-03
Estimated Expiration
2044-09-19

AI Technical Summary

Technical Problem

The prior art cannot effectively monitor and diagnose the operating status of high-voltage voltage transformers online in real time, resulting in possible failures in harsh environments, affecting the safety of the locomotive and the reliability of the equipment.

Method used

Adopt a high-voltage voltage transformer performance monitoring system based on artificial intelligence, collect operating parameters by laying sensor devices, build an online performance warning monitoring platform, use machine learning to analyze historical data, dynamically alarm and adjust sampling frequency, evaluate safe operation index, and realize high voltage Real-time monitoring and intelligent early warning of voltage transformers.

Benefits of technology

Real-time status judgment of high-voltage voltage transformers is achieved, maximum equipment life, reduce maintenance costs, avoid accident risks, and improve operational reliability and safety.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118818407B_ABST
    Figure CN118818407B_ABST
Patent Text Reader

Abstract

The present invention discloses a high-voltage voltage transformer performance monitoring system and method based on artificial intelligence, which relate to the technical field of electrical equipment performance monitoring. The present application can extend the life cycle of the high-voltage voltage transformer to the maximum extent, greatly reduce the maintenance cost, effectively avoid the risk of high-voltage voltage transformer accidents, improve the maintenance efficiency and operation reliability, and improve the safe operation of locomotives by establishing a mode of online monitoring and intelligent early warning of the high-voltage voltage transformer, and at the same time, the present application monitors and adjusts the synchronous sampling state of the sensor device in real time by observing the parameter value changes presented by the sampling of each operating parameter item of the high-voltage voltage transformer through the corresponding sensor device when the high-voltage voltage transformer operation alarm is approaching, thereby ensuring as much as possible that the high-voltage voltage transformer is abnormal before the abnormal state is intervened and the early warning of the abnormal state is intervened.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of electrical equipment performance monitoring, and in particular to a high-voltage voltage transformer performance monitoring system and method based on artificial intelligence. Background Art

[0002] As one of the high-voltage electrical equipment of locomotives, the high-voltage voltage transformer mainly realizes the functions of overhead line voltage measurement, power metering and protection. The resin and silicone rubber materials used inside and outside the high-voltage voltage transformer have high insulation and reliability. The insulation performance of the above materials will be greatly reduced if the transformer is used for a long time in harsh environment or the material is aged and damp, which will cause the insulation performance to decline, resulting in burning, cracking, and even explosion of the relevant parts of the transformer.

[0003] At present, offline maintenance has limitations in judging the health status of high-voltage voltage transformers that have been running for a long time under complex contact networks. Offline detection methods cannot simulate the complex actual environment on site, nor can they monitor the on-site operating status in real time. Voltage transformers that pass the traditional offline status detection method still have the risk of failure when used on site.

[0004] Developing a system that can monitor and diagnose the operating status of high-voltage voltage transformers online and intervene in advance before the high-voltage voltage transformers become abnormal to ensure the safety of locomotive driving has become an urgent task. High-precision sensors are one of the most important components of online monitoring systems. Their performance, quality, and operating status directly affect the measurement effect of the monitoring parameters of high-voltage voltage transformers. Summary of the invention

[0005] The purpose of the present invention is to provide a high-voltage voltage transformer performance monitoring system and method based on artificial intelligence to solve the problems raised in the prior art.

[0006] To achieve the above object, the present invention provides the following technical solution: a high-voltage voltage transformer performance monitoring method based on artificial intelligence, the method comprising:

[0007] Step S1: by deploying a number of sensor devices on the high-voltage voltage transformer of the locomotive, the parameter values ​​of various operating parameter items of the high-voltage voltage transformer in operation are collected; by setting the corresponding standard operating range for each operating parameter item, an online performance early warning monitoring platform for the high-voltage voltage transformer is established; usually when an electrical device is operating normally, there must be an inherent physical connection between the operating parameter items it monitors, so the operating range of the corresponding operating parameter items will conform to a certain characteristic parameter matrix. This application analyzes a large amount of historical data by introducing machine learning and artificial intelligence means to find out the characteristic parameter matrix of the high-voltage voltage transformer during normal operation, so that the online performance early warning monitoring platform built according to the characteristic parameter matrix can realize dynamic alarm for any deviation from the normal operation of the high-voltage voltage transformer; the above-mentioned dynamic alarm process includes: when the relevant operating parameter items deviate from the predicted interval of the matrix, the alarm will issue a warning; that is, each operating parameter item of the high-voltage voltage transformer has its corresponding standard operating range when it is operating normally;

[0008] Step S2: sorting out the value changes of each operating parameter item when the corresponding sensor device samples the parameter value according to the corresponding sampling frequency, and identifying and extracting the characteristic fluctuation value of each operating parameter item;

[0009] Step S3: Traverse the characteristic fluctuation value and standard operation range corresponding to each operation parameter item, extract the relative operation range of each operation parameter item in the corresponding standard operation range through the characteristic fluctuation value of each operation parameter item, and evaluate and calculate the safety operation index of each operation parameter item according to the numerical value inclusion distribution phenomenon presented by the relative operation range and the corresponding standard operation range;

[0010] Step S4: Collect historical fault monitoring and early warning records, obtain the operating parameter items with abnormalities in each historical fault monitoring and early warning record, and use the corresponding standard operating range as the judgment standard;

[0011] If a certain operating parameter item is an operating parameter item that is reported to be abnormal in a certain historical fault monitoring and early warning record, then the certain historical fault monitoring and early warning record is determined to be a target record of the certain operating parameter item;

[0012] Step S5: in each target record of each operating parameter item, by combing the parameter values ​​of each operating parameter item before and after the corresponding target record is generated and the numerical distribution phenomenon presented by the corresponding standard operating interval or relative operating interval, it is determined whether to perform feature marking processing on the corresponding target record;

[0013] Step S6: According to the quantity distribution of target records marked with features for each operating parameter item, the safe operation index of each operating parameter item is adjusted. At the same time, in combination with the frequency of abnormalities reported for each operating parameter item, the stability index of each operating parameter item is calculated, and the sensor device that samples the parameter value of the operating parameter item whose stability index is less than the index threshold is identified, and an information prompt for adjusting the sampling frequency of the corresponding sensor device is sent to the management terminal.

[0014] Preferably, step S2 comprises:

[0015] Step S2-1: for each sensor device, randomly construct N parameter value groups from the parameter value change sequence distributed in chronological order obtained by collecting the corresponding operating parameter items according to the corresponding sampling frequency, wherein each parameter value group is composed of parameter values ​​obtained by two adjacent samplings;

[0016] Step S2-2: Assume that in a certain parameter value group, the parameter value corresponding to the earlier sampling time is Y 1 , the parameter value corresponding to the sampling time after is Y 2 , calculate the fluctuation value β=|Y in a certain parameter value group 2 -Y 1 |; For each operating parameter item, obtain the average value of the fluctuation values ​​extracted from the corresponding N parameter value groups, and set the average value as the characteristic fluctuation value presented by each operating parameter item when the corresponding sensor device samples the parameter value according to the corresponding sampling frequency.

[0017] Preferably, step S3 includes:

[0018] Step S3-1: If the standard operating interval corresponding to a certain operating parameter item is [A, B], the characteristic fluctuation value is α, where A is the minimum threshold in the standard operating interval, and B is the maximum threshold in the standard operating interval; capture the first numerical endpoint R(A)=A+α, and the second numerical endpoint R(B)=B-α in the standard operating interval [A, B]; extract the first numerical range P1=[A, R(A)] and the second numerical range P2=[R(B, B] for a certain operating parameter;

[0019] Step S3-2: When P1∩P2=∅, it is determined that a relative operating interval G=[R(A), R(B)] exists for a certain operating parameter item, and a safe operation index δ=[R(B)-R(A)] / [BA] is calculated for the certain operating parameter item, where 0<δ<1; when P1∩P2≠∅, it is determined that a relative operating interval does not exist for a certain operating parameter item, and the safe operation index δ of the certain operating parameter item is assigned a value of 0;

[0020] The relative operation interval G=[R(A), R(B)] is obtained by referring to the fluctuation of the parameter value sampled from the corresponding sensor device end, that is, according to the characteristic fluctuation value of the corresponding operation parameter item. According to the fluctuation law of the sampled parameter value presented from the corresponding sensor device end, if the parameter value obtained in one sampling of the above operation parameter item is within its relative operation interval G=[R(A), R(B)], because the numerical deviation between the parameter value obtained in the next sampling and the currently obtained parameter value is likely to satisfy a law similar to its characteristic fluctuation value, the parameter value sampled in the next sampling of the operation parameter item is more likely to be within the standard operation interval;

[0021] The larger the value of δ=[R(B)-R(A)] / [BA], the larger the proportion of the relative operating range G=[R(A), R(B)] in the standard operating range [A, B], which means that the safe fluctuation range of the parameter value of the corresponding operating parameter item is larger.

[0022] Preferably, step S5 comprises:

[0023] Step S5-1: extracting target records corresponding to each operating parameter item, and setting each operating parameter item for which a relative operating interval is determined to exist as a first characteristic parameter item, and setting each operating parameter item for which a relative operating interval is determined not to exist as a second characteristic parameter item;

[0024] Step S5-2: If a sensor device that samples the parameter value of a first characteristic parameter item is extracted from a target record corresponding to a first characteristic parameter item, the parameter value d obtained by the most recent sampling of the first characteristic parameter item before the generation of the target record and the relative operation interval D corresponding to the first characteristic parameter item satisfy d∈D, a characteristic mark is made for the target record;

[0025] Step S5-3: If in a target record corresponding to a second characteristic parameter item, an abnormal parameter value corresponding to a second characteristic parameter item when a target record is generated is extracted, and the minimum parameter deviation value between the abnormal parameter value and the standard operating interval corresponding to the second characteristic parameter item, the minimum parameter deviation value here is a numerical deviation result obtained by calculating the deviation between the above abnormal parameter value and the minimum value or maximum value of the standard operating interval; when the deviation threshold is greater than the deviation threshold, a target record is marked as a feature, wherein the setting method of the deviation threshold can be set based on the length of time spent on conducting corresponding troubleshooting on the high-voltage voltage transformer.

[0026] Preferably, step S6 comprises:

[0027] Step S6-1: Obtain the total number k of target records corresponding to each operating parameter item, and respectively accumulate the total number n of target records containing characteristic marks for each operating parameter item; obtain the total number M of historical fault monitoring and early warning records;

[0028] Step S6-2: extract the safe operation index δ of each operating parameter item, and calculate the influence adjustment coefficient θ=n / k for each operating parameter item; when the safe operation index δ of a certain operating parameter item is 0, adjust it to obtain a new safe operation index δ'=δ+1 / θ; when the safe operation index δ of a certain operating parameter item satisfies 0<δ<1, adjust it to obtain a new safe operation index δ'=δ θ ;

[0029] Among them, the larger the n / k, the more frequently the corresponding operating parameter item appears in the relative operating range when the parameter value obtained by the previous sampling is within the parameter value range, but the parameter value obtained by the next sampling exceeds the corresponding standard operating range, indicating that when the high-voltage voltage transformer operation alarm is approaching, the parameter value change of the operating parameter item presented by the sampling of the corresponding sensor device is more unstable;

[0030] Among them, when δ=0, the larger the θ value is, the smaller the corresponding new safe operation index δ' value is, indicating that when the high-voltage voltage transformer operation alarm is approaching, the parameter value change presented by the sampling of the corresponding sensor device of the operation parameter item is more unstable; Among them, when δ satisfies 0<δ<1, the larger the θ value is, the smaller the corresponding new safe operation index δ' value is, indicating that when the high-voltage voltage transformer operation alarm is approaching, the parameter value change presented by the sampling of the corresponding sensor device of the operation parameter item is more unstable;

[0031] Step S6-3: Evaluate the stability index F=1 / (k / M)×δ' of each operating parameter item; set the operating parameter item whose corresponding stability index is less than the index threshold as the target parameter item; feature-mark the sensor device that samples the corresponding parameter value of each target parameter item, and send a message prompting the administrator terminal to adjust the sampling frequency of the sensor device containing the feature mark;

[0032] Among them, the larger the k / M is, the more frequently the operating parameter is reported to be abnormal in all historical fault monitoring and early warning records, and the more frequently the high-voltage voltage transformer has a dynamic alarm due to abnormal parameter values ​​in the operating parameter item.

[0033] In order to better implement the above method, a high-voltage voltage transformer performance monitoring system is also proposed, which includes: a monitoring platform construction management module, a characteristic fluctuation value extraction module, a safe operation index evaluation module, a target record judgment module, a characteristic mark processing module, and a monitoring equipment prompt management module;

[0034] The monitoring platform builds a management module, which is used to deploy several sensor devices on the locomotive's high-voltage voltage transformer and collect parameter values ​​of various operating parameter items of the high-voltage voltage transformer in operation; by setting corresponding standard operating ranges for each operating parameter item, an online performance early warning monitoring platform for the high-voltage voltage transformer is built;

[0035] A characteristic fluctuation value extraction module is used to sort out the value changes of each operating parameter item when the corresponding sensor device samples the parameter value according to the corresponding sampling frequency, and identify and extract the characteristic fluctuation value of each operating parameter item;

[0036] The safety operation index evaluation module is used to traverse the characteristic fluctuation value and standard operation range corresponding to each operation parameter item, extract the relative operation range of each operation parameter item in the corresponding standard operation range through the characteristic fluctuation value of each operation parameter item, and evaluate and calculate the safety operation index of each operation parameter item according to the numerical value inclusion distribution phenomenon presented by the relative operation range and the corresponding standard operation range;

[0037] The target record judgment module is used to collect historical fault monitoring and early warning records, obtain the operating parameter items with abnormal feedback in each historical fault monitoring and early warning record, and use the corresponding standard operating range as the judgment standard; if a certain operating parameter item is an operating parameter item with abnormal feedback in a certain historical fault monitoring and early warning record, then the certain historical fault monitoring and early warning record is judged as the target record of the certain operating parameter item;

[0038] The feature mark processing module is used to determine whether to perform feature mark processing on the corresponding target record by sorting out the parameter values ​​of each operating parameter item before and after the corresponding target record is generated and the numerical distribution phenomenon presented by the corresponding standard operating interval or relative operating interval in each target record of each operating parameter item in turn;

[0039] The monitoring equipment prompt management module is used to adjust the safe operation index of each operating parameter item according to the distribution of the number of target records marked with characteristics for each operating parameter item, and at the same time calculate the stability index of each operating parameter item based on the frequency of abnormal feedback of each operating parameter item, identify the sensor device that samples the parameter value of the operating parameter item whose stability index is less than the index threshold, and send an information prompt to the management personnel terminal to adjust the sampling frequency of the corresponding sensor device.

[0040] Preferably, the safe operation index evaluation module includes: a relative operation interval extraction unit and a safe operation index calculation unit;

[0041] A relative operation interval extraction unit is used to traverse the characteristic fluctuation value and the standard operation interval corresponding to each operation parameter item, and extract the relative operation interval of each operation parameter item in the corresponding standard operation interval through the characteristic fluctuation value of each operation parameter item;

[0042] The safe operation index calculation unit is used to evaluate and calculate the safe operation index of each operating parameter item based on the numerical value inclusion distribution phenomenon presented by the relative operating range and the corresponding standard operating range.

[0043] Preferably, the monitoring equipment prompt management module includes a safe operation index adjustment unit and a stability index calculation management unit;

[0044] A safe operation index adjustment unit, used to adjust the safe operation index of each operating parameter item according to the quantity distribution of target records marked with features corresponding to each operating parameter item;

[0045] The stability index calculation management unit is used to receive the data in the safe operation index adjustment unit, calculate the stability index for each operating parameter item in combination with the frequency of abnormal feedback of each operating parameter item, identify the sensor device that samples the parameter value of the operating parameter item whose stability index is less than the index threshold, and send an information prompt to the management personnel terminal to adjust the sampling frequency of the corresponding sensor device.

[0046] Compared with the prior art, the beneficial effect of the present invention is that the present application can realize online acquisition of some parameters of the high-voltage voltage transformer in operation without affecting the normal operation of other locomotive equipment, and then analyze the data to realize accurate judgment of the operating status of the high-voltage voltage transformer in real time. The present application can maximize the life cycle of the high-voltage voltage transformer, greatly reduce the maintenance cost, effectively avoid the risk of high-voltage voltage transformer accidents, improve the maintenance efficiency and operation reliability, and improve the safe operation of the locomotive by building a method for online monitoring and intelligent early warning of the high-voltage voltage transformer. At the same time, the present application observes the parameter value changes presented by the sampling of the corresponding sensor device when the high-voltage voltage transformer operation alarm is approaching, and monitors and adjusts the synchronous sampling state of the sensor device in real time, thereby ensuring as much as possible that before the high-voltage voltage transformer is abnormal, the early warning of the abnormal state is intervened. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 A schematic flow chart of a method for monitoring performance of a high-voltage voltage transformer based on artificial intelligence according to the present invention;

[0048] Figure 2 The present invention is a structural schematic diagram of a high-voltage voltage transformer performance monitoring system based on artificial intelligence. DETAILED DESCRIPTION

[0049] Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in the field without making any creative work shall fall within the scope of protection of the present invention.

[0050] Example: Figure 1-Figure 2 As shown, the present invention provides a technical solution: a high-voltage voltage transformer performance monitoring method based on artificial intelligence, the method comprising:

[0051] Step S1: by deploying a number of sensor devices on the high-voltage voltage transformer of the locomotive, parameter values ​​of various operating parameter items of the high-voltage voltage transformer in operation are collected; by setting corresponding standard operating ranges for each operating parameter item, an online performance early warning monitoring platform for the high-voltage voltage transformer is established;

[0052] Step S2: sorting out the value changes of each operating parameter item when the corresponding sensor device samples the parameter value according to the corresponding sampling frequency, and identifying and extracting the characteristic fluctuation value of each operating parameter item;

[0053] Wherein, step S2 comprises:

[0054] Step S2-1: for each sensor device, randomly construct N parameter value groups from the parameter value change sequence distributed in chronological order obtained by collecting the corresponding operating parameter items according to the corresponding sampling frequency, wherein each parameter value group is composed of parameter values ​​obtained by two adjacent samplings;

[0055] Step S2-2: Assume that in a certain parameter value group, the parameter value corresponding to the earlier sampling time is Y 1 , the parameter value corresponding to the sampling time after is Y 2 , calculate the fluctuation value β=|Y in a certain parameter value group 2 -Y 1 |; For each operating parameter item, respectively, obtain the average value of the fluctuation values ​​extracted from the corresponding N parameter value groups, and set the average value as the characteristic fluctuation value presented by each operating parameter item when the corresponding sensor device samples the parameter value according to the corresponding sampling frequency;

[0056] Step S3: Traverse the characteristic fluctuation value and standard operation range corresponding to each operation parameter item, extract the relative operation range of each operation parameter item in the corresponding standard operation range through the characteristic fluctuation value of each operation parameter item, and evaluate and calculate the safety operation index of each operation parameter item according to the numerical value inclusion distribution phenomenon presented by the relative operation range and the corresponding standard operation range;

[0057] Wherein, step S3 comprises:

[0058] Step S3-1: If the standard operating interval corresponding to a certain operating parameter item is [A, B], the characteristic fluctuation value is α, where A is the minimum threshold in the standard operating interval, and B is the maximum threshold in the standard operating interval; capture the first numerical endpoint R(A)=A+α, and the second numerical endpoint R(B)=B-α in the standard operating interval [A, B]; extract the first numerical range P1=[A, R(A)] and the second numerical range P2=[R(B, B] for a certain operating parameter;

[0059] Step S3-2: When P1∩P2=∅, it is determined that a relative operating interval G=[R(A), R(B)] exists for a certain operating parameter item, and a safe operation index δ=[R(B)-R(A)] / [BA] is calculated for the certain operating parameter item, where 0<δ<1; when P1∩P2≠∅, it is determined that a relative operating interval does not exist for a certain operating parameter item, and the safe operation index δ of the certain operating parameter item is assigned a value of 0;

[0060] For example, the standard operating range of the voltage of the high-voltage voltage transformer is [80v, 120v], the characteristic fluctuation value α=10v, and the minimum threshold value A=80 and the maximum threshold value B=120 in the standard operating range are known;

[0061] In summary, the first numerical endpoint R(A)=80+10=90v and the second numerical endpoint R(B)=120-10=110v can be captured in the standard operating range [80v, 120v], and the first numerical range P1=[80, 90] and the second numerical range P2=[110, 120] of the voltage can be extracted. Because P1=[80, 90]∩P2=[110, 120]=∅, it is determined that the voltage has a relative operating range G=[90v, 110v];

[0062] At the same time, the voltage can be calculated to obtain the safe operation index δ=[110v-90v] / [120v-80v]=0.5;

[0063] Step S4: Collect historical fault monitoring and early warning records, obtain the operating parameter items with abnormalities in each historical fault monitoring and early warning record, and use the corresponding standard operating range as the judgment standard;

[0064] If a certain operating parameter item is an operating parameter item that is reported to be abnormal in a certain historical fault monitoring and early warning record, then the certain historical fault monitoring and early warning record is determined to be a target record of the certain operating parameter item;

[0065] Step S5: in each target record of each operating parameter item, by combing the parameter values ​​of each operating parameter item before and after the corresponding target record is generated and the numerical distribution phenomenon presented by the corresponding standard operating interval or relative operating interval, it is determined whether to perform feature marking processing on the corresponding target record;

[0066] Wherein, step S5 comprises:

[0067] Step S5-1: extracting target records corresponding to each operating parameter item, and setting each operating parameter item for which a relative operating interval is determined to exist as a first characteristic parameter item, and setting each operating parameter item for which a relative operating interval is determined not to exist as a second characteristic parameter item;

[0068] Step S5-2: If a sensor device that samples the parameter value of a first characteristic parameter item is extracted from a target record corresponding to a first characteristic parameter item, the parameter value d obtained by the most recent sampling of the first characteristic parameter item before the generation of the target record and the relative operation interval D corresponding to the first characteristic parameter item satisfy d∈D, a characteristic mark is made for the target record;

[0069] Step S5-3: If, in a target record corresponding to a second characteristic parameter item, an abnormal parameter value corresponding to a second characteristic parameter item when the target record is generated is extracted, and the minimum parameter deviation value presented between the abnormal parameter value and the standard operating interval corresponding to the second characteristic parameter item satisfies and is greater than the deviation threshold, a characteristic mark is made for the target record;

[0070] Step S6: adjusting the safe operation index of each operating parameter item according to the quantity distribution of the target records marked with features corresponding to each operating parameter item, and calculating the stability index of each operating parameter item in combination with the frequency of abnormal feedback of each operating parameter item, identifying the sensor device that samples the parameter value of the operating parameter item whose stability index is less than the index threshold, and sending a message prompting the administrator terminal to adjust the sampling frequency of the corresponding sensor device;

[0071] Wherein, step S6 comprises:

[0072] Step S6-1: Obtain the total number k of target records corresponding to each operating parameter item, and respectively accumulate the total number n of target records containing characteristic marks for each operating parameter item; obtain the total number M of historical fault monitoring and early warning records;

[0073] Step S6-2: extract the safe operation index δ of each operating parameter item, and calculate the influence adjustment coefficient θ=n / k for each operating parameter item; when the safe operation index δ of a certain operating parameter item is 0, adjust it to obtain a new safe operation index δ'=δ+1 / θ; when the safe operation index δ of a certain operating parameter item satisfies 0<δ<1, adjust it to obtain a new safe operation index δ'=δ θ ;

[0074] Step S6-3: Evaluate the stability index F=1 / (k / M)×δ' of each operating parameter item; set the operating parameter item whose corresponding stability index is less than the index threshold as the target parameter item; feature-mark the sensor device that samples the corresponding parameter value of each target parameter item, and send a message prompting the administrator terminal to adjust the sampling frequency of the sensor device containing the feature mark;

[0075] In order to better implement the above method, a high-voltage voltage transformer performance monitoring system is also proposed, which includes: a monitoring platform construction management module, a characteristic fluctuation value extraction module, a safe operation index evaluation module, a target record judgment module, a characteristic mark processing module, and a monitoring equipment prompt management module;

[0076] The monitoring platform builds a management module, which is used to deploy several sensor devices on the locomotive's high-voltage voltage transformer and collect parameter values ​​of various operating parameter items of the high-voltage voltage transformer in operation; by setting corresponding standard operating ranges for each operating parameter item, an online performance early warning monitoring platform for the high-voltage voltage transformer is built;

[0077] A characteristic fluctuation value extraction module is used to sort out the value changes of each operating parameter item when the corresponding sensor device samples the parameter value according to the corresponding sampling frequency, and identify and extract the characteristic fluctuation value of each operating parameter item;

[0078] The safety operation index evaluation module is used to traverse the characteristic fluctuation value and standard operation range corresponding to each operation parameter item, extract the relative operation range of each operation parameter item in the corresponding standard operation range through the characteristic fluctuation value of each operation parameter item, and evaluate and calculate the safety operation index of each operation parameter item according to the numerical value inclusion distribution phenomenon presented by the relative operation range and the corresponding standard operation range;

[0079] Among them, the safe operation index evaluation module includes: a relative operation interval extraction unit and a safe operation index calculation unit;

[0080] A relative operation interval extraction unit is used to traverse the characteristic fluctuation value and the standard operation interval corresponding to each operation parameter item, and extract the relative operation interval of each operation parameter item in the corresponding standard operation interval through the characteristic fluctuation value of each operation parameter item;

[0081] A safe operation index calculation unit is used to evaluate and calculate the safe operation index of each operating parameter item according to the numerical value inclusion distribution phenomenon presented by the relative operating range and the corresponding standard operating range;

[0082] The target record judgment module is used to collect historical fault monitoring and early warning records, obtain the operating parameter items with abnormal feedback in each historical fault monitoring and early warning record, and use the corresponding standard operating range as the judgment standard; if a certain operating parameter item is an operating parameter item with abnormal feedback in a certain historical fault monitoring and early warning record, then the certain historical fault monitoring and early warning record is judged as the target record of the certain operating parameter item;

[0083] The feature mark processing module is used to determine whether to perform feature mark processing on the corresponding target record by sorting out the parameter values ​​of each operating parameter item before and after the corresponding target record is generated and the numerical distribution phenomenon presented by the corresponding standard operating interval or relative operating interval in each target record of each operating parameter item in turn;

[0084] The monitoring equipment prompt management module is used to adjust the safe operation index of each operating parameter item according to the distribution of the number of target records marked with characteristics corresponding to each operating parameter item, and calculate the stability index of each operating parameter item in combination with the frequency of abnormal feedback of each operating parameter item, identify the sensor device that samples the parameter value of the operating parameter item whose stability index is less than the index threshold, and send an information prompt to the management terminal to adjust the sampling frequency of the corresponding sensor device;

[0085] Among them, the monitoring equipment prompt management module includes a safe operation index adjustment unit and a stability index calculation management unit;

[0086] A safe operation index adjustment unit, used to adjust the safe operation index of each operating parameter item according to the quantity distribution of target records marked with features corresponding to each operating parameter item;

[0087] The stability index calculation management unit is used to receive the data in the safe operation index adjustment unit, calculate the stability index for each operating parameter item in combination with the frequency of abnormal feedback of each operating parameter item, identify the sensor device that samples the parameter value of the operating parameter item whose stability index is less than the index threshold, and send an information prompt to the management personnel terminal to adjust the sampling frequency of the corresponding sensor device.

[0088] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A high-voltage voltage transformer performance monitoring method based on artificial intelligence, characterized in that: The method comprises: Step S1: by deploying a number of sensor devices on the high-voltage voltage transformer of the locomotive, parameter values ​​of various operating parameter items of the high-voltage voltage transformer in operation are collected; by setting corresponding standard operating ranges for each operating parameter item, an online performance early warning monitoring platform for the high-voltage voltage transformer is established; Step S2: sorting out the value changes of each operating parameter item when the corresponding sensor device samples the parameter value according to the corresponding sampling frequency, and identifying and extracting the characteristic fluctuation value of each operating parameter item; Step S3: Traverse the characteristic fluctuation value and standard operation range corresponding to each operation parameter item, extract the relative operation range of each operation parameter item in the corresponding standard operation range through the characteristic fluctuation value of each operation parameter item, and evaluate and calculate the safety operation index of each operation parameter item according to the numerical value inclusion distribution phenomenon presented by the relative operation range and the corresponding standard operation range; Step S4: Collect historical fault monitoring and early warning records, obtain the operating parameter items with abnormalities in each historical fault monitoring and early warning record, and use the corresponding standard operating range as the judgment standard; If a certain operating parameter item is an operating parameter item that is reported to be abnormal in a certain historical fault monitoring and early warning record, then the certain historical fault monitoring and early warning record is determined to be a target record of the certain operating parameter item; Step S5: in each target record of each operating parameter item, by combing the parameter values ​​of each operating parameter item before and after the corresponding target record is generated and the numerical distribution phenomenon presented by the corresponding standard operating interval or relative operating interval, it is determined whether to perform feature marking processing on the corresponding target record; Step S6: adjusting the safe operation index of each operating parameter item according to the quantity distribution of the target records marked with features corresponding to each operating parameter item, and calculating the stability index of each operating parameter item in combination with the frequency of abnormal feedback of each operating parameter item, identifying the sensor device that samples the parameter value of the operating parameter item whose stability index is less than the index threshold, and sending a message prompting the administrator terminal to adjust the sampling frequency of the corresponding sensor device; The step S2 comprises: Step S2-1: for each sensor device, randomly construct N parameter value groups from the parameter value change sequence distributed in chronological order obtained by collecting the corresponding operating parameter items according to the corresponding sampling frequency, wherein each parameter value group is composed of parameter values ​​obtained by two adjacent samplings; Step S2-2: Assume that in a certain parameter value group, the parameter value corresponding to the earlier sampling time is Y 1 , the parameter value corresponding to the sampling time after is Y 2 , calculate the fluctuation value β=|Y in a certain parameter value group 2 -Y 1 |; For each operating parameter item, respectively, obtain the average value of the fluctuation values ​​extracted from the corresponding N parameter value groups, and set the average value as the characteristic fluctuation value presented by each operating parameter item when the corresponding sensor device samples the parameter value according to the corresponding sampling frequency; The step S3 comprises: Step S3-1: If the standard operating interval corresponding to a certain operating parameter item is [A, B], the characteristic fluctuation value is α, wherein A is the minimum threshold value in the standard operating interval, and B is the maximum threshold value in the standard operating interval; capture the first numerical endpoint R(A)=A+α, and the second numerical endpoint R(B)=B-α in the standard operating interval [A, B]; extract the first numerical range P1=[A, R(A)] and the second numerical range P2=[R(B, B] for the certain operating parameter; Step S3-2: When P1∩P2=∅, it is determined that a certain operating parameter item has a relative operating interval G=[R(A), R(B)], and at the same time, the safe operation index δ=[R(B)-R(A)] / [BA] is calculated for the certain operating parameter item, where 0<δ<1; when P1∩P2≠∅, it is determined that a certain operating parameter item does not have a relative operating interval, and at the same time, the safe operation index δ of the certain operating parameter item is assigned a value of 0.

2. The method for monitoring high voltage voltage transformer performance based on artificial intelligence according to claim 1, characterized in that: The step S5 comprises: Step S5-1: extracting target records corresponding to each operating parameter item, and setting each operating parameter item for which a relative operating interval is determined to exist as a first characteristic parameter item, and setting each operating parameter item for which a relative operating interval is determined not to exist as a second characteristic parameter item; Step S5-2: If a sensor device that samples the parameter value of a certain first characteristic parameter item is extracted from a certain target record corresponding to the certain first characteristic parameter item, the parameter value d obtained by the most recent sampling of the certain first characteristic parameter item before the certain target record is generated and the relative operation interval D corresponding to the certain first characteristic parameter item satisfy d∈D, the certain target record is marked as a characteristic; Step S5-3: If in a target record corresponding to a second characteristic parameter item, an abnormal parameter value corresponding to a second characteristic parameter item when the target record is generated is extracted, and the minimum parameter deviation value between the abnormal parameter value and the standard operating range corresponding to the second characteristic parameter item satisfies and is greater than the deviation threshold, the target record is marked as a feature.

3. The method for monitoring high voltage voltage transformer performance based on artificial intelligence according to claim 2 is characterized in that: The step S6 comprises: Step S6-1: Obtain the total number k of target records corresponding to each operating parameter item, and respectively accumulate the total number n of target records containing characteristic marks for each operating parameter item; obtain the total number M of historical fault monitoring and early warning records; Step S6-2: extract the safe operation index δ of each operating parameter item, and calculate the influence adjustment coefficient θ=n / k for each operating parameter item; when the safe operation index δ of a certain operating parameter item is 0, adjust it to obtain a new safe operation index δ'=δ+1 / θ; when the safe operation index δ of a certain operating parameter item satisfies 0<δ<1, adjust it to obtain a new safe operation index δ'=δ θ ; Step S6-3: Evaluate the stability index F=1 / (k / M)×δ' of each operating parameter item; set the operating parameter item whose corresponding stability index is less than the index threshold as the target parameter item; mark the sensor device that samples the corresponding parameter value of each target parameter item with a characteristic mark, and send an information prompt to the management terminal to adjust the sampling frequency of the sensor device containing the characteristic mark.

4. A high-voltage voltage transformer performance monitoring system, used to execute the high-voltage voltage transformer performance monitoring method based on artificial intelligence as described in any one of claims 1 to 3, characterized in that: The system includes: a monitoring platform construction management module, a characteristic fluctuation value extraction module, a safe operation index evaluation module, a target record judgment module, a characteristic mark processing module, and a monitoring equipment prompt management module; The monitoring platform constructs a management module, which is used to deploy a number of sensor devices on the high-voltage voltage transformer of the locomotive, collect parameter values ​​of various operating parameter items of the high-voltage voltage transformer in operation; by setting corresponding standard operating ranges for each operating parameter item, an online performance early warning monitoring platform for the high-voltage voltage transformer is built; The characteristic fluctuation value extraction module is used to sort out the value changes of each operating parameter item when the corresponding sensor device samples the parameter value according to the corresponding sampling frequency, and identify and extract the characteristic fluctuation value of each operating parameter item; The safety operation index evaluation module is used to traverse the characteristic fluctuation value and the standard operation interval corresponding to each operation parameter item, extract the relative operation interval of each operation parameter item in the corresponding standard operation interval through the characteristic fluctuation value of each operation parameter item, and evaluate and calculate the safety operation index of each operation parameter item according to the numerical value inclusion distribution phenomenon presented by the relative operation interval and the corresponding standard operation interval; The target record judgment module is used to collect historical fault monitoring and early warning records, and obtain the operating parameter items with abnormal feedback in each historical fault monitoring and early warning record, taking the corresponding standard operating range as the judgment standard; if a certain operating parameter item is an operating parameter item with abnormal feedback in a certain historical fault monitoring and early warning record, then the certain historical fault monitoring and early warning record is judged as the target record of the certain operating parameter item; The feature mark processing module is used to determine whether to perform feature mark processing on the corresponding target record by sorting out the parameter values ​​of each operating parameter item before and after the corresponding target record is generated and the numerical distribution phenomenon presented by the corresponding standard operating interval or relative operating interval in each target record of each operating parameter item in turn; The monitoring equipment prompt management module is used to adjust the safe operation index of each operating parameter item according to the distribution of the number of target records marked with characteristics for each operating parameter item, and at the same time calculate the stability index of each operating parameter item in combination with the frequency of abnormalities reported for each operating parameter item, identify the sensor device that samples the parameter value of the operating parameter item whose stability index is less than the index threshold, and send an information prompt to the management personnel terminal to adjust the sampling frequency of the corresponding sensor device.

5. A high voltage voltage transformer performance monitoring system according to claim 4, characterized in that: The safe operation index evaluation module includes: a relative operation interval extraction unit and a safe operation index calculation unit; The relative operation interval extraction unit is used to traverse the characteristic fluctuation value and the standard operation interval corresponding to each operation parameter item, and extract the relative operation interval of each operation parameter item in the corresponding standard operation interval according to the characteristic fluctuation value of each operation parameter item; The safety operation index calculation unit is used to evaluate and calculate the safety operation index of each operating parameter item according to the distribution phenomenon of the numerical values ​​presented by the relative operating range and the corresponding standard operating range.

6. A high voltage voltage transformer performance monitoring system according to claim 4, characterized in that: The monitoring equipment prompt management module includes a safe operation index adjustment unit and a stability index calculation management unit; The safe operation index adjustment unit is used to adjust the safe operation index of each operating parameter item according to the quantity distribution of the target records marked with features corresponding to each operating parameter item; The stability index calculation management unit is used to receive the data in the safe operation index adjustment unit, calculate the stability index for each operating parameter item in combination with the frequency of abnormalities reported for each operating parameter item, identify the sensor device that samples the parameter values ​​of the operating parameter items whose stability index is less than the index threshold, and send an information prompt to the management personnel terminal to adjust the sampling frequency of the corresponding sensor device.

Citation Information

Patent Citations

  • A vertical stabilization device, a control method thereof and an image acquisition device

    CN110337622A

  • Transformer state monitoring system based on Internet of Things

    CN113049035A

  • Reactor abnormal working condition diagnosis and control system

    CN113963826A

  • Transformer insulating sleeve fault diagnosis method and device and computer equipment

    CN115629253A

  • Renewable energy-based hydrogen production electrolytic cell electrode deterioration prevention and control system and method

    CN116312836A