Box-type substation based on Internet of Things

Abstract

The invention relates to the technical field of transformer substations, in particular to a box-type transformer substation based on the Internet of Things, which comprises a monitoring module used for judging a monitoring mode according to a set condition so as to obtain a plurality of monitoring points; the analysis module is used for determining the monitoring state of each monitoring point according to the trigger threshold value and the coupling coefficient and determining an analysis mode according to the monitoring state; the trigger analysis module is used for determining a compensation coefficient according to the initial deviation index and the topological connection coefficient and determining a fault threshold value based on the compensation coefficient and an initial fault threshold value in trigger analysis; the interval analysis module is used for determining the length of a time interval according to the comparison correlation degree in interval analysis and determining an interval analysis mode according to the line correlation degree so as to determine a fault threshold value; the early warning module is used for sending an early warning to a user when the fault threshold value is greater than a preset fault threshold value; according to the invention, the fault diagnosis precision of the box-type substation can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of substations, and in particular to a box-type substation based on the Internet of Things. Background Art

[0002] Due to the problems of remote installation location, harsh operating environment and many types of equipment in box-type substations, their daily operation and maintenance are difficult. The Internet of Things technology can meet the information collection needs of substation sites, but how to analyze the monitoring data to improve the accuracy of fault warning is a technical problem that technicians in this field urgently need to solve.

[0003] Chinese Patent Publication No. CN106059086A discloses a box-type substation current-carrying fault prediction system and its application, which includes a process layer, an interval layer, and a station control layer; the process layer is provided with a box-type substation online monitoring terminal, which is connected to the temperature measurement point of the wired temperature measurement module, and the box-type substation online monitoring terminal is also connected to the temperature measurement point of the wireless temperature measurement module via a wireless personal area network; the interval layer is provided with a box-type substation online monitoring IED, which is connected to the box-type substation online monitoring terminal; the station control layer is provided with a station-end monitoring unit, which is connected to the box-type substation online monitoring IED. It can be seen that the above technical solution has the following problems: fault warning is only based on temperature, and the warning method cannot be adaptively adjusted according to the actual status of the monitoring point, and the accuracy of fault diagnosis is poor. Summary of the Invention

[0004] To this end, the present invention provides a box-type substation based on the Internet of Things to overcome the problems in the prior art.

[0005] To achieve the above objectives, the present invention provides a box-type substation based on the Internet of Things, comprising:

[0006] The monitoring module is used to determine the monitoring mode according to the set conditions to obtain a number of monitoring points. The monitoring mode is uniform partition monitoring or associated partition monitoring;

[0007] An analysis module, connected to the monitoring module, for determining the monitoring status of each monitoring point according to the trigger threshold and the coupling coefficient, and determining whether the analysis mode is trigger analysis or interval analysis according to the monitoring status;

[0008] a trigger analysis module connected to the analysis module, configured to determine, in the trigger analysis, based on analysis conditions, whether the video frame acquisition mode is feature acquisition based on a combination of feature segments or uniform acquisition based on a periodic evaluation value, determine an initial fault threshold based on a video frame evaluation coefficient and a feature dimension abnormality, determine a compensation coefficient based on an initial deviation index and a topological connection coefficient, and determine a fault threshold based on the compensation coefficient and the initial fault threshold;

[0009] An interval analysis module, connected to the analysis module, is used to determine the length of the time interval according to the comparison correlation degree in the interval analysis, and to determine the interval analysis method according to the line correlation degree to determine the fault threshold. The interval analysis method is a division comparison analysis or anomaly fusion analysis;

[0010] An early warning module is connected to the trigger analysis module and the interval analysis module respectively, and is used to send an early warning to the user when the fault threshold is greater than a preset fault threshold.

[0011] Furthermore, the monitoring module determines the monitoring mode according to the set conditions, including:

[0012] The monitoring module responds to the setting condition that the architecture clutter is greater than or equal to the preset architecture clutter or the site instability is greater than or equal to the preset site instability, then the monitoring mode is determined to be associated partition monitoring;

[0013] The setting condition for the monitoring module response is that when the architecture clutter is less than the preset architecture clutter and the site instability is less than the preset site instability, the monitoring mode is determined to be uniform partition monitoring.

[0014] Furthermore, the analysis module determines that the analysis mode is trigger analysis when the monitoring state is that the trigger threshold is greater than or equal to the preset trigger threshold or the coupling coefficient is greater than or equal to the preset coupling coefficient.

[0015] Furthermore, when the monitoring state is that the trigger threshold is less than the preset trigger threshold and the coupling coefficient is less than the preset coupling coefficient, the analysis module determines that the analysis method is interval analysis.

[0016] Furthermore, the trigger analysis module determines the video frame acquisition mode according to the analysis conditions, including:

[0017] The analysis condition that triggers the analysis module to respond is that when the end-side correlation degree is less than a preset end-side correlation degree or the data fluctuation coefficient is greater than or equal to a preset data fluctuation coefficient, the video frame acquisition mode is determined to be feature acquisition based on the characteristic segment combination;

[0018] The analysis condition triggering the analysis module to respond is that when the end-side correlation is greater than or equal to the preset end-side correlation and the data fluctuation coefficient is less than the preset data fluctuation coefficient, the video frame acquisition method is determined to be uniform acquisition based on the periodic evaluation value.

[0019] Furthermore, the judgment condition for triggering the analysis module to respond is that when the initial deviation index is greater than or equal to a preset initial deviation index and the topological connection coefficient is greater than or equal to a preset topological connection coefficient, the compensation coefficient is determined based on the supplementary dimension.

[0020] Furthermore, the judgment condition for triggering the analysis module to respond is that when the initial deviation index is less than a preset initial deviation index or the topological connection coefficient is less than a preset topological connection coefficient, the compensation coefficient is determined based on the supplementary video frame.

[0021] Furthermore, the interval analysis module determines the length of the time interval according to the comparison correlation degree;

[0022] The length of the time interval is positively correlated with the comparison correlation degree.

[0023] Furthermore, when the line correlation degree is less than a preset line correlation degree, the interval analysis module determines that the interval analysis method is a division comparison analysis;

[0024] In the division and comparison analysis, the interval division is performed according to the interval stability, and the fault threshold is determined according to the gradient variation value and the variation interaction coefficient.

[0025] Furthermore, the interval analysis module determines that the interval analysis mode is abnormal fusion analysis when the line correlation degree is greater than or equal to the preset line correlation degree;

[0026] In the abnormal fusion analysis, the fusion monitoring points are selected according to the sub-coupling coefficient, and the fault threshold is determined according to the benchmark abnormality.

[0027] Compared with the prior art, the beneficial effect of the present invention lies in that, in the technical solution of the present invention, the layout complexity and operation status of the box-type substation are effectively reflected by setting conditions, and then different monitoring methods are adaptively selected according to the setting conditions, so that the selection of monitoring methods is more in line with the actual application scenario. The use of associated partition monitoring can concentrate resources to carry out detailed monitoring of key areas, which is helpful to timely discover potential faults and abnormal conditions. The use of uniform partition monitoring can ensure that all areas are evenly and effectively monitored, which not only improves the pertinence and effectiveness of monitoring, but also helps to optimize resource allocation, reduce monitoring costs, and enhance the timeliness and accuracy of fault warnings.

[0028] Furthermore, the present invention determines the monitoring status of each monitoring point based on the trigger threshold and the coupling coefficient. The trigger threshold and the coupling coefficient effectively reflect the abnormality of the monitoring indicators and the degree of mutual influence between the monitoring points, and then adaptively selects different analysis methods according to the monitoring status. The trigger analysis can quickly respond to abnormal fluctuations of the monitoring points, and the interval analysis can identify hidden faults. It can timely detect fault signs and issue early warnings, thereby ensuring the stable operation of the box-type substation.

[0029] Furthermore, the present invention effectively reflects the degree of correlation between the video frames on both sides and the degree of fluctuation of the monitoring parameters through analysis conditions, and then adaptively selects different video frame acquisition methods according to the analysis conditions. A feature acquisition method is adopted to focus on key feature parts, avoid collecting irrelevant information, and improve acquisition efficiency; a uniform acquisition method is adopted to ensure that the collected data is highly relevant to the current monitoring needs, providing a high-quality data foundation for subsequent accurate analysis.

[0030] Furthermore, the present invention effectively reflects the degree of anomaly and topological connection status of the accurately captured data through the judgment conditions, and then flexibly selects the supplementary dimensions or supplementary video frames according to the judgment conditions to determine the compensation coefficient. It can comprehensively consider the outlier degree of the supplementary dimensions or the influence deviation of the supplementary video frames, compensate for the initial fault threshold, make the fault threshold closer to the actual fault situation, reduce the misjudgment and missed judgment caused by single indicator judgment, and improve the accuracy of fault warning.

[0031] Furthermore, the present invention determines the length of the time interval by comparing the correlation degree, which can make the analysis more focused on the key time period and improve the accuracy of fault diagnosis. In the divided comparison analysis, it can timely reflect the data change trend and characteristic distribution within the interval, providing a timely and accurate basis for fault warning; in the abnormal fusion analysis, the key monitoring points are selected by the sub-coupling coefficient, and the overall abnormality degree is quickly evaluated using the benchmark abnormality, so that fault warnings are issued in time and the fault handling time is reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 This is a module connection diagram of the box-type substation based on the Internet of Things of the present invention;

[0033] Figure 2 This is a flow chart of the present invention for determining a monitoring mode according to setting conditions;

[0034] Figure 3 This is a flow chart of the present invention for determining an analysis method based on a monitoring state;

[0035] Figure 4 This is a flow chart of the present invention's method for determining interval analysis based on line correlation. DETAILED DESCRIPTION

[0036] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are merely used to explain the present invention and are not intended to limit the present invention.

[0037] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0038] It should be noted that, in the description of the present invention, terms such as "up", "down", "left", "right", "inside", and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the accompanying drawings. This is only for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it cannot be understood as a limitation on the present invention.

[0039] Furthermore, it should be noted that, in the description of the present invention, unless otherwise expressly specified or limited, the terms "mounted," "connected," and "connected" should be understood in a broad sense. For example, they may refer to fixed connections, detachable connections, or integral connections; mechanical connections or electrical connections; direct connections or indirect connections through an intermediate medium; and internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0040] See also Figures 1 to 4 As shown, the present invention provides a box-type substation based on the Internet of Things, comprising:

[0041] The monitoring module is used to determine the monitoring mode according to the set conditions to obtain a number of monitoring points. The monitoring mode is uniform partition monitoring or associated partition monitoring;

[0042] An analysis module, connected to the monitoring module, for determining the monitoring status of each monitoring point according to the trigger threshold and the coupling coefficient, and determining whether the analysis mode is trigger analysis or interval analysis according to the monitoring status;

[0043] a trigger analysis module connected to the analysis module, configured to determine, in the trigger analysis, based on analysis conditions, whether the video frame acquisition mode is feature acquisition based on a combination of feature segments or uniform acquisition based on a periodic evaluation value, determine an initial fault threshold based on a video frame evaluation coefficient and a feature dimension abnormality, determine a compensation coefficient based on an initial deviation index and a topological connection coefficient, and determine a fault threshold based on the compensation coefficient and the initial fault threshold;

[0044] An interval analysis module, connected to the analysis module, is used to determine the length of the time interval according to the comparison correlation degree in the interval analysis, and to determine the interval analysis method according to the line correlation degree to determine the fault threshold. The interval analysis method is a division comparison analysis or anomaly fusion analysis;

[0045] An early warning module is connected to the trigger analysis module and the interval analysis module respectively, and is used to send an early warning to the user when the fault threshold is greater than a preset fault threshold.

[0046] The application scenario of the present invention is fault warning of a box-type substation. The present invention further provides a data acquisition module, which is respectively connected to the monitoring module, the analysis module, and the trigger analysis module to collect monitoring data and video data. The monitoring data and video data are both transmitted to the data acquisition module via an Internet of Things network. The Internet of Things network includes but is not limited to LoRa, ZigBee, and NB-IoT. Single monitoring data includes monitoring parameters monitored in real time by a single monitoring point over the usage time within a single complete monitoring cycle. The monitoring parameters include but are not limited to temperature, humidity, partial discharge intensity, current, and voltage. The temperature, humidity, partial discharge intensity, current, and voltage are measured by a temperature sensor, a humidity sensor, a pulse current sensor, and a voltage transformer, respectively. The video data is a video collected in real time by a single monitoring point over the usage time within a single complete monitoring cycle using an ESP32-CAM camera. This is content that is easy for those skilled in the art to understand and will not be described in detail.

[0047] The present invention is provided with a continuous cycle monitoring cycle. At the end of each monitoring cycle, the monitoring status is determined once. The duration of the monitoring cycle can be set according to the needs of the user. The greater the user's demand for monitoring accuracy, the shorter the duration of the monitoring cycle. A value of the monitoring cycle is provided, and the monitoring cycle is 10 minutes.

[0048] The present invention is provided with several historical records, and any one of the historical records records the fault threshold, line complexity, density complexity, historical fault frequency, fault distribution threshold, architecture clutter, and site instability, etc., in the historical process of at least one box-type substation fault warning, and each historical record corresponds to a qualified mark, which records whether the fault warning process of the box-type substation meets the user's needs. The qualified mark can be recorded manually. It can be understood that the user can determine whether the fault warning process of the box-type substation meets the needs based on the self-set indicators. The self-set indicators can be but not limited to the warning error rate, which will not be elaborated here. Among them, the warning error rate is the number of times no warning is issued when a fault occurs in the box-type substation;

[0049] The value of the preset fault threshold can be determined by the user according to the actual application scenario. The greater the user's demand for ensuring the safe operation of the box-type substation, the smaller the value of the preset fault threshold. A method for determining the value of the preset fault threshold is provided. The detection and early warning module sends a historical record of early warnings to the user, and the minimum value of the fault threshold corresponding to the historical record that can meet the user's needs is recorded as the preset fault threshold;

[0050] The present invention sets a target coefficient and a related threshold value, and the corresponding relationship between the target coefficient and the related threshold value is expressed by a weight formula. The weight formula is: target coefficient = weight coefficient × related threshold value. Specifically, the present invention records the number of divided sub-areas, the number of divided monitoring areas, the value of w, the length of the time interval, the value of a and the fault threshold as the target coefficient, and records the volume of the box-type substation, the comprehensive evaluation coefficient, the cycle evaluation value, the comparison correlation degree, the interval stability degree and the benchmark abnormality degree as the related threshold value. It can be understood that the target coefficients all have corresponding related threshold values. For example, the divided sub-areas There is a positive correlation between the number of regions and the volume of the box-type substation, and the positive correlation between the number of divided sub-regions and the volume of the box-type substation is expressed through the weight formula. The value of the weight coefficient can be determined according to the user's historical experience based on the accuracy requirement of the density complexity, and the value of the weight coefficient can be optimized based on the historical record of the box-type substation fault warning combined with the multi-layer perceptron. The use of the multi-layer perceptron to optimize the value of the weight coefficient is easy for technical personnel in this field to understand, and will not be elaborated on. The value principles of the weight coefficients corresponding to other target coefficients and related thresholds are the same, and will not be elaborated on here.

[0051] Specifically, the monitoring module determines the monitoring mode according to the set conditions, including:

[0052] The monitoring module responds to the setting condition that the architecture clutter is greater than or equal to the preset architecture clutter or the site instability is greater than or equal to the preset site instability, then the monitoring mode is determined to be associated partition monitoring;

[0053] The setting condition for the monitoring module response is that when the architecture clutter is less than the preset architecture clutter and the site instability is less than the preset site instability, the monitoring mode is determined to be uniform partition monitoring.

[0054] Among them, the box-type substation contains several electrical equipment, including but not limited to high-voltage circuit breakers, high-voltage disconnectors, high-voltage load switches, high-voltage fuses, power transformers and power capacitors. A single electrical equipment is connected to other electrical equipment through a connecting line. A single connecting line is a busbar or cable connecting two electrical equipment. This is easy to understand for those skilled in the art and will not be described in detail.

[0055] The substation is divided into several cubic sub-areas with the same volume and shape. The number of sub-areas is positively correlated with the volume of the box-type substation.

[0056] Architecture complexity = line complexity / preset line complexity + density complexity / preset density complexity. Line complexity is the average value of the line tortuosity of each connecting line in the box-type substation. The line tortuosity of a single connecting line = the length of the connecting line / the shortest distance between the two electrical devices connected by the connecting line. Density complexity is the standard deviation of the equipment reference value corresponding to each sub-area. The equipment reference value for a single sub-area is the total number of electrical devices in that sub-area.

[0057] The values ​​of the preset route complexity and preset density complexity can be determined by the user based on the actual application scenario. The greater the user's requirement for the accuracy of capturing the actual complexity changes, the smaller the values ​​of the preset route complexity and preset density complexity will be. The average value of the route complexity and the average value of the density complexity corresponding to the historical records that meet the user's requirements are extracted and recorded as the preset route complexity and preset density complexity respectively.

[0058] Site instability = historical fault frequency / preset historical fault frequency + fault distribution threshold / preset fault distribution threshold;

[0059] The historical fault frequency is the number of faults that have occurred in the box-type substation since its construction. The sub-area where the fault occurred is recorded as the fault sub-area. The fault distribution threshold is the average of the distance thresholds corresponding to each fault sub-area. The distance threshold corresponding to a single fault sub-area is the average of the shortest distances from the fault sub-area to other fault sub-areas.

[0060] The values ​​of the preset historical fault frequency and the preset fault distribution threshold can be determined by the user according to the actual application scenario. The greater the user's requirement for the accuracy of capturing the actual substation stability change, the smaller the values ​​of the preset historical fault frequency and the preset fault distribution threshold. The average value of the corresponding fault frequency and the average value of the fault distribution threshold in the historical records that meet the user's requirements are extracted and recorded as the preset fault frequency and the preset fault distribution threshold respectively;

[0061] The values ​​of the preset architecture clutter and the preset site instability can be determined by the user based on the actual application scenario. The greater the user's demand for improved monitoring accuracy, the smaller the values ​​of the preset architecture clutter and the preset site instability. A method for determining the values ​​of the preset architecture clutter and the preset site instability is provided. The historical records of the user performing uniform partition monitoring are detected, and the average values ​​of the architecture clutter and the average values ​​of the site instability corresponding to the historical records that meet the user's requirements are recorded as the preset architecture clutter and the preset site instability, respectively.

[0062] Association partition monitoring includes: performing association analysis on each electrical device, when performing association analysis on a single electrical device, recording the electrical device as the target electrical device, recording other electrical devices other than the target electrical device that are not included in the associated electrical device combination as reference electrical devices, constructing a combination of the reference electrical devices and the target electrical device whose impact similarity with the target electrical device is less than a preset impact similarity and whose distance reference value is less than a preset distance reference value, and recording the combination as an associated electrical device combination, and continuing to perform association analysis on electrical devices that are not included in the associated electrical device combination until all electrical devices are included in the associated electrical device combination, recording the smallest cubic area that can contain each electrical device in a single associated electrical device combination as a monitoring area; during association partition monitoring, the number of monitoring points corresponding to a single monitoring area is the smallest integer greater than or equal to m1, m1 = (monitoring reference value corresponding to the monitoring area / preset monitoring reference value) × the average number of monitoring points corresponding to each monitoring area in the historical records that can meet user needs; the user can set the location of each monitoring point at the location of each electrical device according to actual needs, without specific restrictions;

[0063] The monitoring reference value corresponding to a single monitoring area = ln(the number of fault sub-areas that the monitoring area passes through × the average value of the line reference values ​​corresponding to each electrical device in the monitoring area + 1). The line reference value corresponding to a single electrical device is the average value of the line tortuosity coefficients corresponding to each connecting line of the electrical device;

[0064] The value of the preset monitoring reference value can be determined by the user based on historical records. The greater the user's demand for monitoring accuracy, the smaller the value of the preset monitoring reference value. The average value of the monitoring reference values ​​corresponding to each monitoring area in the historical records that can meet the user's needs is recorded as the preset monitoring reference value;

[0065] For any two electrical devices, the impact similarity = 1 / (the absolute value of the difference in the device impacts of the two devices + 1). The device impact of a single electrical device = the total number of failures in the sub-area that the electrical device has passed through since its construction to the present / the number of connecting lines corresponding to the electrical device. The reference distance between any two electrical devices is the shortest distance between the two devices.

[0066] The values ​​of the preset impact similarity and the preset distance reference value can be determined by the user according to the actual application scenario. The greater the user's demand for improving monitoring accuracy, the greater the value of the preset impact similarity and the smaller the value of the preset distance reference value. A method for determining the values ​​of the preset impact similarity and the preset distance reference value is provided. The average value of the reference impact similarity and the average value of the reference distance reference value corresponding to the historical records that meet the user's needs are extracted, and are recorded as the preset impact similarity and the preset distance reference value respectively. The reference impact similarity is the impact similarity corresponding to any two electrical devices in a single associated electrical equipment combination, and the reference distance reference value is the distance reference value corresponding to any two electrical devices in a single associated electrical equipment combination;

[0067] Uniform partition monitoring includes: dividing the substation into several cubic monitoring areas of the same volume and shape. The number of divided monitoring areas is positively correlated with the comprehensive evaluation coefficient. During uniform partition monitoring, a monitoring point is set in each monitoring area. The location of the monitoring point can be set by the user at any electrical equipment location according to actual needs, with no specific restrictions.

[0068] Comprehensive evaluation coefficient = architecture complexity / preset architecture complexity + site instability / preset site instability.

[0069] Specifically, the analysis module determines that the analysis mode is trigger analysis when the monitoring state is that the trigger threshold is greater than or equal to the preset trigger threshold or the coupling coefficient is greater than or equal to the preset coupling coefficient.

[0070] The monitoring period that is before the current monitoring period and adjacent to the current monitoring period is recorded as the target monitoring period;

[0071] The trigger threshold corresponding to a single monitoring point is the monitoring abnormality of the characteristic monitoring parameter corresponding to the monitoring point in the target monitoring period;

[0072] The characteristic monitoring parameter corresponding to the target monitoring period is the monitoring parameter with the largest monitoring anomaly in the target monitoring period. The monitoring anomaly corresponding to a single monitoring parameter = the monitoring extreme value corresponding to the monitoring parameter in the target monitoring period / the average value of the monitoring extreme values ​​corresponding to the monitoring parameter in each historical monitoring period + the fluctuation threshold value corresponding to the monitoring parameter in the target monitoring period / the average value of the fluctuation threshold value corresponding to the monitoring parameter in each historical monitoring period;

[0073] The historical monitoring period is the monitoring period before the target monitoring period. The time point is set by the user. A method for setting the time point is provided. The starting time of a single monitoring period is used as the starting point, and an interval point is set every 5 seconds in chronological order. The starting point and each interval point are recorded as time points.

[0074] The monitoring extreme value corresponding to a single monitoring parameter in a single monitoring cycle is the maximum value among the values ​​of the monitoring parameter corresponding to each time point in the monitoring cycle; the fluctuation threshold corresponding to a single monitoring parameter in a single monitoring cycle is the standard deviation among the values ​​of the monitoring parameter corresponding to each time point in the monitoring cycle;

[0075] The monitoring points with a trigger coefficient greater than the preset trigger coefficient are recorded as analysis monitoring points. The coupling coefficient corresponding to a single monitoring point = the average of the sub-coupling coefficients corresponding to the monitoring point and each analysis monitoring point. The sub-coupling coefficient corresponding to the monitoring point and the single analysis monitoring point = 1 / (connectivity coefficient corresponding to the two monitoring points × the shortest distance between the two monitoring points). The connectivity coefficient corresponding to the two monitoring points is the minimum value of the number of connecting lines corresponding to each path from one monitoring point to another monitoring point via the connecting line.

[0076] The values ​​of the preset trigger threshold and the preset coupling coefficient can be determined by the user according to the actual application scenario. The greater the user's demand for trigger analysis, the smaller the values ​​of the preset trigger threshold and the preset coupling coefficient. A method for determining the values ​​of the preset trigger threshold and the preset coupling coefficient is provided. The historical records of the user's trigger analysis are detected, and the average values ​​of the trigger threshold and the average values ​​of the coupling coefficient corresponding to the historical records that can meet the user's needs are recorded as the preset trigger threshold and the preset coupling coefficient respectively.

[0077] Specifically, when the monitoring state is that the trigger threshold is less than the preset trigger threshold and the coupling coefficient is less than the preset coupling coefficient, the analysis module determines that the analysis method is interval analysis.

[0078] Specifically, the trigger analysis module determines the video frame acquisition mode according to the analysis conditions, including:

[0079] The analysis condition that triggers the analysis module to respond is that when the end-side correlation degree is less than a preset end-side correlation degree or the data fluctuation coefficient is greater than or equal to a preset data fluctuation coefficient, the video frame acquisition mode is determined to be feature acquisition based on the characteristic segment combination;

[0080] The analysis condition triggering the analysis module to respond is that when the end-side correlation is greater than or equal to the preset end-side correlation and the data fluctuation coefficient is less than the preset data fluctuation coefficient, the video frame acquisition method is determined to be uniform acquisition based on the periodic evaluation value.

[0081] The device-side correlation degree = the number of matching pixels between the video frame corresponding to the first time point and the video frame corresponding to the last time point in the video data corresponding to a single monitoring point in the target monitoring period / the total number of pixels in a single video frame.

[0082] Perform image registration on the video frames of the first and last time points of a single monitoring point in the target monitoring cycle so that they are aligned to the same coordinate system. This is well understood by those skilled in the art and will not be described in detail. The matching pixels corresponding to the two video frames are the pixels at the same position and with the same pixel value in the two video frames after registration.

[0083] The data fluctuation coefficient is the maximum value among the fluctuation thresholds corresponding to each monitoring parameter in the target monitoring period;

[0084] The values ​​of the preset end-side correlation degree and the preset data fluctuation coefficient can be determined by the user based on the actual application scenario. The greater the user's demand for improved collection accuracy, the larger the value of the preset end-side correlation degree and the smaller the value of the preset data fluctuation coefficient. A method for determining the values ​​of the preset end-side correlation degree and the preset data fluctuation coefficient is provided. The method detects historical records uniformly collected by the user based on periodic evaluation values, and records the average values ​​of the end-side correlation degree and the average values ​​of the data fluctuation coefficient corresponding to the historical records that meet the user's requirements as the preset end-side correlation degree and the preset data fluctuation coefficient, respectively.

[0085] In the feature collection according to the feature paragraph combination, a combination analysis is performed on each feature paragraph in the target monitoring cycle in the order of time from early to late. When the combination analysis is performed on a single feature paragraph, the feature paragraph is recorded as the target feature paragraph, and the feature paragraphs other than the target feature paragraph that are not recorded in the feature paragraph combination are recorded as reference feature paragraphs. A set is constructed of the reference feature paragraphs and the target feature paragraphs whose matching threshold with the target feature paragraph is greater than the preset matching threshold. The set is recorded as a feature paragraph combination, and the association analysis is continued for each feature paragraph that is not recorded in the feature paragraph combination until all feature paragraphs are recorded in the feature paragraph combination, and the association analysis is stopped. A video frame corresponding to any time point in each feature paragraph combination is randomly selected as a collected video frame, and each feature paragraph combination corresponds to one collected video frame.

[0086] For a single monitoring point, a feature image is drawn with the monitoring time corresponding to the target monitoring period as the horizontal axis and the feature monitoring parameter value as the vertical axis. The extreme value points in the feature image are used as isolation points to obtain several feature segments. A single feature segment is a continuous data interval in the feature image containing the values ​​of the feature monitoring parameters corresponding to each time point between the moments corresponding to two adjacent isolation points.

[0087] The matching threshold corresponding to any two feature paragraphs = 1 / (the absolute value of the difference in slope coefficients corresponding to the two feature paragraphs + 1); the slope coefficient corresponding to a single feature paragraph = (the monitoring data value corresponding to the last time point in the feature paragraph - the monitoring data value corresponding to the first time point in the feature paragraph) / the time interval between the first and last time points in the feature paragraph;

[0088] The value of the preset matching threshold can be determined by the user according to the actual application scenario. The larger the value of the preset matching threshold, the greater the user's demand for improving the accuracy of video frame acquisition. A preset matching threshold value is provided to detect the historical records of feature acquisition by the user based on the feature paragraph combination, and the average value of the reference matching thresholds corresponding to each feature paragraph combination in the historical records that can meet the user's needs is recorded as the preset matching threshold. The reference matching threshold corresponding to a single feature paragraph combination is the matching threshold corresponding to any two feature paragraphs in the feature paragraph combination;

[0089] In the uniform acquisition according to the cycle evaluation value, the target monitoring cycle is divided into w equal parts, and the video frames corresponding to the video data corresponding to the target monitoring cycle at each equal division point in the target monitoring cycle are used as the acquired video frames;

[0090] The value of w is positively correlated with the period evaluation value corresponding to the target monitoring period. Period evaluation value = data fluctuation coefficient / preset data fluctuation coefficient + end-side correlation / preset end-side correlation.

[0091] Specifically, the judgment condition for triggering the analysis module to respond is that when the initial deviation index is greater than or equal to the preset initial deviation index and the topological connection coefficient is greater than or equal to the preset topological connection coefficient, the compensation coefficient is determined based on the supplementary dimension.

[0092] Wherein, initial deviation index = |initial fault threshold - average value of comparison thresholds corresponding to each associated record|, and comparison threshold corresponding to a single associated record = (fault threshold corresponding to the associated record / data fluctuation coefficient corresponding to the associated record) × data fluctuation coefficient corresponding to the monitoring point currently being analyzed;

[0093] The values ​​of the characteristic monitoring parameters corresponding to each time point of the characteristic monitoring parameters corresponding to the target monitoring period of a single monitoring point are recorded as analysis data, the historical records that trigger analysis and can meet user needs are recorded as reference records, the analysis data corresponding to each reference record is recorded as comparison analysis data, and the historical records corresponding to the comparison analysis data whose correlation coefficient with the analysis data is greater than the preset correlation coefficient are recorded as association records;

[0094] The calculation formula of the correlation coefficient r corresponding to any two analysis data is:

[0095]

[0096] Where n is the number of time points in the monitoring period corresponding to a single analysis data; x i and y i are the values ​​of the characteristic monitoring parameters corresponding to the i-th time point in the monitoring period corresponding to the two analysis data, is xi The average value of the characteristic monitoring parameter corresponding to each time point in the monitoring period corresponding to the corresponding analysis data, y i The average value of the characteristic monitoring parameter corresponding to each time point in the monitoring period corresponding to the corresponding analysis data, i = 1, 2, 3, ..., n;

[0097] The value of the preset correlation coefficient can be determined by the user based on historical records. The greater the user's demand for improving the accuracy of fault warning, the larger the value of the preset correlation coefficient. A preset correlation coefficient value of 0.8 is provided;

[0098] For a single monitoring point, the topological connection coefficient is the total number of electrical devices directly connected to the electrical device corresponding to the monitoring point; two electrical devices directly connected means that the two electrical devices are connected by only one connecting line;

[0099] The values ​​of the preset initial deviation index and the preset topological connection coefficient can be determined by the user according to the actual application scenario. The smaller the values ​​of the preset initial deviation index and the preset topological connection coefficient, the greater the user's demand for determining the compensation coefficient based on the supplementary dimension. A value of the preset initial deviation index and the preset topological connection coefficient is provided, and the historical records of the user determining the compensation coefficient based on the supplementary dimension are detected. The average values ​​of the initial deviation index and the average value of the topological connection coefficient corresponding to the historical records that can meet the user's needs are recorded as the preset initial deviation index and the preset topological connection coefficient, respectively;

[0100] When determining the initial fault threshold based on the video frame evaluation coefficient and the feature dimension abnormality, the initial fault threshold = video frame evaluation coefficient / preset video frame evaluation coefficient + feature dimension abnormality / preset feature dimension abnormality;

[0101] The video frame evaluation coefficient is the standard deviation of the video frame comparison values ​​corresponding to each collected video frame. The video frame comparison value corresponding to a single collected video frame is the absolute value of the difference between the video frame reference value of the collected video frame and the video frame corresponding to the first time point in the monitoring period. The video frame reference value corresponding to a single video frame = the average value of the pixel values ​​corresponding to each pixel point in the video frame / the standard deviation of the pixel values ​​corresponding to each pixel point in the video frame.

[0102] The characteristic monitoring parameters corresponding to the target monitoring period are recorded as characteristic dimensions. For a single monitoring point, the monitoring point is recorded as the target monitoring point. The monitoring points whose sub-coupling coefficients with the target monitoring point are greater than the preset sub-coupling coefficients are recorded as associated points. The target monitoring point and each associated point are recorded as reference points. The characteristic dimension anomaly corresponding to the target monitoring point is the maximum value of the monitoring anomaly degrees corresponding to the characteristic dimensions of each reference point.

[0103] The value of the preset sub-coupling coefficient can be determined by the user based on historical records. The greater the user's demand for improved monitoring accuracy, the smaller the value of the preset sub-coupling coefficient. A value of the preset sub-coupling coefficient is provided, and the average value of the sub-coupling coefficients corresponding to each associated point in the historical records that can meet the user's needs is recorded as the preset sub-coupling coefficient;

[0104] The values ​​of the preset video frame evaluation coefficient and the preset feature dimension abnormality can be determined by the user according to the actual application scenario. The greater the user's demand for improving the warning sensitivity, the smaller the values ​​of the preset video frame evaluation coefficient and the preset feature dimension abnormality. A preset video frame evaluation coefficient and a preset feature dimension abnormality are provided. The historical records for triggering analysis are detected, and the average values ​​of the video frame evaluation coefficients and the average values ​​of the feature dimension abnormalities corresponding to the historical records that can meet the user's needs are recorded as the preset video frame evaluation coefficient and the preset feature dimension abnormality, respectively.

[0105] Determining a compensation coefficient based on the supplementary dimension includes: recording a to-be-supplemented dimension whose synchronization influence coefficient is greater than a preset synchronization influence coefficient as a supplementary dimension, and determining a compensation coefficient according to an outlier coefficient of the supplementary dimension;

[0106] Other monitoring parameters other than the characteristic dimension are recorded as dimensions to be supplemented. For a single dimension to be supplemented, the dimension to be supplemented is recorded as the target dimension to be supplemented. The synchronization influence coefficient corresponding to the characteristic dimension and the target dimension to be supplemented is the average value of the synchronization thresholds corresponding to each associated point. The synchronization threshold corresponding to a single associated point = 1 / (the absolute value of the difference between the monitoring anomaly degree corresponding to the characteristic dimension corresponding to the target point and the monitoring anomaly degree corresponding to the dimension to be supplemented corresponding to the associated point + 1);

[0107] The outlier coefficient of the supplementary dimension is the average of the sub-outlier coefficients corresponding to each supplementary dimension. The sub-outlier coefficient corresponding to a single supplementary dimension = the maximum value of the monitoring anomaly corresponding to the supplementary dimension at each reference point / the characteristic dimension anomaly;

[0108] Compensation coefficient = (supplementary dimension outlier coefficient - preset supplementary dimension outlier coefficient) / preset supplementary dimension outlier coefficient;

[0109] The values ​​of the preset synchronization influence coefficient and the preset supplementary dimension outlier coefficient can be determined by the user according to the actual application scenario. The greater the user's demand for improving monitoring accuracy, the smaller the value of the preset synchronization influence coefficient. The greater the user's demand for compensation for the initial fault threshold, the smaller the value of the preset supplementary dimension outlier coefficient. The average value of the synchronization influence coefficient corresponding to each compensation dimension and the average value of the supplementary dimension outlier coefficient in the historical records that can meet the user's needs will be recorded as the preset synchronization influence coefficient and the preset supplementary dimension outlier coefficient, respectively.

[0110] Fault threshold = (1 + compensation coefficient) × initial fault threshold.

[0111] Specifically, the judgment condition for triggering the analysis module to respond is that when the initial deviation index is less than a preset initial deviation index or the topological connection coefficient is less than a preset topological connection coefficient, the compensation coefficient is determined based on the supplementary video frame.

[0112] The method of determining the compensation coefficient based on the supplementary video frame includes: using the first and last time points of the target monitoring period and the time points corresponding to each acquired video frame as division points to obtain a number of divided sections, detecting the variation coefficient corresponding to each divided section, selecting the divided sections with a variation coefficient greater than a preset variation coefficient as supplementary sections, randomly selecting a video frame in each supplementary section as the supplementary video frame, and determining the compensation coefficient according to the impact deviation value corresponding to the supplementary video frame;

[0113] A single segment is a time segment between two adjacent segment points. The variance coefficient corresponding to a single segment = the absolute value of the difference between the video frame reference values ​​of the two video frames corresponding to the segment × the standard deviation of the value of the feature monitoring parameter corresponding to each time point in the segment;

[0114] Impact deviation value = |average value of video frame comparison values ​​corresponding to each supplementary video frame - average value of video frame comparison values ​​corresponding to each acquired video frame|;

[0115] Compensation coefficient = (impact deviation value - preset impact deviation value) / preset impact deviation value;

[0116] The values ​​of the preset variation coefficient and the preset impact deviation value can be determined by the user according to the actual application scenario. The greater the user's demand for improving the monitoring accuracy, the smaller the value of the preset variation coefficient. The average value of the variation coefficients corresponding to the supplementary paragraphs in the historical records that can meet the user's needs is recorded as the preset variation coefficient. The greater the user's demand for compensation for the initial fault threshold, the smaller the value of the preset impact deviation value. The average value of the impact deviation values ​​corresponding to the historical records that can meet the user's needs is recorded as the preset impact deviation value.

[0117] Fault threshold = (1 + compensation coefficient) × initial fault threshold;

[0118] It can be understood that by determining the compensation coefficient based on the supplementary video frame, it is possible to capture abnormal conditions such as changes in the appearance of the device, abnormal movements, smoke, sparks, etc. that are difficult to perceive by other means, thereby more accurately determining the fault threshold.

[0119] Specifically, the interval analysis module determines the length of the time interval based on the comparison correlation degree;

[0120] The length of the time interval is positively correlated with the comparison correlation degree.

[0121] The last time point in the target monitoring period is recorded as the first time point, and the time between any time point before the first time point and the first time point is recorded as the time interval;

[0122] The length of a time interval is the time interval from the earliest time point to the first time point in the time interval;

[0123] The historical records that undergo interval analysis and can meet user needs are recorded as first historical records, and the analysis data corresponding to each first historical record is recorded as reference analysis data. The comparison correlation degree = the number of historical records that send warnings to users in the historical records corresponding to the reference analysis data whose correlation coefficient with the analysis data is greater than the preset correlation coefficient / the number of historical records corresponding to the reference analysis data whose correlation coefficient with the analysis data is greater than the preset correlation coefficient.

[0124] Specifically, when the line correlation degree is less than the preset line correlation degree, the interval analysis module determines that the interval analysis method is the division comparison analysis;

[0125] In the division and comparison analysis, the interval division is performed according to the interval stability, and the fault threshold is determined according to the gradient variation value and the variation interaction coefficient.

[0126] The line correlation degree corresponding to a single monitoring point is the number of monitoring points whose sub-coupling coefficient with the monitoring point is greater than the preset sub-coupling coefficient;

[0127] The interval division is performed according to the interval stability, including: dividing the time interval into a equal parts to obtain a number of subintervals, where a subinterval is the time range between two adjacent equal division points, the value of a is positively correlated with the interval stability, the characteristic monitoring parameter corresponding to the target monitoring period is recorded as the reference parameter, and the interval stability is the standard deviation of the value of the reference parameter corresponding to each time point in the time interval;

[0128] The value of the preset line correlation degree can be determined by the user based on the actual application scenario. The greater the user's demand for segmentation and comparison analysis, the greater the value of the preset line correlation degree. A value of the preset line correlation degree is provided, and the historical records of the user's segmentation and comparison analysis are detected. The average value of the line correlation degrees corresponding to the historical records that can meet the user's needs is recorded as the preset line correlation degree;

[0129] Fault threshold = gradient variation value / preset gradient variation value + variation interaction coefficient / preset variation interaction coefficient. The values ​​of the preset gradient variation value and the preset variation interaction coefficient can be determined by the user based on historical records. The greater the user's demand for improved fault warning accuracy, the smaller the values ​​of the preset gradient variation value and the preset variation interaction coefficient. A method for determining the preset gradient variation value and the preset variation interaction coefficient is provided. The average value of the gradient variation value and the average value of the variation interaction coefficient corresponding to the historical records that can meet the user's needs are extracted and recorded as the preset gradient variation value and the preset variation interaction coefficient, respectively.

[0130] Gradient variation value = (total amount of increasing subintervals / total amount of subintervals) + (incremental mean / preset incremental mean);

[0131] For a single subinterval, if the incremental coefficient corresponding to the subinterval is greater than 0, the subinterval is an increasing subinterval, and the incremental coefficient corresponding to the single subinterval = the interval characteristic value corresponding to the subinterval - the interval characteristic value corresponding to the left adjacent subinterval. The left adjacent subinterval corresponding to the single subinterval is the subinterval located on the left side of the subinterval and adjacent to the subinterval; the characteristic monitoring parameter corresponding to the target monitoring period is recorded as the analysis parameter, and the interval characteristic value corresponding to the single subinterval is the average value of the value of the analysis parameter corresponding to each time point in the subinterval. It should be noted that the subinterval located at the leftmost side of the time interval does not have a left adjacent subinterval, so the incremental coefficient corresponding to the subinterval located at the leftmost side of the time interval is recorded as 0;

[0132] The incremental mean is the average value of the incremental coefficients corresponding to the incremental subintervals, and the preset incremental mean is the average value of the incremental means corresponding to the historical records that can meet user needs;

[0133] The subinterval whose interval characteristic value is greater than the preset interval characteristic value is recorded as a characteristic subinterval; the preset interval characteristic value is the average value of the interval characteristic values ​​corresponding to each characteristic subinterval in the historical records that can meet the user's needs;

[0134] The interaction coefficient of anomaly = (the minimum value of the time interval values ​​corresponding to the rightmost subinterval of the time series and each characteristic subinterval) / (characteristic interval fluctuation value + 1), where the time interval value corresponding to any two subintervals is the number of subintervals between the subintervals; the characteristic interval fluctuation value is the standard deviation of the interval reference value corresponding to each characteristic subregion, and the interval reference value corresponding to a single characteristic subregion is the minimum value of the time interval values ​​corresponding to the characteristic subregion and other subregions;

[0135] It can be understood that the increase and increment of data in different time periods of the gradient anomaly value and the anomaly interaction coefficient can effectively reflect the distribution pattern of the characteristic sub-interval in the time dimension, and thus effectively reflect the behavior pattern of historical data in the past. When determining the fault threshold, the normal range of historical data can be used as a reference, thereby effectively predicting the fault condition of the monitoring point.

[0136] Specifically, the interval analysis module determines that the interval analysis method is abnormal fusion analysis when the line correlation degree is greater than or equal to the preset line correlation degree;

[0137] In the abnormal fusion analysis, the fusion monitoring points are selected according to the sub-coupling coefficient, and the fault threshold is determined according to the benchmark abnormality.

[0138] The fusion monitoring point selection is performed according to the sub-coupling coefficient, including: for a single monitoring point, recording the monitoring point as a target monitoring point, and recording the monitoring points whose sub-coupling coefficients with the target monitoring point are greater than the preset sub-coupling coefficients and the target monitoring point as fusion monitoring points;

[0139] The benchmark anomaly degree is the maximum value of the anomaly evaluation values ​​corresponding to each fusion monitoring point;

[0140] Abnormal assessment value corresponding to a single fusion monitoring point = [(sub-coupling coefficient corresponding to the fusion monitoring point - preset sub-coupling coefficient) / preset sub-coupling coefficient] × [gradient abnormality value corresponding to the fusion monitoring point / preset gradient abnormality value + abnormality interaction coefficient corresponding to the fusion monitoring point / preset abnormality interaction coefficient];

[0141] The fault threshold is positively correlated with the baseline abnormality.

[0142] Thus far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art may make equivalent changes or substitutions to the relevant technical features, and the technical solutions after such changes or substitutions will fall within the scope of protection of the present invention.

[0143] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that the present invention is susceptible to various modifications and variations. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

Claims

1. A box-type substation based on the Internet of Things, characterized in that: include: The monitoring module is used to determine the monitoring mode according to the set conditions to obtain a number of monitoring points. The monitoring mode is uniform partition monitoring or associated partition monitoring; An analysis module, connected to the monitoring module, for determining the monitoring status of each monitoring point according to the trigger threshold and the coupling coefficient, and determining whether the analysis mode is trigger analysis or interval analysis according to the monitoring status; a trigger analysis module connected to the analysis module, configured to determine, in the trigger analysis, based on analysis conditions, whether the video frame acquisition mode is feature acquisition based on a combination of feature segments or uniform acquisition based on a periodic evaluation value, determine an initial fault threshold based on a video frame evaluation coefficient and a feature dimension abnormality, determine a compensation coefficient based on an initial deviation index and a topological connection coefficient, and determine a fault threshold based on the compensation coefficient and the initial fault threshold; An interval analysis module, connected to the analysis module, is used to determine the length of the time interval according to the comparison correlation degree in the interval analysis, and to determine the interval analysis method according to the line correlation degree to determine the fault threshold. The interval analysis method is a division comparison analysis or anomaly fusion analysis; An early warning module is connected to the trigger analysis module and the interval analysis module respectively, and is used to send an early warning to the user when the fault threshold is greater than a preset fault threshold.

2. The box-type substation based on the Internet of Things according to claim 1 is characterized in that: The monitoring module determines the monitoring mode according to the set conditions, including: The monitoring module responds to the setting condition that the architecture clutter is greater than or equal to the preset architecture clutter or the site instability is greater than or equal to the preset site instability, then the monitoring mode is determined to be associated partition monitoring; The setting condition for the monitoring module response is that when the architecture clutter is less than the preset architecture clutter and the site instability is less than the preset site instability, the monitoring mode is determined to be uniform partition monitoring.

3. The box-type substation based on the Internet of Things according to claim 2 is characterized in that: The analysis module determines that the analysis mode is trigger analysis when the monitoring state is that the trigger threshold is greater than or equal to the preset trigger threshold or the coupling coefficient is greater than or equal to the preset coupling coefficient.

4. The box-type substation based on the Internet of Things according to claim 3 is characterized in that: The analysis module determines that the analysis method is interval analysis when the monitoring state is that the trigger threshold is less than the preset trigger threshold and the coupling coefficient is less than the preset coupling coefficient.

5. The box-type substation based on the Internet of Things according to claim 3 is characterized in that: The trigger analysis module determines the video frame acquisition mode according to the analysis conditions, including: The analysis condition that triggers the analysis module to respond is that when the end-side correlation degree is less than a preset end-side correlation degree or the data fluctuation coefficient is greater than or equal to a preset data fluctuation coefficient, the video frame acquisition mode is determined to be feature acquisition based on the characteristic segment combination; The analysis condition triggering the analysis module to respond is that when the end-side correlation is greater than or equal to the preset end-side correlation and the data fluctuation coefficient is less than the preset data fluctuation coefficient, the video frame acquisition method is determined to be uniform acquisition based on the periodic evaluation value.

6. The box-type substation based on the Internet of Things according to claim 5 is characterized in that: The judgment condition for triggering the analysis module to respond is that when the initial deviation index is greater than or equal to the preset initial deviation index and the topological connection coefficient is greater than or equal to the preset topological connection coefficient, the compensation coefficient is determined based on the supplementary dimension.

7. The box-type substation based on the Internet of Things according to claim 6 is characterized in that: The judgment condition for triggering the analysis module to respond is that when the initial deviation index is less than a preset initial deviation index or the topological connection coefficient is less than a preset topological connection coefficient, the compensation coefficient is determined based on the supplementary video frame.

8. The box-type substation based on the Internet of Things according to claim 4 is characterized in that: The interval analysis module determines the length of the time interval according to the comparison correlation degree; The length of the time interval is positively correlated with the comparison correlation degree.

9. The box-type substation based on the Internet of Things according to claim 8, characterized in that: When the line correlation degree is less than the preset line correlation degree, the interval analysis module determines that the interval analysis method is the division comparison analysis; In the division and comparison analysis, the interval division is performed according to the interval stability, and the fault threshold is determined according to the gradient variation value and the variation interaction coefficient.

10. The box-type substation based on the Internet of Things according to claim 9, characterized in that: The interval analysis module determines that the interval analysis method is abnormal fusion analysis when the line correlation degree is greater than or equal to the preset line correlation degree; In the abnormal fusion analysis, the fusion monitoring points are selected according to the sub-coupling coefficient, and the fault threshold is determined according to the benchmark abnormality.

Images