Edge computing driven power collection loop real-time monitoring method

Abstract

The application discloses an edge computing driven current collecting ring real-time monitoring method and relates to the technical field of current collecting ring temperature monitoring. The application comprises the following steps: synchronously collecting the temperatures of all carbon brushes when the current collecting ring is normally operated to obtain carbon brush temperature sample data; performing sample division, calculating the difference degree of the carbon brush temperatures under different temperature levels, performing sample screening to obtain carbon brush temperature difference sample data; performing normal difference analysis to obtain the normal difference range of the carbon brush temperatures under different temperature levels and obtain normal difference reference data; and monitoring and analyzing the temperature of the carbon brush according to the normal difference reference data. The application is used to solve the problem that the existing current collecting ring temperature monitoring technology cannot set different temperature difference thresholds at different temperature levels according to the temperature difference of the carbon brush under normal operation of the current collecting ring and reliably monitor the temperature difference when monitoring the temperature difference between multiple carbon brushes.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of collector ring temperature monitoring, in particular to an edge computing driven collector ring real-time monitoring method. BACKGROUND

[0002] The collector ring temperature monitoring technology is a running state monitoring technology for key heat generating components of the collector ring system. The core is to detect the temperature parameters of the contact interface between the collector ring and the carbon brush, the carbon brush body, the surface of the collector ring and the insulating support and other key heat generating parts in real time or periodically through the whole process of sensing collection, data transmission and algorithm analysis, evaluate the temperature distribution characteristics and fluctuation rules, and thus guarantee the stable operation of the collector ring.

[0003] The existing collector ring temperature monitoring technology often synchronously collects the temperature of each carbon brush to calculate the corresponding temperature range, and then compares it with the fixed threshold set by humans. If the temperature exceeds the threshold, it is judged that the temperature of the carbon brush is abnormal. However, under normal working conditions, the temperature difference of the collector ring carbon brush will show a reasonable small increase with the increase of the overall temperature level. The heat generation of the carbon brush is jointly dominated by contact loss and friction loss. When the motor load increases, the overall temperature level of the carbon brush rises, and the current flowing through the carbon brush increases. Even if the contact resistance difference of each carbon brush remains unchanged, the difference in contact loss will increase significantly due to the amplification effect of the current, eventually resulting in a natural increase in the temperature difference between the carbon brushes. In a high temperature environment, the resistivity of the carbon brush will change nonlinearly with the increase of temperature, and the original small material uniformity difference will be amplified, further exacerbating the normal temperature difference fluctuation. The fixed threshold completely ignores the influence of the carbon brush temperature level on the temperature difference, and uses one threshold to judge the temperature difference at different temperature levels, which is easy to cause temperature monitoring errors and high misjudgment rate, and leads to low fault recognition, which cannot meet the precise needs of some scenes. Therefore, the existing collector ring temperature monitoring technology cannot set different temperature difference thresholds according to the temperature difference of the carbon brush under normal operation of the collector ring at different temperature levels, and cannot reliably monitor the temperature difference. SUMMARY

[0004] The present application aims to at least solve one of the technical problems in the prior art by synchronously collecting the temperatures of all carbon brushes when the collector ring is normally operating to obtain carbon brush temperature sample data; and performing sample division, calculating the difference degree of carbon brush temperatures at different temperature levels, and performing sample screening to obtain carbon brush temperature difference sample data; and performing normal difference analysis to obtain the normal difference range of carbon brush temperatures at different temperature levels to obtain normal difference reference data; and monitoring and analyzing the temperature of the carbon brush according to the normal difference reference data; to solve the problem that the existing collector ring temperature monitoring technology cannot set different temperature difference thresholds at different temperature levels according to the temperature difference of the carbon brush under normal operation of the collector ring and reliably monitor the temperature difference when monitoring the temperature difference between multiple carbon brushes.

[0005] To achieve the above-mentioned purpose, the present application provides an edge computing driven collector ring real-time monitoring method, comprising the following steps:

[0006] Synchronously collecting the temperatures of all carbon brushes when the collector ring is normally operating to obtain carbon brush temperature sample data;

[0007] Performing sample division on the carbon brush temperature sample data, calculating the difference degree of carbon brush temperatures at different temperature levels, and performing sample screening to obtain carbon brush temperature difference sample data;

[0008] Performing normal difference analysis according to the carbon brush temperature difference sample data to obtain the normal difference range of carbon brush temperatures at different temperature levels to obtain normal difference reference data;

[0009] Collecting the temperatures of all carbon brushes when the collector ring is normally operating and monitoring and analyzing the temperature of the carbon brush according to the normal difference reference data.

[0010] Further, synchronously collecting the temperatures of all carbon brushes when the collector ring is normally operating to obtain carbon brush temperature sample data comprises the following sub-steps:

[0011] Any collector ring with not less than n carbon brushes is denoted as a first collector ring, and the carbon brushes of the first collector ring are denoted as carbon brush 1 to carbon brush n in turn;

[0012] A position for measuring the temperature of the carbon brush is selected in the same position of carbon brush 1 to carbon brush n in turn, and is denoted as temperature measurement position 1 to temperature measurement position n in turn; the temperature range of the carbon brush when the first collector ring is normally operating is obtained and is denoted as carbon brush normal temperature range [AT, BT].

[0013] Further, synchronously collecting the temperatures of all carbon brushes when the collector ring is normally operating to obtain carbon brush temperature sample data further comprises the following sub-steps:

[0014] When the collector ring is in normal operation, the temperatures of the carbon brushes 1 to n are synchronously collected from the temperature measuring position 1 to the temperature measuring position n, recorded as carbon brush temperature information, and any one of the carbon brush temperature information is recorded as first temperature information;

[0015] The median of the first temperature information is obtained, recorded as a carbon brush temperature level corresponding to the first temperature information; a plurality of carbon brush temperature information is repeatedly collected, and the corresponding carbon brush temperature level is obtained to obtain carbon brush temperature sample data.

[0016] Further, the carbon brush temperature sample data is divided into samples, and the difference degree of the carbon brush temperature at different temperature levels is calculated, and sample screening is performed to obtain carbon brush temperature difference sample data, including the following sub-steps:

[0017] The carbon brush normal temperature range [AT, BT] is divided into a plurality of temperature intervals with a length of k1, recorded as temperature level intervals, and any one of the temperature level intervals is recorded as a first temperature interval, wherein k1 is the set interval length;

[0018] For the carbon brush temperature sample data, all carbon brush temperature information is divided according to the temperature level interval in which the corresponding carbon brush temperature level is located; the carbon brush temperature information in the first temperature interval is recorded as first sample information.

[0019] Further, the carbon brush temperature sample data is divided into samples, and the difference degree of the carbon brush temperature at different temperature levels is calculated, and sample screening is performed to obtain carbon brush temperature difference sample data, including the following sub-steps:

[0020] Any one of the carbon brush temperature information in the first sample information is recorded as second temperature information, the range and standard deviation of the second temperature information are calculated, the range and standard deviation of all carbon brush temperature information in the first sample information are repeatedly calculated, and the temperature range and standard deviation of the first temperature interval are obtained respectively;

[0021] The range and standard deviation in the temperature range and standard deviation set are one-to-one corresponding to the carbon brush temperature information they belong to, and are combined into a data pair, recorded as D={(AR1, AB1), (AR2, AB2), …, (ARm, ABm)}}, wherein m is the total number of data pairs; and any one data pair is recorded as (ARi, ABi), i∈[1, m];

[0022] The average value AR0 of the temperature range set and the average value AB0 of the temperature standard deviation set are calculated respectively, and (AR0, AB0) is recorded as the initial core point; the standard deviation RB0 of the temperature range set and the standard deviation BB0 of the temperature standard deviation set are calculated respectively.

[0023] Further, the carbon brush temperature sample data is divided into samples, and the difference degree of the carbon brush temperature at different temperature levels is calculated, and sample screening is performed to obtain carbon brush temperature difference sample data, which further includes the following sub-steps:

[0024] Calculate , denoted as (ARi, ABi) to the initial core point. The standard distance of all data pairs to the initial core point is repeatedly calculated, and the average value DP of all standard distances is calculated;

[0025] Data pairs with a standard distance greater than 2*DP to the initial core point are denoted as suspected data pairs, otherwise they are denoted as valid data pairs;

[0026] Based on all valid data pairs, the average value YR0 of the range and the average value YB0 of the standard deviation are calculated again; (YR0, YB0) is denoted as the effective core point; and the standard deviation RB1 of the range and the standard deviation BB1 of the standard deviation are calculated again;

[0027] According to RB1 and BB1, the standard distance of each valid data pair to the effective core point is calculated, and the average value DY and the standard deviation DB of all standard distances to the effective core point are calculated; DY+k2*DB is denoted as the normal distance threshold DT, wherein k2 is a set proportion coefficient.

[0028] Further, the carbon brush temperature sample data is divided into samples, and the difference degree of the carbon brush temperature at different temperature levels is calculated, and sample screening is performed to obtain carbon brush temperature difference sample data, which further includes the following sub-steps:

[0029] If (ARi, ABi) is a valid data pair, ARi / (ABi+e0) is calculated, denoted as the effective proportion of (ARi, ABi); the effective proportion of all valid data pairs is repeatedly calculated, and the average value RP and the standard deviation RB of the effective proportion are calculated, wherein e0 is a set minimum constant;

[0030] For any data pair (ARi, ABi), the standard distance ADi to the effective core point is calculated according to RB1 and BB1; if ADi is greater than DT, and ARi / (ABi+e0) is not located in [RP-k3*RB, RP+k3*RB], then ARi and ABi are determined as outliers, and are removed from the corresponding temperature range set and temperature standard deviation set, wherein k3 is a set proportion coefficient;

[0031] All outliers in the temperature range set and the temperature standard deviation set are repeatedly obtained and removed, and after completion, the standard range set and the standard standard deviation set of the first temperature interval are obtained, denoted as the difference sample data of the first temperature interval. Repeat to obtain the difference sample data of all temperature level intervals to obtain the carbon brush temperature difference sample data.

[0032] Further, the normal difference analysis is performed according to the carbon brush temperature difference sample data to obtain a normal difference range of the carbon brush temperature at different temperature levels, and normal difference reference data is obtained, including the following sub-steps:

[0033] According to the standard range set and the standard standard deviation set of the first temperature interval, all corresponding data pairs are obtained, denoted as standard data pairs. The range is taken as the x-axis, and the standard deviation is taken as the y-axis. A data scatter plot is plotted on a two-dimensional plane and denoted as data scatter plot.

[0034] The average value PV of the standard range set and the average value PU of the standard standard deviation set are calculated respectively. (PV, PU) is denoted as a standard core point. The slope of each standard data pair relative to the origin of the data scatter plot is calculated and denoted as an individual slope.

[0035] The average value KP and the standard deviation KB of all individual slopes are calculated, and [KP-e1*KB, KP+e1*KB] is denoted as a slope allowable range, wherein e1 is a set proportion coefficient.

[0036] Further, the normal difference analysis is performed according to the carbon brush temperature difference sample data to obtain a normal difference range of the carbon brush temperature at different temperature levels, and normal difference reference data is obtained, including the following sub-steps:

[0037] The Euclidean distance of each point in the data scatter plot to the standard core point is calculated, and all Euclidean distances are arranged from small to large, and the k4 percentile is taken and denoted as a Euclidean distance threshold, wherein k4 is a set percentile.

[0038] According to the slope of the point relative to the origin being within the slope allowable range and the Euclidean distance of the point to the standard core point being not greater than the Euclidean distance threshold, a sector region is demarcated in the data scatter plot and denoted as a normal difference region.

[0039] The maximum value MC of the range and the maximum value MB of the standard deviation of the standard data pairs located in the normal difference region are obtained, and MC and MB are denoted as the range threshold and the standard deviation threshold of the first temperature interval, respectively.

[0040] The range threshold and the standard deviation threshold of all temperature level intervals are repeatedly obtained to obtain the normal difference reference data.

[0041] Further, the temperature of all carbon brushes is collected when the collector ring is normally operated, and the temperature of the carbon brushes is monitored and analyzed according to the normal difference reference data, including the following sub-steps:

[0042] When the first collector ring is operated, the temperature of all carbon brushes is collected synchronously at the temperature measurement positions 1 to n at a first time interval, and the current collection is denoted as current temperature information, wherein the first time interval is t1.

[0043] Obtain the median of the current temperature information, denoted as the current temperature level, and calculate the range GC and the standard deviation GB of the current temperature information;

[0044] According to the temperature level interval where the current temperature level is located, the corresponding range threshold and the standard deviation threshold are obtained, if GC is greater than the corresponding range threshold or GB is greater than the corresponding standard deviation threshold, it is judged that the carbon brush temperature difference of the first collector ring is abnormal, otherwise it is judged that the carbon brush temperature difference of the first collector ring is normal; repeat the monitoring of the carbon brush temperature.

[0045] The beneficial effects of the present application: the present application synchronously collects the temperature of all carbon brushes when the collector ring is normally running, obtains carbon brush temperature sample data; sample division is performed on the carbon brush temperature sample data, the difference degree of carbon brush temperature under different temperature levels is calculated, sample screening is performed, and carbon brush temperature difference sample data is obtained; normal difference analysis is performed according to the carbon brush temperature difference sample data, the normal difference range of carbon brush temperature under different temperature levels is obtained, and normal difference reference data is obtained; the temperature of all carbon brushes is collected when the collector ring is normally running, and the temperature of the carbon brush is monitored and analyzed according to the normal difference reference data; when the temperature difference between multiple carbon brushes is monitored, different temperature difference thresholds can be set at different temperature levels according to the temperature difference of the carbon brush under normal operation of the collector ring, the reliability and stability of temperature monitoring are improved;

[0046] The present application calculates the initial core point and the standard deviation first, identifies the suspected data pairs based on the standard distance; and re-calculates the effective core point and obtains a new standard deviation for the effective data pairs, then obtains the normal distance threshold, and combines the ratio of the range and the standard deviation to double judge and eliminate the abnormality; it can accurately distinguish between normal data and error data, can maximize the retention of such normal values, improve the accuracy of eliminating abnormal samples, reduce the pollution of abnormal samples to threshold statistics, so as to obtain more reliable difference sample data; scatter points are drawn on the plane with the range as the y-axis and the standard deviation as the x-axis, and a fan-shaped normal area is delimited with the slope range and the Euclidean distance threshold to the core point; both the relationship between the two dimensions and the distance from the core point are limited, so that the obtained threshold is not a result of single-dimensional statistics, but a threshold after two-dimensional correlation constraint; both the relative shape and the absolute deviation are considered, the accuracy and reliability of temperature difference monitoring are improved. BRIEF DESCRIPTION OF DRAWINGS

[0047] Figure 1 The step flow chart of the method of the present application;

[0048] Figure 2 The range threshold and the standard deviation threshold acquisition flow chart of the present application;

[0049] Figure 3 The normal difference area schematic diagram of the present application;

[0050] Figure 4 Structure diagram of an electronic device of the present application. DETAILED DESCRIPTION

[0051] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0052] Embodiment 1, please refer to Figure 1 As shown in the figure, the present application provides an edge computing driven power collection ring real-time monitoring method, comprising the following steps:

[0053] Step S1, synchronously collecting the temperatures of all carbon brushes when the power collection ring is normally running to obtain carbon brush temperature sample data; step S1 comprises the following sub-steps:

[0054] Step S101, recording any one power collection ring with no less than n carbon brushes as a first power collection ring, and recording the carbon brushes of the first power collection ring as carbon brush 1 to carbon brush n in turn; n is generally 4, that is, the number of carbon brushes of the power collection ring is required to be no less than 4, if the number of carbon brushes of the power collection ring is too small, the corresponding standard deviation cannot be calculated, affecting subsequent analysis and processing; the power collection ring is a kind of electrical contact sliding connection device, through the sliding or rolling contact of the conductive part, electrostatic coupling or electromagnetic coupling, a reliable electrical path is established between the rotating structure and the stationary structure, realizing the continuous transmission of electric energy and electric signal;

[0055] Step S101, selecting a position for measuring the temperature of the carbon brush in the same position of carbon brush 1 to carbon brush n, recording them as temperature measurement position 1 to temperature measurement position n in turn; obtaining the temperature range of the carbon brush when the first power collection ring is normally running, recording it as carbon brush normal temperature range [AT, BT], that is, the temperature range that the carbon brush can reach during the running process of the power collection ring;

[0056] Step S101, synchronously collecting the temperatures of carbon brush 1 to carbon brush n from temperature measurement position 1 to temperature measurement position n when the power collection ring is normally running, recording it as carbon brush temperature information, and recording any one piece of carbon brush temperature information as first temperature information;

[0057] Step S104, obtaining the median of the first temperature information, recording it as the carbon brush temperature level corresponding to the first temperature information; repeatedly collecting multiple pieces of carbon brush temperature information and obtaining the corresponding carbon brush temperature level to obtain carbon brush temperature sample data; the median is more robust to single or a small number of abnormal points than the arithmetic mean, and is resistant to outliers, such as a carbon brush instantaneous failure with abnormally high temperature, so the median can more stably represent the overall temperature level of the carbon brush;

[0058] In the implementation process, the temperature measurement positions of the carbon brushes 1 to n are at the same geometric positions on each carbon brush, such as the middle points of the side surfaces of the carbon brushes or the outer edges of the contact areas, otherwise the temperature measurement positions are different, which introduces a deviation; for example, the temperature measurement points on different brushes are inconsistent, and the subsequent range and standard deviation obtained are mixed with the position difference instead of reflecting the temperature difference.

[0059] Step S2, sample division is performed on the carbon brush temperature sample data, the difference degree of the carbon brush temperature at different temperature levels is calculated, sample screening is performed, and carbon brush temperature difference sample data is obtained; step S2 includes the following substeps:

[0060] Step S201, the carbon brush normal temperature range [AT, BT] is divided into a plurality of temperature intervals with a length of k1, denoted as temperature level intervals, and any one temperature level interval is denoted as a first temperature interval, wherein k1 is the set interval length; in this embodiment, k1=5℃, that is, every 5℃ is a temperature level interval, for example, (25, 30], (30, 35], the unit is ℃, and k1 can be flexibly set, generally [3℃, 10℃]; the absolute difference between the carbon brushes is amplified with the overall temperature rise, so different temperature intervals should establish difference distribution respectively;

[0061] Step S202, for the carbon brush temperature sample data, all carbon brush temperature information is divided according to the temperature level interval in which the corresponding carbon brush temperature level is located; the carbon brush temperature information located in the first temperature interval is denoted as first sample information.

[0062] Step S203, any one carbon brush temperature information in the first sample information is denoted as second temperature information, the range and standard deviation of the second temperature information are calculated, the range and standard deviation of all carbon brush temperature information in the first sample information are repeatedly calculated, and the temperature range and standard deviation of the first temperature interval are obtained respectively; the range reflects the maximum temperature difference between the carbon brushes, and the standard deviation reflects the overall fluctuation of the temperature between the carbon brushes; the combination of the two can distinguish different fault types of single-point jump and overall fluctuation increase;

[0063] Step S204, the range and standard deviation in the temperature range and standard deviation set are one-to-one corresponding to the carbon brush temperature information to which they belong, and are combined into data pairs, denoted as D={ (AR1, AB1), (AR2, AB2), …, (ARm, ABm)}, wherein m is the total number of data pairs; and any one data pair is denoted as (ARi, ABi), i∈[1, m];

[0064] Under normal working conditions, the range and standard deviation of the carbon brush temperature are strongly positively correlated. The greater the range, the greater the temperature difference between the carbon brushes, and the greater the corresponding standard deviation. The proportion of the two is stable. The originally independent two one-dimensional indicators are converted into two-dimensional coordinate points, i.e., data pairs. Subsequently, abnormal values can be identified through double-dimensional constraints to avoid misjudgment in traditional single-index analysis.

[0065] In step S205, the average value AR0 of the temperature range set and the average value AB0 of the temperature standard deviation set are calculated, and (AR0, AB0) is recorded as the initial core point. The standard deviation RB0 of the temperature range set and the standard deviation BB0 of the temperature standard deviation set are calculated. The initial core point is the average position of all data pairs, representing the initial correlation center of the range and the standard deviation, and providing a reference for subsequent distance calculation.

[0066] In step S206, the standard distance of each data pair to the initial core point is calculated as (ARi, ABi). Dividing by the respective standard deviations is to eliminate the dimensional differences. The standard distances of all data pairs to the initial core point are repeatedly calculated, and the average value DP of all standard distances is calculated.

[0067] In step S207, the data pairs with a standard distance greater than 2×DP to the initial core point are recorded as suspected data pairs, and otherwise as valid data pairs. The standard distance of an extreme abnormal value will be much greater than the average distance of normal data. Using 2×DP as the threshold can quickly exclude such points and avoid their interference with the calculation of the stable core. 2 as the proportional coefficient can also be flexibly adjusted according to the actual application scenario.

[0068] In step S208, based on all valid data pairs, the average value YR0 of the range and the average value YB0 of the standard deviation are calculated again. (YR0, YB0) is recorded as the effective core point. The standard deviation RB1 of the range and the standard deviation BB1 of the standard deviation are calculated again. The effective core point is calculated based on the valid data pairs after removing the extreme abnormal values, and can represent the true average position of the range and the standard deviation under normal working conditions.

[0069] In step S209, the standard distance of each valid data pair to the effective core point is calculated according to RB1 and BB1, and the average value DY and the standard deviation DB of all standard distances to the effective core point are calculated. DY+k2×DB is recorded as the normal distance threshold DT, where k2 is the set proportional coefficient. In this embodiment, k2=2, which can be flexibly set according to the actual application scenario, generally [1.5, 2.5]. DT considers both typical deviations and change ranges, thereby distinguishing between slightly deviated and abnormal points.

[0070] Step S210, if (ARi, ABi) is a valid data pair, calculate ARi / (ABi+e0), denoted as the valid ratio of (ARi, ABi), repeat the calculation of the valid ratio of all valid data pairs, and calculate the mean RP and the standard deviation RB of the valid ratio, wherein e0 is a set minimum constant;

[0071] Step S211, for any data pair (ARi, ABi), calculate the standard distance ADi to the valid core point according to RB1 and BB1; if ADi is greater than DT, and ARi / (ABi+e0) is not located in [RP-k3×RB, RP+k3×RB], then determine that ARi and ABi are outliers, and remove them from the corresponding temperature range set and temperature standard deviation set, wherein k3 is a set proportion coefficient; in this embodiment, k3 is 1.5, which can be flexibly set, generally [1, 2]; the proportion of the range and the standard deviation is stable under normal working conditions, so the proportion will be unbalanced when there is a fault, and the double threshold value can improve the accuracy of the abnormality determination;

[0072] Step S212, repeat the acquisition of all outliers in the temperature range set and the temperature standard deviation set, and remove them, and after completion, the standard range set and the standard standard deviation set of the first temperature interval are obtained, denoted as the difference sample data of the first temperature interval, and the difference sample data of all temperature level intervals are repeatedly acquired to obtain the carbon brush temperature difference sample data.

[0073] For example, D={ (2.1, 0.7), (2.3, 0.8), (1.8, 0.6), (9.0, 2.5), (2.5, 0.9), (2.0, 0.7), (2.2, 0.8), (1.9, 0.6), (2.4, 0.9), (2.2, 0.8)}, then AR0=2.84, AB0=0.93, and the initial core point is (2.84, 0.93) ; RB0=2.13, BB0=0.54; the standard distances to the initial core point are {0.55, 0.48, 0.72, 4.10, 0.31, 0.62, 0.45, 0.68, 0.28, 0.45} respectively, then DP=0.864, 2×DP=1.728, and (9.0, 2.5) is a suspected data pair;

[0074] After removal, YR0=2.16, YB0=0.76, and the effective core point is (2.16, 0.76); RB1=0.25, BB1=0.13; then the standard distance to the effective core point is recalculated, and DY=0.32 and DB=0.15 are obtained; then DT=0.62;

[0075] The effective proportion of the calculated effective data pairs is {3.0, 2.88, 3.0, 2.78, 2.86, 2.75, 3.17, 2.67, 2.75} respectively, RP=2.88, RB=0.15; [RP-k3*RB, RP+k3*RB] is [2.66, 3.11]; for (9.0, 2.5), the standard distance to the effective core point is 30.5, 9.0 / 2.5=3.6, so it is an abnormal value;

[0076] In the specific implementation process, only the points far away from the center but still in the normal proportion interval may be rare but proportional samples, and only the points far away from the center and abnormal in proportion are truly excluded, thereby retaining the sample data to the maximum extent.

[0077] Step S3, performing normal difference analysis on the carbon brush temperature difference sample data to obtain a normal difference range of the carbon brush temperature at different temperature levels, and obtaining normal difference reference data; step S3 includes the following sub-steps:

[0078] Step S301, referring to Figure 2 As shown in the figure, according to the standard range set and the standard standard deviation set of the first temperature interval, all corresponding data pairs are obtained, denoted as standard data pairs. The range is taken as the x-axis, and the standard deviation is taken as the y-axis. A graph is drawn on a two-dimensional plane, denoted as a data scatter plot.

[0079] Step S302, calculating the average value PV of the standard range set and the average value PU of the standard standard deviation set respectively, and taking (PV, PU) as the standard core point. The slope of each standard data pair relative to the origin of the data scatter plot is calculated, denoted as the individual slope, that is, the standard deviation of each standard data is proportional to the range. The standard core point is the average position of all standard data pairs, representing the core correlation state of the range and the standard deviation under the normal working condition of the carbon brush.

[0080] Step S303, calculating the average value KP and the standard deviation KB of all individual slopes, and taking [KP-e1*KB, KP+e1*KB] as the slope allowable range, wherein e1 is a set proportion coefficient, in this embodiment, e1=1.5, which can be flexibly set according to the actual application scene, generally [1, 2]. Under the normal working condition, the range and the standard deviation of the carbon brush temperature are strongly positively correlated. The larger the temperature difference is, the higher the dispersion degree of the temperature distribution is, and the ratio of the two is relatively stable. KP is the central value of the normal proportion, and the standard deviation KB is the normal fluctuation range of the proportion. Through KP and KB, some omitted abnormal points can be excluded, thereby avoiding affecting the calculation of the subsequent threshold.

[0081] Step S304: Calculate the Euclidean distance from each point to the standard core point in the scatter plot of the data, arrange all Euclidean distances from smallest to largest, and take the k4 percentile as the Euclidean distance threshold, where k4 is the set percentile; in this embodiment, k4=95, generally 90 to 98, which can be flexibly set; the Euclidean distance to the standard core point is the straight-line distance from the point in the two-dimensional plane to the standard core point, which comprehensively reflects the degree to which the range and standard deviation of the point deviate from the normal mean. The larger the distance, the more the temperature difference characteristics of the sample deviate from the normal working conditions; taking the 95th percentile as the threshold can adapt to data with different distribution characteristics and avoid the bias of a fixed threshold;

[0082] For step S305, please refer to [link / reference]. Figure 3 As shown, based on the fact that the slope of a point relative to the origin is within the allowable slope range and the Euclidean distance from the point to the standard core point is not greater than the Euclidean distance threshold, a sector-shaped region is defined in the data scatter plot, denoted as the normal difference region; that is, the region formed by the two slope boundaries overlaps with the circular region with the standard core point as the circle and the Euclidean distance threshold as the radius; the normal difference region is the geometric manifestation of the proportional constraint and the distance constraint. Only points that simultaneously satisfy both constraints are normal data with reasonable proportions and small deviations.

[0083] Step S306: Obtain the maximum value of the range MC and the maximum value of the standard deviation MB of the standard data pair located in the normal difference region, and record MC and MB as the range threshold and standard deviation threshold of the first temperature range, respectively.

[0084] Step S307: Repeatedly obtain the range threshold and standard deviation threshold for all temperature level ranges to obtain normal difference reference data.

[0085] For example, in the first temperature range, the standard data pairs are {(1.8, 0.6), (2.0, 0.7), (2.1, 0.7), (1.9, 0.6), (2.3, 0.8), (2.2, 0.7), (1.7, 0.5), (2.4, 0.8), (2.0, 0.7), (2.1, 0.7), (1.9, 0.6), (2.2, 0.8)}, then PV=2.05, PU=0.68, and the standard nucleus... The center point is (2.05, 0.68); the slopes of the standard data pairs relative to the origin are {0.333, 0.350, 0.333, 0.316, 0.348, 0.318, 0.294, 0.333, 0.350, 0.333, 0.316, 0.364}; then KP=0.332, KB=0.018; then [KP-e1×KB, KP+e1×KB] is [0.305, 0.359];

[0086] Compute all the standard data pairs to the standard core point distance, and arrange in descending order, get {0.030, 0.051, 0.072, 0.080, 0.101, 0.102, 0.120, 0.150, 0.160, 0.180, 0.220, 0.262}, then the Euclidean distance threshold is 0.262, and the normal difference area is constructed;

[0087] Among them, the standard data pairs (1.7, 0.5) and (2.2, 0.8), the corresponding slope does not locate in [0.305, 0.359], so it is excluded from statistics, and MC=2.4, MB=0.8 is obtained from the remaining standard data pairs.

[0088] In the specific implementation process, the range and the standard deviation as the statistical quantity to describe the temperature dispersion degree, the physical meaning is impossible to be negative, so taking the upper limit of the range and the standard deviation in the normal difference area as the threshold value.

[0089] Step S4, collecting the temperature of all carbon brushes when the collector ring is normally running, and monitoring and analyzing the temperature of the carbon brush according to the normal difference reference data; Step S4 includes the following sub-steps:

[0090] Step S401, when the first collector ring is running, synchronously collecting the temperature of all carbon brushes at the temperature measurement position 1 to the temperature measurement position n with the first time interval, and recording the current collection as the current temperature information, wherein the first time interval is t1; In this embodiment, t1=1 second, that is, monitoring once per second, which can be flexibly set;

[0091] Step S402, obtaining the median of the current temperature information, recording as the current temperature level, and calculating the range GC and the standard deviation GB of the current temperature information;

[0092] Step S403, according to the temperature level interval where the current temperature level is located, obtaining the corresponding range threshold and the standard deviation threshold, if GC is greater than the corresponding range threshold or GB is greater than the corresponding standard deviation threshold, it is judged that the carbon brush temperature difference of the first collector ring is abnormal, that is, the temperature difference or the temperature dispersion degree between the carbon brushes exceeds the reasonable interval under the normal working condition; Otherwise, it is judged that the carbon brush temperature difference of the first collector ring is normal; Repeat the monitoring of the carbon brush temperature;

[0093] In the specific implementation process, the judgment of the carbon brush temperature difference only depends on the lightweight statistical operation such as the median, the range and the standard deviation, the calculation amount is small, the memory occupation is low, and it can be completed in real time and low delay on the acquisition module or the industrial edge gateway; The complex threshold value construction can be placed in the cloud, and the threshold value of each temperature level interval can be updated by periodic delivery to realize edge computing driving.

[0094] Embodiment 2, please refer to Figure 4 As shown in the figure, Figure 4An example is provided for a structural diagram of an electronic device, which can include a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory complete the communication among each other through the communication bus. The memory stores computer readable instructions, and the processor can call the instructions in the memory. When the computer readable instructions are executed by the processor, the steps in the edge computing driven carbon ring real-time monitoring method are run to realize the following functions: synchronously collecting the temperatures of all carbon brushes when the carbon ring is normally operated to obtain carbon brush temperature sample data; performing sample division on the carbon brush temperature sample data, calculating the difference degree of the carbon brush temperature at different temperature levels, and performing sample screening to obtain carbon brush temperature difference sample data; performing normal difference analysis according to the carbon brush temperature difference sample data to obtain the normal difference range of the carbon brush temperature at different temperature levels and obtain normal difference reference data; collecting the temperatures of all carbon brushes when the carbon ring is normally operated, and monitoring and analyzing the temperatures of the carbon brushes according to the normal difference reference data.

[0095] In addition, the logical instructions in the memory described above can be implemented in the form of a software functional unit and sold or used as an independent product, which can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application essentially or the part that contributes to the prior art or part of the technical solutions can be embodied in the form of a software product, which is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in the embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0096] In embodiment 3, the present application also provides a computer readable storage medium, and the present application provides a storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps in the edge computing driven carbon ring real-time monitoring method are run to realize the following functions: synchronously collecting the temperatures of all carbon brushes when the carbon ring is normally operated to obtain carbon brush temperature sample data; performing sample division on the carbon brush temperature sample data, calculating the difference degree of the carbon brush temperature at different temperature levels, and performing sample screening to obtain carbon brush temperature difference sample data; performing normal difference analysis according to the carbon brush temperature difference sample data to obtain the normal difference range of the carbon brush temperature at different temperature levels and obtain normal difference reference data; collecting the temperatures of all carbon brushes when the carbon ring is normally operated, and monitoring and analyzing the temperatures of the carbon brushes according to the normal difference reference data.

[0097] Through the description of the above embodiments, the embodiments of the present application can be provided as a method, a system or a computer program product. Based on such understanding, the above technical solutions can be embodied in the form of a software product, which can be stored in a computer readable storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, etc., and includes a plurality of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute the methods described in various embodiments or some parts of the embodiments.

[0098] In the embodiments provided by the present application, it should be understood that the disclosed system or method can be implemented in other manners. The embodiments described above are merely schematic, and should not be construed as limiting. For example, the division of the modules or the units is merely logical function division, and there can be other division manners in actual implementation. For example, a plurality of modules or units can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections between different modules can be indirect couplings or communication connections through some interfaces, and there can be electric, mechanical or other forms.

[0099] Finally, it should be noted that the above-mentioned embodiments are merely used to illustrate the technical solutions of the present application, rather than limit them; even if the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacements to some technical features; and these modifications or replacements do not cause the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. Edge computing driven power collection loop real-time monitoring method, characterized in that, Comprising the following steps: Synchronously collecting the temperature of all carbon brushes when the collector ring is in normal operation to obtain carbon brush temperature sample data; Dividing the carbon brush temperature sample data into samples, calculating the difference degree of the carbon brush temperature at different temperature levels, and performing sample screening to obtain carbon brush temperature difference sample data; Performing normal difference analysis according to the carbon brush temperature difference sample data to obtain the normal difference range of the carbon brush temperature at different temperature levels and obtain normal difference reference data; comprising the following sub-steps: according to the standard range set and the standard standard deviation set of the first temperature interval, obtaining all corresponding data pairs, denoted as standard data pairs, drawing a data scatter plot with the range as the x-axis and the standard deviation as the y-axis; calculating the average value PV of the standard range set and the average value PU of the standard standard deviation set, and denoted as the standard core point (PV, PU); and calculating the slope of each standard data pair relative to the origin of the data scatter plot, denoted as the individual slope; calculating the average value KP and the standard deviation KB of all individual slopes, and denoted as the slope allowable range [KP-e1*KB, KP+e1*KB], wherein e1 is a set proportion coefficient; calculating the Euclidean distance of each point in the data scatter plot to the standard core point, and arranging all Euclidean distances from small to large, and taking the k4 percentile, denoted as the Euclidean distance threshold, wherein k4 is a set percentile; according to the slope of the point relative to the origin being within the slope allowable range and the Euclidean distance of the point to the standard core point being not greater than the Euclidean distance threshold, a sector region is demarcated in the data scatter plot, denoted as the normal difference region; obtaining the maximum value MC of the range and the maximum value MB of the standard deviation of the standard data pairs located in the normal difference region, and denoted as the range threshold and the standard deviation threshold of the first temperature interval; repeating the above steps to obtain the range threshold and the standard deviation threshold of all temperature level intervals to obtain the normal difference reference data; Collecting the temperature of all carbon brushes when the collector ring is in normal operation, and monitoring and analyzing the temperature of the carbon brushes according to the normal difference reference data. 2.The edge computing driven current collector ring real-time monitoring method according to claim 1, wherein, Synchronously collecting the temperature of all carbon brushes when the collector ring is in normal operation to obtain carbon brush temperature sample data comprising the following sub-steps: Denote any collector ring with not less than n carbon brushes as a first collector ring, and denote the carbon brushes of the first collector ring as carbon brush 1 to carbon brush n in turn; Select a position for measuring the temperature of the carbon brush in the same position of carbon brush 1 to carbon brush n in turn, denoted as temperature measurement position 1 to temperature measurement position n; obtain the temperature range of the carbon brush when the first collector ring is in normal operation, denoted as the carbon brush normal temperature range [AT, BT].

3. The edge computing driven slip ring real-time monitoring method according to claim 2, characterized in that, Synchronously collecting the temperature of all carbon brushes when the collector ring is in normal operation to obtain carbon brush temperature sample data further comprises the following sub-steps: Synchronously collecting the temperature of carbon brush 1 to carbon brush n from temperature measurement position 1 to temperature measurement position n when the collector ring is in normal operation, denoted as carbon brush temperature information, and denoting any one piece of carbon brush temperature information as first temperature information; Obtain the median of the first temperature information, denoted as the carbon brush temperature level corresponding to the first temperature information; repeat the collection of multiple pieces of carbon brush temperature information and obtain the corresponding carbon brush temperature levels to obtain the carbon brush temperature sample data.

4. The edge computing driven slip ring real-time monitoring method according to claim 3, characterized in that, The carbon brush temperature sample data is divided into samples, the difference degree of the carbon brush temperature under different temperature levels is calculated, and sample screening is performed to obtain carbon brush temperature difference sample data, including the following sub-steps: Divide the carbon brush normal temperature range [AT, BT] into multiple temperature intervals with interval length k1, denoted as temperature level intervals, and any one temperature level interval is denoted as a first temperature interval, wherein k1 is the set interval length; For the carbon brush temperature sample data, all carbon brush temperature information is divided according to the temperature level interval where the corresponding carbon brush temperature level is located; the carbon brush temperature information located in the first temperature interval is denoted as first sample information.

5. The edge computing driven slip ring real-time monitoring method according to claim 4, characterized in that, The carbon brush temperature sample data is divided into samples, the difference degree of the carbon brush temperature under different temperature levels is calculated, and sample screening is performed to obtain carbon brush temperature difference sample data, including the following sub-steps: Any one carbon brush temperature information in the first sample information is denoted as second temperature information, the range and standard deviation of the second temperature information are calculated, and the range and standard deviation of all carbon brush temperature information in the first sample information are repeatedly calculated to obtain the temperature range and standard deviation of the first temperature interval. The range and standard deviation in the temperature range and standard deviation set are one-to-one corresponding to the carbon brush temperature information they belong to, and are combined into data pairs, denoted as D={ (AR1, AB1), (AR2, AB2), …, (ARm, ABm)}, wherein m is the total number of data pairs; and any one data pair is denoted as (ARi, ABi), i∈[1, m]; The average value AR0 of the temperature range set and the average value AB0 of the temperature standard deviation set are calculated respectively, and (AR0, AB0) is denoted as the initial core point. The standard deviation RB0 of the temperature range set and the standard deviation BB0 of the temperature standard deviation set are calculated respectively.

6. The edge computing driven slip ring real-time monitoring method according to claim 5, characterized in that, The carbon brush temperature sample data is divided into samples, the difference degree of the carbon brush temperature under different temperature levels is calculated, and sample screening is performed to obtain carbon brush temperature difference sample data, including the following sub-steps: Computing the standard distance of all data pairs to the initial core point and computing the average value DP of all standard distances; Data pairs with a standard distance greater than 2×DP from the initial core point are denoted as suspected data pairs, otherwise they are denoted as valid data pairs; Based on all valid data pairs, the average value YR0 of the range and the average value YB0 of the standard deviation are calculated again. (YR0, YB0) is denoted as the effective core point; and the standard deviation RB1 of the range and the standard deviation BB1 of the standard deviation are calculated again. According to RB1 and BB1, the standard distance of each valid data pair to the effective core point is calculated, and the average value DY and the standard deviation DB of all standard distances to the effective core point are calculated; DY+k2×DB is denoted as the normal distance threshold DT, wherein k2 is a set proportion coefficient.

7. The edge computing driven slip ring real-time monitoring method according to claim 6, characterized in that, The carbon brush temperature sample data is divided into samples, the difference degree of the carbon brush temperature under different temperature levels is calculated, and sample screening is performed to obtain carbon brush temperature difference sample data, including the following sub-steps: If (ARi, ABi) is a valid data pair, calculate ARi / (ABi+e0), denoted as the valid ratio of (ARi, ABi), repeat the calculation of the valid ratio of all valid data pairs, and calculate the mean value RP and the standard deviation RB of the valid ratio, wherein e0 is a set minimum constant; For any data pair (ARi, ABi), calculate the standard distance ADi to the valid core point according to RB1 and BB1; if ADi is greater than DT, and ARi / (ABi+e0) is not located in [RP-k3×RB, RP+k3×RB], then determine that ARi and ABi are outliers, and remove them from the corresponding temperature range set and temperature scale set, wherein k3 is a set proportion coefficient; Repeat the acquisition and removal of all outliers in the temperature range set and the temperature scale set, and after completion, the standard range set and the standard scale set of the first temperature interval are obtained, denoted as the difference sample data of the first temperature interval, repeat the acquisition of the difference sample data of all temperature level intervals, and obtain the carbon brush temperature difference sample data.

8. The edge computing driven slip ring real-time monitoring method according to claim 7, characterized in that, Collect the temperatures of all carbon brushes when the collector ring is normally running, and monitor and analyze the temperatures of the carbon brushes according to the normal difference reference data, including the following sub-steps: When the first collector ring is running, collect the temperatures of all carbon brushes at the temperature measurement positions 1 to n synchronously at a first time interval, and denote the current collection as current temperature information, wherein the first time interval is t1; Obtain the median of the current temperature information, denoted as the current temperature level, and calculate the range GC and the standard deviation GB of the current temperature information; According to the temperature level interval where the current temperature level is located, obtain the corresponding range threshold and scale threshold, if GC is greater than the corresponding range threshold or GB is greater than the corresponding scale threshold, then determine that the carbon brush temperature difference of the first collector ring is abnormal, otherwise, determine that the carbon brush temperature difference of the first collector ring is normal; repeat the monitoring of the carbon brush temperature.

Images