Barrier gate cross bar operation fault identification and repair system based on Internet of Things

Abstract

The invention relates to the technical field of barrier gate cross rod operation data processing, in particular to a barrier gate cross rod operation fault recognition and repair system based on the Internet of Things. The system comprises a data acquisition module used for acquiring a data sequence of operation data of each cross bar; the initial box number determination module is used for determining the initial box number of the data sequence based on the experience box number and the box number increase and decrease number; the normal data acquisition module is used for acquiring normal data in the data sequence; the division effect evaluation module is used for calculating a division effect score corresponding to the number of the initial boxes; the fault identification module is used for obtaining the optimization rate of the initial box number based on the spatial complexity and the division effect score, and then determining the optimal box number; and according to the optimal number of boxes and based on an HBOS algorithm, obtaining an abnormal score, and according to the abnormal score, reminding a worker to overhaul the cross bar. According to the invention, the accuracy of barrier gate cross bar operation fault detection can be improved, and workers are reminded to carry out maintenance.

Description

Technical Field

[0001] The present invention relates to the technical field of gate crossbar operation data processing, and in particular to a gate crossbar operation fault identification and repair system based on the Internet of Things. Background Art

[0002] As a key equipment at the entrance of a parking lot, the barrier gate bears the important responsibility of restricting the driving of motor vehicles and ensuring the safety of the parking lot. In actual applications, the barrier gate crossbar often has operational failures due to factors such as frequent lifting and lowering operations, bad weather and human damage, such as motor failure, sensor failure, mechanical component damage, etc., which seriously affects the normal operation of the parking lot and user experience. The traditional barrier gate fault identification and repair method mainly relies on manual inspection and regular maintenance. This method has many disadvantages such as slow response speed, low efficiency and high maintenance cost. In recent years, the Internet of Things technology has been widely used in the intelligent transformation of the barrier gate system. Through sensors, wireless communications, cloud computing and other technical means, the Internet of Things technology has realized remote monitoring, fault warning and intelligent maintenance of barrier gate equipment, greatly improving the reliability and maintenance efficiency of the barrier gate system.

[0003] By detecting anomalies in the crossbar operation data, fault information is sent to the cloud server in a timely manner when abnormal data is detected, so that the staff can repair the fault in time. Since the crossbar operation data is multi-dimensional feature data, the HBOS algorithm can be used. When the operation data of the gate crossbar is abnormal, the HBOS algorithm can quickly identify potential fault modes, such as unstable crossbar movement, abnormal speed or slow response, so as to help maintenance personnel intervene and repair in time. In the HBOS algorithm, the number of boxes is an important parameter. The number of boxes determines the width of each box in the histogram, which in turn affects the sensitivity and accuracy of anomaly detection. Narrower box widths can help the algorithm more accurately identify low-density anomalies, while wider boxes may cause outliers to mix with normal values ​​around them, thereby reducing the sensitivity of detection. Increasing the number of boxes can improve the accuracy of anomaly detection, but it will also increase the computational complexity of the algorithm. Therefore, it is necessary to find the optimal number of boxes to achieve better anomaly detection results and avoid excessive complexity. Summary of the invention

[0004] In order to solve the above technical problems, the purpose of the present invention is to provide a barrier crossbar operation fault identification and repair system based on the Internet of Things. The technical solutions adopted are as follows: An embodiment of the present invention provides a barrier crossbar operation fault identification and repair system based on the Internet of Things, the system comprising: The data acquisition module is used to collect the crossbar operation data, arrange each type of crossbar operation data according to the data size, and obtain the data sequence corresponding to each type of crossbar operation data; The module for determining the number of initial boxes is used to determine the number of empirical boxes in the data sequence; determine the number of boxes to increase or decrease based on the maximum value, minimum value and variance in the data sequence; and determine the initial number of boxes in the data sequence based on the number of empirical boxes and the number of boxes to increase or decrease. A normal data acquisition module is used to acquire data at a middle position among the data with the highest frequency in the data sequence, and the position of the data is the reference position; an abnormal value of each data is acquired according to the reference position and the data in the data sequence, and normal data is obtained by using the abnormal value; The partitioning effect evaluation module is used to partition the data sequence using the initial number of boxes, and calculate the partitioning effect score corresponding to the initial number of boxes according to the partitioning result; The fault identification module is used to calculate the spatial complexity according to the initial number of boxes; obtain the priority rate of the initial number of boxes according to the division effect score and spatial complexity corresponding to the initial number of boxes; obtain the optimal number of boxes for the data sequence based on the priority rate; obtain the abnormal score based on the optimal number of boxes and the HBOS algorithm, and remind the staff to inspect the crossbar according to the abnormal score.

[0005] Preferably, determining the number of empirical bins of the data sequence comprises: Input the number of data in a data series into the Sturges formula to obtain the number of empirical boxes for the data series.

[0006] Preferably, the calculation formula for the increase or decrease in the number of boxes is: , Among them, n represents the number of boxes increased or decreased in a data sequence; M represents the number of initial boxes increased or decreased; and They represent the maximum and minimum values ​​in the data sequence respectively; δ represents the skewness coefficient of the data sequence; N represents the number of data in the data sequence; It means that the window size is centered on the i-th data in the data sequence. The variance of the data, K represents the number of empirical boxes corresponding to the data sequence; S represents the variance of the data sequence; ⌊⌋ represents rounding down.

[0007] Preferably, determining the initial number of boxes of the data sequence based on the number of empirical boxes and the number of increases and decreases in the number of boxes includes: The upper and lower boundaries are determined according to the increase or decrease in the number of boxes and the number of empirical boxes, where the upper boundary is the sum of the increase or decrease in the number of boxes and the number of empirical boxes, and the lower boundary is the difference between the number of empirical boxes and the number of increase or decrease in the number of boxes; the initial number of boxes corresponding to the data sequence is the upper boundary, the lower boundary, and the integer between the lower boundary and the upper boundary.

[0008] Preferably, the calculation formula of the outlier value is: , in, Represents the outlier value of the jth data in the data sequence; Indicates the position of the jth data in the data sequence; Indicates the reference position in the data sequence; Represents the slope of the jth data in the data sequence and the previous data; Represents the slope between the jth data and the next data in the data sequence; Represents the data value of the jth data in the data sequence; represents the data value of the reference position data in the data sequence; c represents the adjustment coefficient.

[0009] Preferably, using abnormal values ​​to obtain normal data includes: An abnormality threshold is set. Data with an abnormal value greater than or equal to the abnormality threshold is used as reference data in a data sequence, and the rest of the data in the data sequence is used as normal data.

[0010] Preferably, the calculation formula for the partitioning effect score corresponding to the initial number of boxes is: , Among them, β represents the partitioning effect score corresponding to an initial number of boxes in the data sequence; It represents the remainder after dividing the number of normal data in the data sequence by the number of data in the box; k represents the initial number of boxes; Indicates the width of the u-th box; represents the sum of outliers of all data in the u-th box; c represents the adjustment coefficient.

[0011] Preferably, the space complexity is calculated according to the initial number of boxes, including: The difference between the preset value and the reciprocal of the initial number of boxes is the space complexity corresponding to the initial number of boxes.

[0012] Preferably, the optimization rate of the initial number of boxes is obtained according to the partition effect score and space complexity corresponding to the initial number of boxes, including: The preset value is obtained and the space complexity corresponding to the initial number of boxes is subtracted to obtain the first coefficient of space complexity; the partition effect score of the initial number of boxes is added to the first coefficient of space complexity and multiplied by the first adjustment coefficient to obtain the optimization rate of the initial number of boxes.

[0013] Preferably, the optimal number of boxes for the data sequence is obtained based on the priority rate, including: The initial number of boxes with the largest priority among the initial numbers of boxes corresponding to a data sequence is taken as the optimal number of boxes for the data sequence.

[0014] The embodiments of the present invention have at least the following beneficial effects: the present invention obtains various data when the barrier gate crossbar is running, and then obtains a data sequence of each crossbar operation data, and obtains the empirical number of boxes corresponding to each data sequence, and adjusts it on the basis of the empirical number of boxes, so that the subsequent optimization of the number of boxes is more accurate; further, the increase and decrease number of the number of boxes corresponding to each data sequence is obtained, and the number of boxes is adjusted within a certain range to obtain multiple initial numbers of boxes corresponding to each data sequence, and then the data sequence is divided by the initial number of boxes, and the division situation is analyzed to obtain the division effect score corresponding to each initial number of boxes, and then the space complexity of each initial number of boxes is calculated, and the division effect score and space complexity of the initial number of boxes are combined to obtain the optimization rate of each initial number of boxes, and then the optimal number of boxes for each data sequence is obtained, and the detection effect and complexity are considered to improve the accuracy of barrier gate fault detection while avoiding high calculation complexity, so that the staff can quickly and accurately receive fault information and then perform maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0016] Figure 1 A system block diagram of a barrier gate crossbar operation fault identification and repair system based on the Internet of Things provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0017] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation method, structure, features and effects of a barrier crossbar operation fault identification and repair system based on the Internet of Things proposed by the present invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures, or characteristics in one or more embodiments may be combined in any suitable form.

[0018] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0019] The following is a detailed description of a specific solution of a barrier crossbar operation fault identification and repair system based on the Internet of Things provided by the present invention in conjunction with the accompanying drawings.

[0020] Example: The main application scenarios of the present invention are: multiple sensors collect various data during the operation of the cross bar, mainly including the lifting speed, lowering speed, lifting response time, lowering reset time, cross bar displacement and other data. Through these data, the abnormal data during the operation of the cross bar can be detected based on the HBOS algorithm. When using the HBOS algorithm for monitoring, the number of boxes will affect the abnormal detection effect and the calculation complexity, resulting in extreme effects. Therefore, it is necessary to determine the optimal number of boxes corresponding to each type of data.

[0021] This application calculates the optimization rate of each number of boxes by analyzing the effect value and space complexity of dynamic box width division for different numbers of boxes, and finds the optimal number of boxes, so that the anomaly detection effect of the dynamic box width constructed with the optimal number of boxes is better, while avoiding excessive computational complexity.

[0022] See also Figure 1 , which shows a system block diagram of a barrier crossbar operation fault identification and repair system based on the Internet of Things provided by an embodiment of the present invention, the system includes the following modules: The data acquisition module is used to collect the crossbar operation data, arrange each crossbar operation data according to the data size, and obtain the data sequence corresponding to each crossbar operation data.

[0023] The main purpose of the present invention is to determine the optimal number of boxes according to the preferred ratio of different numbers of boxes, so that when the optimal number of boxes is used as the preset number of boxes for the dynamic box width of HBOS, the abnormality detection effect is better while avoiding excessive calculation complexity.

[0024] When the barrier crossbar is in operation, a variety of different crossbar operation data can be collected for abnormal analysis. Therefore, a variety of sensors are arranged on the crossbar to collect data when the crossbar is in operation. Preferably, in an embodiment of the present invention, data such as the bar lifting speed, bar lowering speed, bar lifting response time, bar lowering reset time, and crossbar displacement are collected. Each type of data is arranged according to the size of the data value, specifically, arranged in ascending order, thereby obtaining a data sequence for each type of data.

[0025] The module for determining the number of initial boxes is used to determine the empirical number of boxes in the data sequence; determine the increase or decrease in the number of boxes using the maximum value, minimum value and variance in the data sequence; and determine the initial number of boxes in the data sequence based on the empirical number of boxes and the increase or decrease in the number of boxes.

[0026] Due to external force majeure factors, such as changes in natural wind may create additional resistance to the crossbar, temperature changes may affect the material properties and lubrication conditions of the crossbar, thereby affecting the smoothness and accuracy of movement, dust and pollution may accumulate on the moving parts of the crossbar, and the current and magnetic field generated during operation may produce electromagnetic coupling with other electromagnetic devices, resulting in signal interference or degradation of motor performance.

[0027] Therefore, the data is in a fluctuating state every time the crossbar runs, and normal data may also be mutated. The same is true for abnormal data. For the crossbar, the difference between early abnormal data and normal data may not be obvious, so the data difference may be small, while normal data may be dense, abnormal data is relatively discrete, and the overall data value may have a skewed distribution. The effect of using different number of boxes is different. For data with a small difference between the maximum and minimum values, a smaller number of boxes is required. For data with a relatively concentrated density in a certain area, a larger number of boxes is required. For data with a skewed distribution, a larger number of boxes is required. Therefore, the number of boxes can be calculated according to the distribution and difference of the data values. First, the empirical box number K is calculated according to the Sturges formula, the initial box increase or decrease number is set, and the specific increase or decrease number of boxes is calculated according to the distribution and difference of the data, and the lower and upper boundaries of the box number range are obtained.

[0028] For a data sequence, the number of empirical boxes K is calculated according to the Sturges formula. The specific calculation formula is: , Where N represents the number of data in the data sequence. Since the data targeted by the Sturges formula is suitable for normal distribution or approximately normal distribution data, but the actual collected data may be skewed, the empirical box number cannot be used directly. This grouping may lead to uneven data and make the data processed too generalized. Therefore, it is necessary to use the empirical box number as a basis, obtain the approximate range of the box number according to the distribution characteristics of the data itself, and find the optimal box number within this range.

[0029] Furthermore, based on experience, the initial number of boxes is set to M, where the initial number of boxes can be adjusted according to actual conditions, but M needs to be greater than or equal to ; Then, the increase or decrease in the number of boxes corresponding to each data sequence is calculated according to the data distribution characteristics of each data sequence. Specifically, the maximum value, minimum value and variance in the data sequence are used to determine the increase or decrease in the number of boxes. The calculation formula is: , Among them, n represents the number of boxes increased or decreased in a data sequence; M represents the number of initial boxes increased or decreased; and They represent the maximum and minimum values ​​in the data sequence respectively; δ represents the skewness coefficient of the data sequence; N represents the number of data in the data sequence; It means that the window size is centered on the i-th data in the data sequence. The variance of the data, K represents the number of empirical boxes corresponding to the data sequence; S represents the variance of the data sequence; ⌊⌋ represents rounding down.

[0030] It indicates the difference between the maximum and minimum values ​​in the data sequence. The greater the difference between the maximum and minimum values, the greater the difference. δ indicates the skewness coefficient of the skewness of the data sequence distribution. The larger δ is, the greater the skewness of the data distribution. Indicates the skewness of the data series distribution; Indicates the concentration of the window composed of the i-th data, The smaller the value of is, the more concentrated the window composed of the i-th data is. The larger the value, the more highly concentrated areas there are.

[0031] Thus, we can get the number of boxes n corresponding to each data sequence, and determine the upper and lower boundaries according to the number of boxes n and the number of boxes from experience, where the upper boundary is the sum of the number of boxes and the number of boxes from experience, and the lower boundary is the difference between the number of boxes from experience and the number of boxes from increase or decrease; the initial number of boxes corresponding to the data sequence is the integer between the upper boundary, the lower boundary, and the lower boundary and the upper boundary. The lower boundary is Kn, the upper boundary is K+n, and the initial number of boxes for a data sequence is multiple, namely k=Kn, K-n+1, ​​K-n+2, ..., K+n, and each data sequence will have multiple corresponding initial numbers of boxes.

[0032] The normal data acquisition module is used to obtain the data in the middle position of the data with the highest frequency in the data sequence, and the position of this data is the reference position; the abnormal value of each data is obtained according to the reference position and the data in the data sequence, and the normal data is obtained using the abnormal value.

[0033] Furthermore, it is necessary to analyze the data in the data sequence to obtain the characteristics of the data in the data sequence, calculate the abnormal value of each data according to the characteristics of the abnormal data, obtain normal data and abnormal data (reference data), and perform dynamic box width division according to the number of boxes obtained above to obtain multiple dynamic box width histograms with different numbers of boxes. Select any k value for construction and analysis. The number of boxes will affect the number of data in each box and the starting and ending values ​​of each box, which in turn affects the width of the box and the depth of the box. The optimal k value should separate all normal data from abnormal data as much as possible, and only the width of abnormal data is larger, while the width of the rest of the normal data is smaller. Therefore, it is necessary to calculate whether the number of boxes divided by the selected k value can separate normal data from abnormal data and conform to the relationship between abnormal values ​​and width.

[0034] Furthermore, the data with the highest frequency in the data sequence is obtained, and the position of the data in the middle of these data is used as the reference position. Because most normal data are in the middle position and have a high repetition rate, the reference position is likely to be normal data. Then, the abnormal value of each data is obtained according to the reference position and the data in the data sequence. The calculation formula of the abnormal value is: , in, Represents the outlier value of the jth data in the data sequence; Indicates the position of the jth data in the data sequence; Indicates the reference position in the data sequence; Represents the slope of the jth data in the data sequence and the previous data; Represents the slope between the jth data and the next data in the data sequence; Represents the data value of the jth data in the data sequence; Represents the data value of the reference position data in the data sequence; c represents the adjustment coefficient to prevent the denominator from being 0.

[0035] It should be noted that the position here refers to the position of the data in the data sequence. For example, the first data is at position 1 and the second is at position 2. It represents the abnormal probability of the jth data at the position. If the data position is farther away from the position base point, the probability that the data is abnormal data increases; Indicates the probability of data mutation. If the probability of data mutation is greater, the probability that the data is abnormal data increases; It represents the numerical difference between the jth data and the data with the highest probability of being normal data. The larger the numerical difference, the greater the probability that the selected data is abnormal data.

[0036] Since interference often causes normal data to deviate, it is often located far away from the center in the data sequence. Data with a larger anomaly threshold has the above characteristics and is more likely to be abnormal data. Therefore, more attention is paid to such data. The anomaly threshold can be set to 0.7, and data with an anomaly value greater than or equal to 0.7 is recorded as the reference data (abnormal data) in the current calculated data sequence, and the rest of the data in the data sequence is recorded as normal data.

[0037] The partitioning effect evaluation module is used to partition the data sequence using the initial number of boxes, and calculate the partitioning effect score corresponding to the initial number of boxes based on the partitioning result.

[0038] Furthermore, an initial number of boxes in any data sequence is selected to divide the data sequence, and the division effect score corresponding to the initial number of boxes is calculated according to the division result. The position range and number of normal data in the data sequence are obtained in the normal data acquisition module. If the selected k value is optimal, the number of data in each box obtained according to the k value should be able to completely divide the normal data and the reference data, and the width of each box and the proportion of abnormal values ​​in the box are small. Therefore, the division effect score of the selected k value can be calculated based on the above analysis, and the specific calculation formula is: , Among them, β represents the partitioning effect score corresponding to an initial number of boxes in the data sequence; It represents the remainder after dividing the number of normal data in the data sequence by the number of data in the box; k represents the initial number of boxes; Indicates the width of the u-th box; represents the sum of outliers of all data in the u-th box; c represents the adjustment coefficient, which is not 0.

[0039] In the above formula, if the number of normal data in the data sequence is a multiple of the number of data in the box, that is, the remainder Y is 0, then the probability that the normal data and the reference data are completely separated is high; Indicates the proportion of the width of the u-th box in the sum of the widths of all boxes, The proportion of the sum of the outliers in the u-th box to the sum of the outliers in all boxes, It indicates the consistency of the relationship between the outlier value and the box width. The greater the consistency, the smaller the probability of error in the division of normal data and reference data.

[0040] Thus, the partitioning effect score for each initial number of boxes corresponding to the data sequence can be obtained.

[0041] The fault identification module is used to calculate the spatial complexity according to the initial number of boxes; obtain the priority rate of the initial number of boxes according to the division effect score and spatial complexity corresponding to the initial number of boxes; obtain the optimal number of boxes for the data sequence based on the priority rate; obtain the abnormal score based on the optimal number of boxes and the HBOS algorithm, and remind the staff to inspect the crossbar according to the abnormal score.

[0042] Since the number of boxes affects the space complexity, the more boxes there are, the higher the space complexity is. The calculation formula for the total space complexity is O(d*k), where d represents the data dimension, that is, the type of collected data, k represents the number of boxes, and d is a constant. Therefore, k is the largest factor affecting the space complexity. The space complexity is calculated from this. The difference between the preset value and the reciprocal of the initial number of boxes is the space complexity corresponding to the initial number of boxes. The preset value is 1. The specific calculation formula is: , Among them, σ represents the space complexity corresponding to the initial number of boxes; k represents the initial number of boxes.

[0043] Since the larger the initial number of boxes, the better the division effect, but the higher the space complexity, it is necessary to analyze these two factors to determine the optimal initial number of boxes, obtain the preset value and the space complexity corresponding to the initial number of boxes, and subtract them to get the first coefficient of space complexity; add the division effect score of the initial number of boxes to the first coefficient of space complexity and multiply them by the first adjustment coefficient to get the optimal ratio of the initial number of boxes. The specific calculation formula is: , Among them, μ represents the preference rate of the initial number of boxes; tj represents the first adjustment coefficient, which can be adjusted by the implementer according to the actual situation. In this application, the value is 0.5. In this way, the preference rate corresponding to each initial number of boxes of a data sequence can be calculated. When the preference rate is the highest, the division effect is high and the space complexity is low. It is the best balance point. Therefore, the k value corresponding to the highest preference rate at this time is the optimal number of boxes. When HBOS is used for the currently selected data, the preset number of boxes is the k value. The preset number of boxes for HBOS corresponding to the remaining data can be obtained in the same way. The initial number of boxes with the largest priority rate among the initial number of boxes corresponding to the data sequence is selected as the optimal number of boxes for the data sequence, thereby obtaining the optimal number of boxes for each data sequence.

[0044] Furthermore, the optimal number of boxes of each data sequence is used as the preset number of boxes of the HBOS algorithm for anomaly detection to obtain anomaly scores, wherein the HBOS algorithm is a prior art and will not be described in detail here. The larger the anomaly score, the more abnormal the data is, and a fault may occur. At this time, the anomaly score is sent to the staff to remind them to inspect the crossbar. Specifically, the inspection threshold is set according to the actual situation. When the anomaly score is greater than or equal to the inspection threshold, the staff is reminded to inspect the crossbar.

[0045] To summarize, the present application installs multiple sensors to collect various data during the operation of the crossbar, obtains the initial value and range of the number of boxes according to the distribution and characteristics of each data, performs dynamic box width division according to the number of boxes, calculates the abnormal value of each data according to the characteristics of the abnormal data, obtains normal data and reference data, and after dividing the data with the selected k, obtains the division effect value of each number of boxes according to the division situation. Since the number of boxes will affect the space complexity, the space complexity of each number of boxes is calculated, and the optimization rate of each number of boxes is obtained according to the division effect value and the space complexity. The optimization rate of each number of boxes is calculated within the set range, and the number of boxes with the highest optimization rate is used as the optimal number of boxes. Then, the HBOS algorithm is used for anomaly detection based on the optimal number of boxes, and then the staff is reminded to perform maintenance, which can reduce labor while taking into account the complexity of calculation and the accuracy of detection.

[0046] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The above is a description of a specific embodiment of this specification. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0047] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0048] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A barrier crossbar operation fault identification and repair system based on the Internet of Things, characterized in that: The system includes: The data acquisition module is used to collect the crossbar operation data, arrange each type of crossbar operation data according to the data size, and obtain the data sequence corresponding to each type of crossbar operation data; The module for determining the number of initial boxes is used to determine the number of empirical boxes in the data sequence; determine the number of boxes to increase or decrease based on the maximum value, minimum value and variance in the data sequence; and determine the initial number of boxes in the data sequence based on the number of empirical boxes and the number of boxes to increase or decrease. A normal data acquisition module is used to acquire data at a middle position among the data with the highest frequency in the data sequence, and the position of the data is the reference position; an abnormal value of each data is acquired according to the reference position and the data in the data sequence, and normal data is obtained by using the abnormal value; The partitioning effect evaluation module is used to partition the data sequence using the initial number of boxes, and calculate the partitioning effect score corresponding to the initial number of boxes according to the partitioning result; The fault identification module is used to calculate the spatial complexity according to the initial number of boxes; obtain the priority rate of the initial number of boxes according to the division effect score and spatial complexity corresponding to the initial number of boxes; obtain the optimal number of boxes for the data sequence based on the priority rate; obtain the abnormal score based on the optimal number of boxes and the HBOS algorithm, and remind the staff to inspect the crossbar according to the abnormal score.

2. According to the Internet of Things-based barrier crossbar operation fault identification and repair system of claim 1, it is characterized in that: The step of determining the number of empirical boxes of a data sequence comprises: Input the number of data in a data series into the Sturges formula to obtain the number of empirical boxes for the data series.

3. According to the Internet of Things-based barrier crossbar operation fault identification and repair system of claim 1, it is characterized in that: The calculation formula for the increase or decrease in the number of boxes is: ， Among them, n represents the number of boxes corresponding to a data sequence, M represents the number of initial boxes; and They represent the maximum and minimum values ​​in the data sequence respectively; δ represents the skewness coefficient of the data sequence; N represents the number of data in the data sequence; It means that the window size is centered on the i-th data in the data sequence. The variance of the data, K represents the number of empirical boxes corresponding to the data sequence; S represents the variance of the data sequence; ⌊⌋ represents rounding down.

4. According to the Internet of Things-based barrier crossbar operation fault identification and repair system of claim 1, it is characterized in that: The determining of the initial number of boxes of the data sequence based on the number of empirical boxes and the number of increases and decreases in the number of boxes includes: The upper and lower boundaries are determined according to the increase or decrease in the number of boxes and the number of empirical boxes, where the upper boundary is the sum of the increase or decrease in the number of boxes and the number of empirical boxes, and the lower boundary is the difference between the number of empirical boxes and the number of increase or decrease in the number of boxes; the initial number of boxes corresponding to the data sequence is the upper boundary, the lower boundary, and the integer between the lower boundary and the upper boundary.

5. According to the Internet of Things-based barrier crossbar operation fault identification and repair system of claim 1, it is characterized in that: The calculation formula of the outlier value is: ， in, Represents the outlier value of the jth data in the data sequence; Indicates the position of the jth data in the data sequence; Indicates the reference position in the data sequence; Represents the slope of the jth data in the data sequence and the previous data; Represents the slope between the jth data and the next data in the data sequence; Represents the data value of the jth data in the data sequence; represents the data value of the reference position data in the data sequence; c represents the adjustment coefficient.

6. According to the Internet of Things-based barrier crossbar operation fault identification and repair system of claim 1, it is characterized in that: The use of abnormal values ​​to obtain normal data includes: An abnormality threshold is set. Data with an abnormal value greater than or equal to the abnormality threshold is used as reference data in a data sequence, and the rest of the data in the data sequence is used as normal data.

7. According to the Internet of Things-based barrier crossbar operation fault identification and repair system of claim 1, it is characterized in that: The calculation formula for the partitioning effect score corresponding to the initial number of boxes is: ， Among them, β represents the partitioning effect score corresponding to an initial number of boxes in the data sequence; It represents the remainder after dividing the number of normal data in the data sequence by the number of data in the box; k represents the initial number of boxes; Indicates the width of the u-th box; represents the sum of outliers of all data in the u-th box; c represents the adjustment coefficient.

8. The barrier crossbar operation fault identification and repair system based on the Internet of Things according to claim 1 is characterized in that: The space complexity is calculated according to the initial number of boxes, including: The difference between the preset value and the reciprocal of the initial number of boxes is the space complexity corresponding to the initial number of boxes.

9. The barrier crossbar operation fault identification and repair system based on the Internet of Things according to claim 1 is characterized in that: The step of obtaining the optimal ratio of the initial number of boxes according to the partition effect score and space complexity corresponding to the initial number of boxes includes: The preset value is obtained and the space complexity corresponding to the initial number of boxes is subtracted to obtain the first coefficient of space complexity; the partition effect score of the initial number of boxes is added to the first coefficient of space complexity and multiplied by the first adjustment coefficient to obtain the optimization rate of the initial number of boxes.

10. The barrier crossbar operation fault identification and repair system based on the Internet of Things according to claim 1 is characterized in that: The method of obtaining the optimal number of boxes for the data sequence based on the priority rate includes: The initial number of boxes with the largest priority among the initial numbers of boxes corresponding to a data sequence is taken as the optimal number of boxes for the data sequence.

Images