A method and device for analyzing terminal data

Abstract

The present invention relates to the technical field of data analysis, and particularly relates to a method and device for analyzing monitoring terminal data. The method includes: establishing a database of glass tilt types and corresponding faults based on historical data; when it is determined that the glass is tilted in the image data and the dirt type on the glass surface is no dirt: comparing four groups of data with the corresponding thresholds set for each group of data through the data analysis module, and respectively outputting the first probabilities of four groups of faults based on the range of the exceeded thresholds; when it is determined that the glass is tilted in the image data and the dirt type on the glass surface is dirty; comparing two groups of data with the corresponding thresholds set for each group of data through the data analysis module, and respectively outputting the second probabilities of two groups of faults based on the range of the exceeded thresholds. The present invention can give relatively comprehensive fault prediction, and the staff can select subsequent fault troubleshooting plans according to the high and low probability values.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and in particular to an analysis method and device for monitoring terminal data. Background Art

[0002] A glass washing machine is used to wash glass efficiently and precisely. A glass washing machine usually consists of a transmission unit, a brushing unit, a water flushing unit, an electric control unit, etc. The relevant status data during the operation of these units is usually monitored through a monitoring terminal, and then the monitoring terminal analyzes the data to determine whether a relevant unit fails. The current judgment method usually sets a threshold, and then detects multiple data in real time. When a certain data exceeds the corresponding threshold, it is determined that this part fails. Currently, the monitored data usually only includes wind pressure, water temperature, water pressure, motor speed, etc. The monitored data is not comprehensive enough. When it is detected that the glass is tilted during the washing process, a relatively comprehensive fault prediction cannot be given. Summary of the Invention

[0003] In view of this, the purpose of the present invention is to provide an analysis method and device for monitoring terminal data to solve the problem that the currently monitored data is not comprehensive enough, and a relatively comprehensive fault prediction cannot be given when it is detected that the glass is tilted during the washing process.

[0004] Based on the above purpose, the present invention provides an analysis method for monitoring terminal data for analyzing the working data of a glass washing machine. The method is executed by a monitoring terminal device, and the monitoring terminal device includes: a first camera module, a second camera module, an image processing module, a water pressure detection module, a brush height detection module, a pressure roller height detection module, and a data analysis module. The method includes:

[0005] Based on historical data, establish a database of glass tilt types and corresponding faults;

[0006] Obtain the image data of the glass after washing through the first camera module;

[0007] Based on the image data, use the image processing module to judge the tilt type of the glass and the dirt type on the glass surface in the image data. When it is judged that the glass in the image data is tilted and the dirt type on the glass surface is no dirt:

[0008] Obtain the rotation data of the conveyor roller in the glass washing machine through the second camera module and the image processing module;

[0009] Obtain the first pressure data on both sides of the water pipe of the flushing unit in the glass washing machine through the water pressure inspection module;

[0010] Detect the first height data on both sides of the brush in the glass washing machine through the brush height detection module;

[0011] Detect the height data on both sides of the pressure roller in the glass washing machine through the pressure roller height detection module;

[0012] Compare the above four groups of data with the corresponding thresholds set for each group of data through the data analysis module, and respectively output the first probabilities of four groups of faults based on the range of the exceeded thresholds;

[0013] When it is judged that the glass is inclined in the image data and the type of dirt on the glass surface is dirty;

[0014] Obtain the second pressure data on both sides of the water pipe of the flushing unit in the glass washing machine through the water pressure inspection module;

[0015] Detect the second height data on both sides of the brush in the glass washing machine through the brush height detection module;

[0016] Compare the above two groups of data with the corresponding thresholds set for each group of data through the data analysis module, and respectively output the second probabilities of two groups of faults based on the range of the exceeded thresholds.

[0017] Optionally, when the first probabilities are the same, obtain the inclination angle of the glass in the image data, and respectively output the third probabilities of four groups of faults based on the number of times of four groups of faults corresponding to the glass at this inclination angle in the historical fault records.

[0018] Optionally, when the second probabilities are the same, obtain the inclination angle of the glass in the image data, and respectively output the fourth probabilities of two groups of faults based on the number of times of two groups of faults corresponding to the glass at this inclination angle in the historical fault records.

[0019] Optionally, when the third probabilities are the same, obtain the position information of the conveyor roller corresponding to the third probability, obtain the position information of the water pipe corresponding to the third probability, obtain the position information of the brush corresponding to the third probability, and obtain the position information of the pressure roller corresponding to the third probability;

[0020] Based on the above four groups of position information, obtain the adjacent relationship of the conveyor roller, water pipe, brush and pressure roller corresponding to the third probability. When at least two of the conveyor roller, water pipe, brush and pressure roller corresponding to the third probability are in an adjacent state, add the third probabilities corresponding to the adjacent states as the fifth probability for output.

[0021] Optionally, the second camera module includes multiple groups of cameras. The cameras are used to photograph one end of the conveyor roller, and one end of the conveyor roller is wrapped with a sticker. The sticker is evenly distributed with two different colors in a circular shape. The rotation state of the conveyor roller is judged by the change of colors in the pictures taken at a predetermined time interval.

[0022] Optionally, the water pressure detection module includes pressure sensors arranged on both sides of the top of the water pipe. When one of the nozzles at the bottom of the water pipe is blocked and the other nozzle is not blocked, at this time, the reaction force received by the water pipe on the side where the nozzle is not blocked is greater than the reaction force received by the water pipe on the side where the nozzle is blocked, so that the value of the pressure sensor on the water pipe on the side where the nozzle is not blocked is greater than the value of the pressure sensor on the water pipe on the side where the nozzle is blocked.

[0023] Optionally, both the brush height detection module and the pressure roller height detection module detect whether the heights at both ends of the brush and the pressure roller are the same through distance sensors, so as to judge whether the brush and the pressure roller are inclined.

[0024] Based on the same inventive concept, the present invention also provides a device for analyzing the data of a monitoring terminal. The method is executed by a monitoring terminal device, and the monitoring terminal device includes: a first camera module, a second camera module, an image processing module, a water pressure detection module, a brush height detection module, a pressure roller height detection module, and a data analysis module, wherein:

[0025] The first camera module is used to acquire the image data of the glass after cleaning is completed;

[0026] The image processing module is used to perform corresponding processing on the acquired image data and output the processed data;

[0027] The second camera module is used to acquire the image data of the conveyor roller in the glass washing machine;

[0028] The water pressure inspection module is used to acquire the first pressure data on both sides of the water pipe of the flushing unit in the glass washing machine;

[0029] The brush height detection module is used to detect the first height data on both sides of the brush in the glass washing machine;

[0030] The pressure roller height detection module is used to detect the height data on both sides of the pressure roller in the glass washing machine;

[0031] The data analysis module is used to analyze the acquired data and output the failure probability value of the corresponding unit according to the analysis result.

[0032] During operation, image data of the glass after cleaning is obtained through the first camera module; based on the image data, the inclination type of the glass and the dirt type on the glass surface are judged through the image processing module. When it is judged that the glass in the image data is inclined and the dirt type on the glass surface is no dirt, it indicates that the brush and the flushing unit can clean the glass. There are various possible reasons for the glass to be inclined. At this time, the rotation data of the conveyor rollers in the glass cleaning machine is obtained through the second camera module and the image processing module; the first pressure data on both sides of the water pipe of the flushing unit in the glass cleaning machine is obtained through the water pressure inspection module; the first height data on both sides of the brush in the glass cleaning machine is detected through the brush height detection module; the height data on both sides of the pressure roller in the glass cleaning machine is detected through the pressure roller height detection module; through the data analysis module, the above four groups of data are compared with the corresponding thresholds set for each group of data. Based on the range of the exceeded thresholds, the first probabilities of four groups of faults are respectively output. The larger the range of the exceeded threshold, the higher the probability of the fault occurring in the unit. At this time, the corresponding first probability value output for the unit is larger. The staff can select the subsequent fault troubleshooting plan according to the level of the first probability value.

[0033] When it is judged that the glass in the image data is inclined and the dirt type on the glass surface is dirty, it indicates that the brush and the flushing unit cannot clean the glass. It may be that the brush is inclined or the water pressure on one side of the water pipe in the flushing unit is large and the water pressure on the other side is small, resulting in the inclination of the glass. Therefore, the fault conditions of these two units can be preferentially checked. The second pressure data on both sides of the water pipe of the flushing unit in the glass cleaning machine is obtained through the water pressure inspection module; the second height data on both sides of the brush in the glass cleaning machine is detected through the brush height detection module; through the data analysis module, the above two groups of data are compared with the corresponding thresholds set for each group of data. Based on the range of the exceeded thresholds, the second probabilities of two groups of faults are respectively output. The staff can select the subsequent fault troubleshooting plan according to the level of the second probability value.

[0034] As can be seen from the above, when the present invention detects that the glass is inclined, it can detect data of multiple units causing the inclination, the data is more comprehensive, and then analyze the corresponding fault probabilities. The fault prediction is presented through the probabilities, which can give a relatively comprehensive fault prediction. The staff can select the subsequent fault troubleshooting plan according to the level of the probability value. At the same time, some of the data in this method do not need to be detected in real time, reducing the energy consumption and data processing volume. Description of the Drawings

[0035] To more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the accompanying drawings required for the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only those of the present invention. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can also be obtained based on these drawings.

[0036] Figure 1 The sectional view of the sticker of the embodiment of the present invention;

[0037] Figure 2 The structural schematic diagram of the water pipe of the embodiment of the present invention.

[0038] 1. Sticker; 2. Water pipe; 3. Pressure sensor. Specific embodiments

[0039] To make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the following further details the present invention in conjunction with specific embodiments.

[0040] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should have the ordinary meanings understood by those of ordinary skill in the art to which the present invention belongs. The "first", "second" and similar terms used in the present invention do not denote any order, quantity or importance, but are only used to distinguish different components. The terms such as "comprising" or "including" mean that the elements or items appearing before this term cover the elements or items listed after this term and their equivalents, without excluding other elements or items. The terms such as "connected" or "coupled" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The terms such as "upper", "lower", "left", "right" are only used to represent relative positional relationships. When the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0041] As Figure 1 shown, an analysis method for monitoring terminal data is used for the analysis of the working data of a glass washing machine. The method is executed by a monitoring terminal device, and the monitoring terminal device includes: a first camera module, a second camera module, an image processing module, a water pressure detection module, a brush height detection module, a pressure roller height detection module and a data analysis module. The method includes:

[0042] Based on historical data, establish a database of glass tilt types and corresponding faults;

[0043] Obtain the image data of the glass after cleaning through the first camera module;

[0044] Based on the image data, the image processing module determines the tilt type of the glass and the dirt type on the glass surface. When it is determined that the glass in the image data is tilted and the dirt type on the glass surface is no dirt:

[0045] The rotation data of the conveyor rollers in the glass washing machine is obtained through the second camera module and the image processing module;

[0046] The first pressure data on both sides of the water pipe of the flushing unit in the glass washing machine is obtained through the water pressure inspection module;

[0047] The first height data on both sides of the brush in the glass washing machine is detected through the brush height detection module;

[0048] The height data on both sides of the pressure roller in the glass washing machine is detected through the pressure roller height detection module;

[0049] The data analysis module compares the above four groups of data with the corresponding thresholds set for each group of data, and based on the range of the exceeded thresholds, respectively outputs the first probabilities of four groups of faults;

[0050] When it is determined that the glass in the image data is tilted and the dirt type on the glass surface is dirty;

[0051] The second pressure data on both sides of the water pipe of the flushing unit in the glass washing machine is obtained through the water pressure inspection module;

[0052] The second height data on both sides of the brush in the glass washing machine is detected through the brush height detection module;

[0053] The data analysis module compares the above two groups of data with the corresponding thresholds set for each group of data, and based on the range of the exceeded thresholds, respectively outputs the second probabilities of two groups of faults.

[0054] This method is mainly for fault prediction when the glass shows a tilted condition after cleaning. In this method, at the beginning, only the first camera module and the image processing module need to work, and other modules can be temporarily not working or working periodically. The database contains historical detection data, the corresponding fault conditions for the historical data, and the corresponding glass tilt types for the fault conditions. The types of glass tilt can be divided by the tilt angle. For example, when the brush is tilted in the historical data, what type of glass tilt is generally caused. During operation, the first camera module is used to obtain the image data of the glass after cleaning; based on the image data, the image processing module judges the tilt type of the glass and the dirt type on the glass surface in the image data. When it is judged that the glass in the image data is tilted and the dirt type on the glass surface is no dirt, it indicates that the brush and the flushing unit can clean the glass. There are various possible reasons for the glass to be tilted. At this time, the second camera module and the image processing module are used to obtain the rotation data of the conveyor rollers in the glass cleaning machine; the water pressure inspection module is used to obtain the first pressure data on both sides of the water pipe of the flushing unit in the glass cleaning machine; the brush height detection module is used to detect the first height data on both sides of the brush in the glass cleaning machine; the pressure roller height detection module is used to detect the height data on both sides of the pressure roller in the glass cleaning machine; the data analysis module compares the above four groups of data with the corresponding thresholds set for each group of data, and based on the range of the exceeded thresholds, respectively outputs the first probabilities of the four groups of faults. The larger the range of the exceeded threshold, the higher the probability of the fault occurring in this unit, and the larger the first probability value corresponding to this unit output at this time. The staff can select the subsequent fault troubleshooting plan according to the level of the first probability value.

[0055] When it is judged that the glass in the image data is tilted and the dirt type on the glass surface is dirty, it indicates that the brush and the flushing unit cannot clean the glass. It may be that the brush is tilted or the water pressure on one side of the water pipe in the flushing unit is large and the water pressure on the other side is small, resulting in the glass being tilted. Therefore, the fault conditions of these two units can be preferentially checked. The water pressure inspection module is used to obtain the second pressure data on both sides of the water pipe of the flushing unit in the glass cleaning machine; the brush height detection module is used to detect the second height data on both sides of the brush in the glass cleaning machine; the data analysis module compares the above two groups of data with the corresponding thresholds set for each group of data, and based on the range of the exceeded thresholds, respectively outputs the second probabilities of the two groups of faults. The staff can select the subsequent fault troubleshooting plan according to the level of the second probability value.

[0056] As can be seen from the above, when the present invention detects that the glass is tilted, it can perform data detection on multiple groups of units that cause the tilt, and the data is more comprehensive. Then, it analyzes the corresponding failure probabilities, and uses the probabilities to represent the failure prediction, which can give a relatively comprehensive failure prediction. The staff can select the subsequent failure troubleshooting plan according to the level of the probability value. At the same time, some data in this method do not need to be detected in real time, reducing the energy consumption and data processing volume.

[0057] In some embodiments, when the first probabilities are the same or differ very little, for example, less than 2%, obtain the tilt angle of the glass in the image data, and respectively output the third probabilities of four groups of failures corresponding to the four groups of failures when the glass has this tilt angle based on the number of times of the corresponding four groups of failures in the historical failure records.

[0058] When the first probabilities are the same or differ very little, for example, less than 2%, it means that the failures of several units may all cause the glass to tilt. At this time, in order to better provide guidance to the staff, respectively output the third probabilities of four groups of failures corresponding to the four groups of failures when the glass has this tilt angle based on the number of times of the corresponding four groups of failures in the historical failure records. That is, although the first probabilities are the same, at the same probability, the higher the number of times of the corresponding four groups of failures when this tilt angle finally appears, the greater the third probability output at this time. For example, when the tilt angle is 20 degrees, the number of times caused by the conveyor roller is the largest, and the probability of the conveyor roller failure output at this time is the largest. In this way, it can better provide guidance to the staff.

[0059] In some embodiments, when the second probabilities are the same, obtain the tilt angle of the glass in the image data, and respectively output the fourth probabilities of two groups of failures corresponding to the two groups of failures when the glass has this tilt angle based on the number of times of the corresponding two groups of failures in the historical failure records.

[0060] The principle of this design is the same as that of the previous embodiment. By calculating the probability again through the number of failures, it can better provide guidance to the staff.

[0061] In some embodiments, when the third probabilities are the same or differ very little, for example, less than 2%, it means that the statistics of historical times cannot provide guidance to the staff either. At this time, obtain the position information of the conveyor roller corresponding to the third probability, obtain the position information of the water pipe corresponding to the third probability, obtain the position information of the brush corresponding to the third probability, and obtain the position information of the pressure roller corresponding to the third probability;

[0062] Based on the above four groups of position information, obtain the adjacent relationships of the conveyor roller, water pipe, brush, and pressure roller corresponding to the third probability. When at least two of the conveyor roller, water pipe, brush, and pressure roller corresponding to the third probability are in an adjacent state, add the third probabilities corresponding to the adjacent states as the fifth probability output.

[0063] When the third probability also fails to provide guidance to the staff, in order to further provide guidance to the staff, obtain the above four groups of position information, and obtain the adjacent relationships of the conveyor rollers, water pipes, brushes, and pressure rollers corresponding to the third probability. When at least two of the conveyor rollers, water pipes, brushes, and pressure rollers corresponding to the third probability are in an adjacent state, add the third probabilities corresponding to the adjacent states as the fifth probability for output. When at least two of the faulty conveyor rollers, water pipes, brushes, and pressure rollers are in an adjacent state, it indicates that when the glass passes through the adjacent area, it will be affected by multiple faults. Therefore, the probability of this adjacent area being the fault point increases significantly. Therefore, adding the third probabilities corresponding to the adjacent states as the fifth probability for output can further provide guidance to the staff.

[0064] In some embodiments, the top-down projection of the glass washing machine is set in a natural coordinate system, and the coordinate positions of all conveyor rollers, water pipes, brushes, and pressure rollers in the coordinate system are obtained. When the third probability K1 is the same or differs very little, for example, less than 2%, at this time, obtain the coordinate position information A of the conveyor roller corresponding to the third probability K1, obtain the coordinate position information B of the water pipe corresponding to the third probability K1, obtain the coordinate position information C of the brush corresponding to the third probability K1, obtain the coordinate position information D of the pressure roller corresponding to the third probability K1, and obtain the dimension L of the glass being washed along the traveling direction when it is washed by the glass washing machine. The sixth probability K2 is calculated through the following formula:

[0065] |X - Y| < L, K2 = 2 * K1;

[0066] Wherein, X represents A or B or C or D, and Y represents A or B or C or D different from X.

[0067] In this embodiment, it is judged whether the distance between the faulty points is less than the dimension L of the glass traveling direction through the above inequality. When it is judged to be less than, it indicates that when the glass passes through these two faulty points, it will be affected by these two faults at the same time. Therefore, it is judged that the probability of the glass tilting here is greater than the probability of a single faulty point position. Therefore, the sixth probability K2 is output as twice K1. In this way, the distinction can further provide guidance on the faulty position to the staff.

[0068] Such as Figure 1As shown, in some embodiments, the second camera module includes multiple groups of cameras. The cameras are used to capture one end of the conveyor roller, and one end of the conveyor roller is wrapped with a sticker. The sticker is evenly distributed with two different colors in a circular shape. By observing the color changes in the pictures taken at a predetermined time interval, the rotation state of the conveyor roller is determined. The color change situation is completed by the image processing module. In this embodiment, when there is no color change in the pictures taken at a predetermined time interval, or the rotation state of the conveyor roller can be judged according to the degree of change. If there is basically no change, it means that this conveyor roller has a transmission failure and cannot rotate and convey the glass normally.

[0069] As Figure 2 As shown, in some embodiments, the water pressure detection module includes pressure sensors arranged on both sides of the top of the water pipe. When one side of the nozzle at the bottom of the water pipe is blocked and the other side is not blocked, at this time, the reaction force received by the water pipe on the unblocked side of the nozzle is greater than the reaction force received by the water pipe on the blocked side of the nozzle, so that the value of the pressure sensor on the water pipe on the unblocked side of the nozzle is greater than the value of the pressure sensor on the water pipe on the blocked side of the nozzle. By detecting the magnitude of the pressure difference between the two sides of the water pipe, the degree of blockage on one side of the water pipe is judged. The greater the pressure difference, the more serious the blockage on one side of the water pipe.

[0070] In some embodiments, both the brush height detection module and the pressure roller height detection module detect whether the heights at both ends of the brush and the pressure roller are consistent through distance sensors, so as to judge whether the brush and the pressure roller are tilted. By detecting the magnitude of the height difference between both ends of the brush and the pressure roller, the degree of tilt of the brush is judged. The greater the height difference, the greater the degree of tilt of the brush.

[0071] To further implement the present invention, the present invention also provides a device for analyzing the data of the monitoring terminal. The method is executed by the monitoring terminal device. The monitoring terminal device includes: a first camera module, a second camera module, an image processing module, a water pressure detection module, a brush height detection module, a pressure roller height detection module, and a data analysis module, wherein:

[0072] The first camera module is used to obtain the image data of the glass after cleaning;

[0073] The image processing module is used to perform corresponding processing on the obtained image data and output the processed data;

[0074] The second camera module is used to obtain the image data of the conveyor roller in the glass washing machine;

[0075] The water pressure inspection module is used to obtain the first pressure data on both sides of the water pipe of the flushing unit in the glass washing machine;

[0076] The brush height detection module is used to detect the first height data on both sides of the brush in the glass washing machine;

[0077] The pressure roller height detection module is used to detect the height data on both sides of the pressure roller in the glass washing machine;

[0078] The data analysis module is used to analyze the acquired data and output the failure probability value of the corresponding unit according to the analysis result.

[0079] When the device detects that the glass is tilted, it can detect the data of multiple groups of units that cause the tilt, and the data is more comprehensive. Then, it analyzes the corresponding failure probability, and uses the probability to represent the failure prediction, which can give a relatively comprehensive failure prediction. The staff can select the subsequent failure troubleshooting plan according to the high or low probability value. At the same time, some data in this method do not need to be detected in real time, which reduces the energy consumption and data processing volume.

[0080] Those of ordinary skill in the art should understand that the discussion of any embodiment above is only exemplary, and is not intended to imply that the scope of the present invention (including the claims) is limited to these examples; under the concept of the present invention, the technical features in the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations in different aspects of the present invention as above, which are not provided in detail for the sake of brevity.

[0081] The present invention aims to cover all such substitutions, modifications, and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for analyzing monitoring terminal data, used for analyzing working data of a glass washing machine, characterized in that: The method is performed by a monitoring terminal device, which includes: a first camera module, a second camera module, an image processing module, a water pressure detection module, a brush height detection module, a pressure roller height detection module and a data analysis module. The method includes: Based on historical data, a database of glass tilt types and corresponding faults is established; Acquiring image data of the glass after cleaning through the first camera module; Based on the image data, the image processing module determines the tilt type of the glass and the dirt type on the glass surface in the image data. When it is determined that the glass in the image data is tilted and the dirt type on the glass surface is not dirty: Acquiring the rotation data of the conveying roller in the glass washing machine through the second camera module and the image processing module; Acquire first pressure data on both sides of the water pipe of the flushing unit in the glass washing machine through the water pressure inspection module; Detecting first height data of both sides of the brush in the glass cleaning machine by the brush height detection module; Detecting the height data of both sides of the pressure roller in the glass washing machine by the pressure roller height detection module; The data analysis module compares the four groups of data with the corresponding thresholds set for each group of data, and outputs the first probabilities of the four groups of faults respectively based on the range of the exceeded thresholds; When it is determined that the glass is tilted in the image data and the dirt type on the glass surface is dirty; Acquire second pressure data on both sides of the water pipe of the flushing unit in the glass washing machine through the water pressure inspection module; Detecting second height data of both sides of the brush in the glass cleaning machine by the brush height detection module; The data analysis module compares the two groups of data with the corresponding thresholds set for each group of data, and outputs the second probabilities of the two groups of faults based on the range of the exceeded thresholds.

2. The method for analyzing monitoring terminal data according to claim 1, characterized in that: When the first probabilities are the same, the tilt angle of the glass in the image data is obtained, and the third probabilities of the four groups of faults are output respectively based on the number of times the glass appears at the tilt angle corresponding to the four groups of faults in the historical fault records.

3. The method for analyzing monitoring terminal data according to claim 1, characterized in that: When the second probabilities are the same, the tilt angle of the glass in the image data is obtained, and the fourth probabilities of the two groups of faults are outputted respectively based on the number of times the glass appears at the tilt angle corresponding to the two groups of faults in the historical fault records.

4. The method for analyzing monitoring terminal data according to claim 2, characterized in that: When the third probabilities are the same, obtain the position information of the conveying roller corresponding to the third probability, obtain the position information of the water pipe corresponding to the third probability, obtain the position information of the brush corresponding to the third probability, and obtain the position information of the pressing roller corresponding to the third probability; Based on the above four groups of position information, the adjacent relationship of the conveying rollers, water pipes, brushes and pressure rollers corresponding to the third probability is obtained. When at least two groups of conveying rollers, water pipes, brushes and pressure rollers corresponding to the third probability are in an adjacent state, the third probabilities corresponding to the adjacent states are added together as the fifth probability output.

5. The method for analyzing monitoring terminal data according to claim 1, characterized in that: The second camera module includes multiple groups of cameras, which are used to shoot one end of the conveyor roller. One end of the conveyor roller is wrapped with a sticker, and the sticker has two groups of different colors evenly distributed in a ring shape. The rotation state of the conveyor roller is judged by the color changes in the pictures taken at predetermined time intervals.

6. The method for analyzing monitoring terminal data according to claim 1, characterized in that: The water pressure detection module includes pressure sensors arranged on both sides of the top of the water pipe. When the nozzle on one side of the bottom of the water pipe is blocked and the nozzle on the other side is not blocked, the reaction force received by the water pipe on the side not blocked by the nozzle is greater than the reaction force received by the water pipe on the side blocked by the nozzle, so that the value of the pressure sensor on the water pipe on the side not blocked by the nozzle is greater than the value of the pressure sensor on the water pipe on the side blocked by the nozzle.

7. The method for analyzing monitoring terminal data according to claim 1, characterized in that: The brush height detection module and the pressure roller height detection module both detect whether the heights of the two ends of the brush and the pressure roller are consistent through distance sensors, thereby determining whether the brush and the pressure roller are tilted.

8. A device for executing the method for analyzing monitoring terminal data according to claim 1, characterized in that: The method is performed by a monitoring terminal device, which includes: a first camera module, a second camera module, an image processing module, a water pressure detection module, a brush height detection module, a pressure roller height detection module and a data analysis module, wherein: The first camera module is used to obtain image data of the glass after cleaning; The image processing module is used to process the acquired image data accordingly and output the processed data; The second camera module is used to obtain image data of a conveyor roller in a glass cleaning machine; The water pressure inspection module is used to obtain first pressure data on both sides of the water pipe of the flushing unit in the glass washing machine; The brush height detection module is used to detect first height data of both sides of the brush in the glass cleaning machine; The pressure roller height detection module is used to detect the height data of both sides of the pressure roller in the glass washing machine; The data analysis module is used to analyze the acquired data and output the failure probability value of the corresponding unit according to the analysis result.

Images