Online monitoring system for sandblasting equipment motion data based on artificial intelligence

Abstract

The present invention relates to the field of abnormality monitoring technology, and specifically to an online monitoring system for motion data of sandblasting equipment based on artificial intelligence. The system includes a processor and a memory, and the processor executes a computer program stored in the memory to implement the following steps: obtaining a target difference characterization value sequence corresponding to an initial difference characterization value sequence, and obtaining a difference degree characterization value corresponding to each target difference characterization value in the target difference characterization value sequence according to historical target difference characterization values ​​of each target difference characterization value in the target difference characterization value sequence; obtaining a set of abnormality degree characterization values ​​at the current moment according to the difference degree characterization value, and obtaining a result of monitoring the working state of the sandblasting equipment according to the abnormality degree characterization value in the abnormality degree characterization value set; and the present invention can improve the accuracy of monitoring the working state of the sandblasting equipment.

Description

Technical Field

[0001] The present invention relates to the technical field of abnormal monitoring, and in particular to an online monitoring system for motion data of sandblasting equipment based on artificial intelligence. Background Art

[0002] Sandblasting equipment is an automated equipment used for industrial cleaning, surface treatment and preparation before painting. The main function of sandblasting equipment is to remove pollutants, corrosion, burrs, etc. on the surface of the workpiece. It can also be used for lubrication and polishing to improve processing efficiency and product quality. The sandblasting equipment currently used is generally a sandblasting robot. However, when the sandblasting equipment is abnormal or fails during operation, it will directly affect the processing quality and efficiency. Therefore, it is very important to monitor the working status of the sandblasting equipment in real time during the operation of the sandblasting equipment.

[0003] In the prior art, the motion data of the sandblasting equipment during operation is generally monitored, and then whether the sandblasting equipment has an abnormality or failure is determined based on the monitoring results. If the motion data at a certain moment exceeds a preset threshold or the degree of change of the motion data at a certain moment is not within the specified change range, it is determined that the sandblasting equipment has an abnormality or failure, and the nozzle movement speed and sandblasting flow rate of the sandblasting equipment during operation are all motion data of the sandblasting equipment during operation; however, this existing detection method may have false detection when detecting whether the sandblasting equipment has an abnormality or failure. The main reason is that the processing complexity and surface shape of different areas of the processed workpiece are different, and during the processing, The sandblasting equipment will automatically adjust the parameters of the sandblasting equipment according to changes in other processing environments such as processing complexity and the surface shape of the workpiece being processed. The change in parameters will cause the motion data of the sandblasting equipment to change during operation. Therefore, when processing a workpiece area with high processing complexity or a surface area with a complex shape, the monitored motion data may exceed the preset threshold or the degree of change in the monitored motion data may not be within the specified range of change. This situation does not mean that the sandblasting equipment has an abnormality or failure, but according to the existing detection methods, it will be misjudged as an abnormality or failure of the sandblasting equipment. Therefore, in the process of real-time detection of the working status of the sandblasting equipment, improving the detection accuracy has become an urgent problem to be solved. Summary of the invention

[0004] In order to solve the above problems, the present invention provides an online monitoring system for motion data of sandblasting equipment based on artificial intelligence, and the technical solutions adopted are as follows:

[0005] An embodiment of the present invention provides an online monitoring system for motion data of sandblasting equipment based on artificial intelligence, comprising a processor and a memory, wherein the processor executes a computer program stored in the memory to implement the following steps:

[0006] Obtain a motion data sequence set corresponding to the sandblasting equipment in the current working time period, wherein the motion data sequence set consists of A different types of motion data sequences, where A is greater than 1;

[0007] Acquire a differential sequence of the motion data sequence, and obtain, according to the differential sequence, an initial difference characterization value sequence corresponding to the differential sequence and a first label value of each initial difference characterization value in the initial difference characterization value sequence;

[0008] According to each initial difference characterization value in the initial difference characterization value sequence and the first mark value of each initial difference characterization value, a target difference characterization value sequence corresponding to the initial difference characterization value sequence is obtained; according to the historical target difference characterization value of each target difference characterization value in the target difference characterization value sequence, a difference degree characterization value corresponding to each target difference characterization value in the target difference characterization value sequence is obtained;

[0009] According to the difference degree characterization value, a set of abnormality degree characterization values ​​at the current moment is obtained, and according to the abnormality degree characterization values ​​in the abnormality degree characterization value set, a result of monitoring the working state of the sandblasting equipment is obtained.

[0010] Beneficial effect: The present invention first obtains a set of motion data sequences corresponding to the sandblasting equipment in the current working time period; then, according to the differential sequence of the motion data sequence, obtains an initial difference characterization value sequence corresponding to the differential sequence and a first mark value of each initial difference characterization value in the initial difference characterization value sequence; according to each initial difference characterization value in the initial difference characterization value sequence and the first mark value of each initial difference characterization value, obtains a target difference characterization value sequence corresponding to the initial difference characterization value sequence; according to the historical target difference characterization value of each target difference characterization value in the target difference characterization value sequence, obtains a difference degree characterization value corresponding to each target difference characterization value in the target difference characterization value sequence; then, according to the difference degree characterization value, obtains an abnormality degree characterization value set at the current moment; finally, according to the abnormality degree characterization value in the abnormality degree characterization value set, obtains a result of monitoring the working state of the sandblasting equipment; and the present invention can monitor the working state of the sandblasting equipment more accurately based on the abnormality degree characterization value in the abnormality degree characterization value set at the current moment, that is, it can improve the accuracy of monitoring the working state of the sandblasting equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] 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.

[0012] Figure 1 The present invention is a flowchart of an online monitoring method for motion data of sandblasting equipment based on artificial intelligence. DETAILED DESCRIPTION

[0013] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field belong to the scope of protection of the embodiments of the present invention.

[0014] 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 pertains.

[0015] This embodiment provides an online monitoring system for motion data of sandblasting equipment based on artificial intelligence, including a processor and a memory, wherein the processor executes a computer program stored in the memory to implement an online monitoring method for motion data of sandblasting equipment based on artificial intelligence, such as Figure 1 As shown, the online monitoring method of the motion data of the sandblasting equipment based on artificial intelligence includes the following steps:

[0016] Step S001, obtaining a motion data sequence set corresponding to the sandblasting equipment in the current working time period.

[0017] Since the traditional detection method has low accuracy when performing real-time detection of the working status of the sandblasting equipment, the main purpose of this embodiment is to improve the detection accuracy during the real-time detection of the working status of the sandblasting equipment; since the type of sandblasting equipment to be detected in this embodiment is a sandblasting robot, the sandblasting robot that appears in the subsequent content of this embodiment is the sandblasting equipment, that is, the sandblasting robot belongs to the sandblasting equipment; in addition, for ease of understanding, this embodiment will be described later by taking the real-time monitoring process of any sandblasting robot during the working process as an example, that is, all sandblasting robots that appear later in this embodiment are the same sandblasting robot.

[0018] Since the motion data of the sandblasting robot during the working process can reflect the working state of the sandblasting robot, this embodiment will next collect and obtain multiple types of motion data of the sandblasting robot when it is working, and in this embodiment, the number of types of motion data to be collected at each collection moment is set to A, and A is greater than 1; in addition, since this embodiment needs to achieve the purpose of real-time detection, at each collection moment in the current working time period of the sandblasting robot, it is necessary to use sensors or collection equipment to collect and obtain A different types of motion data, and then the time series constructed by the same type of motion data obtained in the current working time period are all recorded as the motion data sequence corresponding to the sandblasting robot in the current working time period; then the set constructed by all the motion data sequences obtained in the current working time period is recorded as the motion data sequence set corresponding to the sandblasting device in the current working time period, and since the number of types of motion data collected at each collection moment is A, the number of motion data sequences in the motion data sequence set finally obtained is also A.

[0019] In specific applications, the implementer needs to set the time interval between the current working time period and adjacent collection moments according to actual conditions. For example, in this embodiment, the current working time period can be set to 10 minutes, and the time interval between adjacent collection moments can be set to 1 second. It is also required that the current moment must be included in the current working time period, that is, the last collection moment in the current working time period is the current moment.

[0020] In this embodiment, the implementer needs to determine the value of A and the specific type of motion data sequence according to the actual situation. If the value of A can be set to 3 in this embodiment, the motion data sequence set can be set to be composed of a moving speed sequence, a sandblasting flow sequence and a sandblasting nozzle angle sequence respectively; and the moving speed in the moving speed sequence refers to the moving speed of the nozzle of the sandblasting robot, which can be collected by a speed sensor; the sandblasting flow in the sandblasting flow sequence refers to the sandblasting flow of the nozzle of the sandblasting robot, which can be collected by a sandblasting flow sensor; the sandblasting nozzle angle in the sandblasting nozzle angle sequence refers to the sandblasting nozzle angle of the sandblasting robot, which can be collected by an angle sensor. And it should be noted that the sandblasting robot will automatically adjust various parameters of the sandblasting robot according to changes in other processing environments such as processing complexity and the surface shape of the workpiece, such as automatically adjusting parameters such as moving speed, sandblasting flow, nozzle angle, etc., and the moving speed, sandblasting flow, nozzle angle, etc. are all motion data of the sandblasting robot. For example, when the surface shape of the workpiece becomes complex, the sandblasting robot needs to automatically adjust the angle of the sandblasting nozzle and the sandblasting flow, etc. to achieve the purpose of removing pollutants, corrosion, burrs, etc. on the surface of the workpiece.

[0021] In addition, for ease of understanding, this embodiment will be described below by taking the acquisition process of any type of motion data sequence as an example. For example, this embodiment will be described below by taking the acquisition process of the sandblasting flow sequence corresponding to the current working time period of the sandblasting robot as an example. Then, the acquisition process of the sandblasting flow sequence corresponding to the current working time period of the sandblasting robot is as follows: at each acquisition moment in the current working time period of the sandblasting robot, the sandblasting flow of the sandblasting robot nozzle is collected by using relevant sensors or equipment, and the time series sequence composed of the sandblasting flow of all the sandblasting robot nozzles acquired in the current working time period is recorded as the sandblasting flow sequence corresponding to the sandblasting robot in the current working time period.

[0022] Therefore, the corresponding motion data sequence set can be obtained through the above process.

[0023] Step S002, obtaining a differential sequence of the motion data sequence, and obtaining an initial difference characterization value sequence corresponding to the differential sequence and a first label value of each initial difference characterization value in the initial difference characterization value sequence based on the differential sequence; obtaining a target difference characterization value sequence corresponding to the initial difference characterization value sequence based on each initial difference characterization value in the initial difference characterization value sequence and the first label value of each initial difference characterization value; obtaining a difference degree characterization value corresponding to each target difference characterization value in the target difference characterization value sequence based on a historical target difference characterization value of each target difference characterization value in the target difference characterization value sequence.

[0024] In this embodiment, the purpose of monitoring the working state of the sandblasting equipment is mainly achieved by determining the abnormality degree characterization value set at the current moment, that is, the real-time working state of the sandblasting robot is monitored mainly by determining the abnormality degree characterization value set at the current moment. Therefore, this embodiment needs to first obtain the abnormality degree characterization value set at the current moment. Before obtaining the abnormality degree characterization value set at the current moment, it is necessary to first obtain the initial difference characterization value sequence corresponding to the differential sequence and the first mark value of each initial difference characterization value in the initial difference characterization value sequence according to the differential sequence of the motion data sequence, and then obtain the target difference characterization value corresponding to the initial difference characterization value sequence according to each initial difference characterization value in the initial difference characterization value sequence and the first mark value of each initial difference characterization value. A value sequence is obtained, and then the difference degree representation value corresponding to each target difference representation value in the target difference representation value sequence is obtained according to the historical target difference representation value of each target difference representation value in the target difference representation value sequence. Finally, the abnormal degree representation value set at the current moment is obtained according to the difference degree representation value corresponding to each target difference representation value in the target difference representation value sequence. Therefore, based on the above description, it can be known that in this embodiment, the initial difference representation value sequence corresponding to the difference sequence of each motion data sequence in the motion data sequence set and the first label value of each initial difference representation value in the initial difference representation value sequence are obtained according to the difference sequence of the motion data sequence. That is, the specific acquisition process of the initial difference representation value sequence and the first label value of each initial difference representation value in the initial difference representation value sequence is:

[0025] First, a first-order difference operation is performed on each motion data sequence in the motion data sequence set, and the result of the operation is recorded as the difference sequence of each motion data sequence, that is, multiple difference sequences can be obtained through the above process, and the number of obtained difference sequences is consistent with the number of motion data sequences; in addition, since the process of performing a first-order difference operation on the sequence is a prior art, it will not be described in detail in this embodiment.

[0026] Then, a preset sliding window is constructed, and the preset sliding window is slid on the differential sequence of each motion data sequence with a preset sliding step size to obtain each window located on the differential sequence of each motion data sequence, and the first data in the first window on any differential sequence is the first data in the corresponding differential sequence; in addition, during the sliding process, when the last data in the differential sequence appears for the first time in the preset sliding window, the sliding needs to be stopped immediately.

[0027] In a specific application, the implementer needs to set the size of the preset sliding window and the preset sliding step length according to the actual situation, but the size of the preset sliding step length must not be greater than the length of the preset sliding window. For example, in this embodiment, the size of the preset sliding window can be set to or , you can set the preset sliding step size to 3 or 5.

[0028] Next, based on the data in the adjacent windows on the differential sequence of each motion data sequence, the initial difference characterization values ​​between the adjacent windows on the differential sequence of each motion data sequence are obtained; then, a sequence constructed by the initial difference characterization values ​​between all adjacent windows on each differential sequence is obtained, and recorded as the initial difference characterization value sequence corresponding to the corresponding differential sequence; for example, for any differential sequence, the initial difference characterization value sequence corresponding to the differential sequence is a sequence constructed by the initial difference characterization values ​​between all adjacent windows on the differential sequence, and the a-th initial difference characterization value in the initial difference characterization value sequence corresponding to the differential sequence is the initial difference characterization value between the a-th window and the a+1-th window on the differential sequence.

[0029] Then, each initial difference characterization value in the initial difference characterization value sequence is marked according to the position of each initial difference characterization value in the initial difference characterization value sequence, and the first marking value of each initial difference characterization value in each initial difference characterization value sequence is obtained, and the first marking value of the rth initial difference characterization value in each initial difference characterization value sequence is r; in addition, it should be noted that although all initial difference characterization values ​​with the same first marking value in different initial difference characterization value sequences come from different initial difference characterization value sequences, the time intervals corresponding to the windows for calculating these initial difference characterization values ​​with the same first marking values ​​are the same; for example, for the ath initial difference characterization value in the initial difference characterization value sequence B2 corresponding to the differential sequence B1 and the initial difference table corresponding to the differential sequence B3 For the ath initial difference characterization value in the eigenvalue sequence B4, the ath initial difference characterization value in the initial difference characterization value sequence B2 and the ath initial difference characterization value in the initial difference characterization value sequence B4 have the same first mark value, and the ath initial difference characterization value in the initial difference characterization value sequence B2 is the initial difference characterization value between the ath window and the a+1th window on the differential sequence B1, the ath initial difference characterization value in the initial difference characterization value sequence B4 is the initial difference characterization value between the ath window and the a+1th window on the differential sequence B3, and the acquisition time interval corresponding to the data in the ath window and the a+1th window on the differential sequence B1 is the same as the acquisition time interval corresponding to the data in the ath window and the a+1th window on the differential sequence B3.

[0030] And this embodiment will then describe in detail the specific process of obtaining the initial difference characterization value between adjacent windows on the differential sequence of each motion data sequence based on the data in adjacent windows on the differential sequence of each motion data sequence; in addition, for ease of understanding, this embodiment will describe the process of obtaining the initial difference characterization value between the b-th window and the b+1-th window on the differential sequence B of any motion data sequence, that is, the process of obtaining the initial difference characterization value between the b-th window and the b+1-th window on the differential sequence B is: first, the sequence formed by the data in the b-th window and the b+1-th window are recorded as the first sequence and the second sequence respectively; then, the coefficient of variation and the change trend factor of the first sequence and the coefficient of variation and the change trend factor of the second sequence are obtained, and the absolute value of the difference between the coefficient of variation of the first sequence and the coefficient of variation of the second sequence is calculated, and the absolute value of the difference between the coefficient of variation of the first sequence and the coefficient of variation of the second sequence is positively correlated and mapped. The obtained value is recorded as the coefficient of variation difference value, and the absolute value of the difference between the change trend factor of the first sequence and the change trend factor of the second sequence is calculated. The value obtained after positive correlation mapping of the absolute value of the difference between the change trend factor of the first sequence and the change trend factor of the second sequence is recorded as the change trend factor difference value; then the sum of the obtained coefficient of variation difference value and the change trend factor difference value is used as the initial difference characterization value between the b-th window and the b+1-th window; and the larger the initial difference characterization value, the greater the difference between the change and stability of the data in the b-th window and the change and stability of the data in the b+1-th window, that is, it indicates that the change trend and variation degree of the data in the b-th window are different from the change trend and variation degree of the data in the b+1-th window; it should be noted that this embodiment uses the exponential function exp() with a constant e as the base for positive correlation mapping, and the purpose of mapping here is to eliminate the influence of the dimension on the operation by division.

[0031] The reason for calculating the initial difference characterization value between adjacent windows is that if the sandblasting robot has an abnormality or failure during the working process, the difference between the data change trend and the degree of variation in the adjacent windows on the differential sequence of the motion data sequence will be large, so the initial difference characterization value can reflect the working status of the sandblasting robot.

[0032] In addition, the coefficient of variation of any sequence is the ratio of the standard deviation of the sequence to the mean value, and the larger the coefficient of variation of the sequence, the greater the degree of variation based on the mean, and the smaller the coefficient of variation, the smaller the degree of variation. The change trend factor of any sequence is the cumulative sum of all adjacent difference values ​​in the adjacent difference value sequence corresponding to the sequence, and the kth adjacent difference value in the adjacent difference value sequence corresponding to any sequence refers to the absolute value of the difference between the kth data and the k+1th data in the sequence.

[0033] Therefore, this embodiment can obtain the initial difference characterization value sequence corresponding to the differential sequence through the above process. Since it does not mean that the sandblasting robot has an abnormality or failure as long as the initial difference characterization value is large, in order to further ensure the accuracy of subsequent monitoring results, this embodiment needs to obtain the target difference characterization value sequence corresponding to the initial difference characterization value sequence according to each initial difference characterization value in the initial difference characterization value sequence and the first mark value of each initial difference characterization value. The target difference characterization value sequence is the key to subsequently obtaining the abnormality degree characterization value set, and the abnormality degree characterization value set is the key to accurately monitoring the working state of the sandblasting robot. Therefore, in this embodiment, the target difference characterization value sequence is obtained. The specific process of the target difference characterization value sequence corresponding to the initial difference characterization value sequence is as follows: first, according to each initial difference characterization value in the initial difference characterization value sequence and the first mark value of each initial difference characterization value, the target difference characterization value corresponding to each initial difference characterization value in the initial difference characterization value sequence is obtained; then, the sequence constructed by the target difference characterization values ​​corresponding to each initial difference characterization value in each initial difference characterization value sequence is recorded as the target difference characterization value sequence corresponding to the corresponding initial difference characterization value sequence, that is, the zth target difference characterization value in the target difference characterization value sequence corresponding to any initial difference characterization value sequence is the target difference characterization value corresponding to the zth initial difference characterization value in the initial difference characterization value sequence.

[0034] For ease of understanding, this embodiment will be described below by taking the specific process of obtaining the target difference characterization value corresponding to the f-th initial difference characterization value in any initial difference characterization value sequence F as an example, that is, the specific process of obtaining the target difference characterization value corresponding to the f-th initial difference characterization value in the initial difference characterization value sequence F is:

[0035] First, all remaining initial difference characterization value sequences except the initial difference characterization value sequence F are recorded as the sequences to be analyzed corresponding to the initial difference characterization value sequence F; then, in each sequence to be analyzed corresponding to the initial difference characterization value sequence F, all initial difference characterization values ​​that are the same as the first tag value of the f-th initial difference characterization value in the initial difference characterization value sequence F are obtained, and are all recorded as the characterization values ​​to be analyzed corresponding to the f-th initial difference characterization value in the initial difference characterization value sequence F; then, the set constructed from all the characterization values ​​to be analyzed corresponding to the f-th initial difference characterization value in the initial difference characterization value sequence F is recorded as the f-th initial difference characterization value. The characterization value set to be analyzed corresponds to the difference characterization value, and the cumulative sum of all the characterization values ​​to be analyzed in the characterization value set to be analyzed is recorded as the first cumulative characterization value; then the value after the f-th initial difference characterization value in the initial difference characterization value sequence F is subtracted from the first cumulative characterization value is obtained, and recorded as the first index value; then the first index value is positively correlated mapped using an exponential function, and the mapping result is recorded as the target difference characterization value corresponding to the f-th initial difference characterization value in the initial difference characterization value sequence F; in addition, the specific formula for obtaining the target difference characterization value corresponding to the f-th initial difference characterization value in the initial difference characterization value sequence F is:

[0036]

[0037] in, is the target difference representation value corresponding to the fth initial difference representation value in the initial difference representation value sequence F, exp() is an exponential function with a constant e as the base, is the fth initial difference characterization value in the initial difference characterization value sequence F, is the total number of data in the to-be-analyzed characterization value set corresponding to the f-th initial difference characterization value in the initial difference characterization value sequence F, It is the mth characterization value to be analyzed in the characterization value set to be analyzed corresponding to the fth initial difference characterization value in the initial difference characterization value sequence F; and the reason for using the exponential function for mapping is to prevent the target difference characterization value from being a negative number.

[0038] And because the larger the target difference representation value corresponding to the initial difference representation value between the two windows closest to the current moment on the differential sequence of a certain type of motion data sequence, and the smaller the target difference representation value corresponding to the initial difference representation value between the two windows closest to the current moment on the differential sequence of motion data sequences of other types except this type, then it indicates that for this type of motion data sequence, the possibility of abnormal changes in the motion data closest to the current moment is greater, which also indicates that the change in the motion data of this type at the current moment may not be caused by the automatic adjustment of the sandblasting robot, but due to the abnormality or failure of the sandblasting robot, that is, the greater the possibility of abnormality or failure of the sandblasting robot at the current moment; therefore, based on the above analysis, it can be seen that the target difference representation value corresponding to the initial difference representation value can reflect the working state of the sandblasting robot, and when the target difference representation value corresponding to the initial difference representation value is larger, it indicates that the possibility of abnormality or failure of the sandblasting robot is greater, and The bigger and The smaller the value, the larger the target difference representation value corresponding to the f-th initial difference representation value.

[0039] Therefore, this embodiment can obtain the target difference characterization value sequence corresponding to the initial difference characterization value sequence through the above process, and the target difference characterization values ​​in the target difference characterization value sequence can also reflect the working state of the sandblasting robot. However, the working state of the sandblasting robot cannot be monitored accurately based on the target difference characterization values. Therefore, after obtaining the target difference characterization value sequence, this embodiment will also obtain the difference degree characterization value corresponding to each target difference characterization value by analyzing the difference between each target difference characterization value in the target difference characterization value sequence and its historical target difference characterization value. That is, in order to reliably and accurately monitor the sandblasting robot (sandblasting equipment), this embodiment will then obtain the difference degree characterization value corresponding to each target difference characterization value in the target difference characterization value sequence according to the historical target difference characterization value of each target difference characterization value in the target difference characterization value sequence. That is, the specific acquisition process of the difference degree characterization value corresponding to the target difference characterization value is as follows:

[0040] First, the second label value corresponding to each target difference characterization value in the target difference characterization value sequence is obtained, and the second label value corresponding to the d-th target difference characterization value in any target difference characterization value sequence is d, and in this embodiment, the number of target difference characterization value sequences is consistent with the value of A.

[0041] Then, this embodiment describes the process of obtaining the difference degree representation value corresponding to any target difference representation value g in the target difference representation value sequence G as an example, that is, the process of obtaining the difference degree representation value corresponding to the target difference representation value g in the target difference representation value sequence G is: in the target difference representation value sequence G, all target difference representation values ​​before the target difference representation value g are obtained, and the sequence formed by all target difference representation values ​​before the target difference representation value g is recorded as the historical sequence corresponding to the target difference representation value g; then, the second eigenvalue sequence corresponding to the target difference representation value g is obtained, and the mean of the second eigenvalue sequence corresponding to the target difference representation value g is recorded as the difference degree representation value corresponding to the target difference representation value g.

[0042] And the process of obtaining the second eigenvalue sequence corresponding to the target difference characterization value g is as follows: for the h-th target difference characterization value in the historical sequence corresponding to the target difference characterization value g, the absolute value of the difference between the second mark value corresponding to the h-th target difference characterization value in the historical sequence and the second mark value corresponding to the target difference characterization value g is recorded as the first distance interval value corresponding to the h-th target difference characterization value, the inverse of the first distance interval value corresponding to the h-th target difference characterization value is used as the distance weight value corresponding to the h-th target difference characterization value, the absolute value of the difference between the h-th target difference characterization value and the target difference characterization value g is used as the first eigenvalue corresponding to the h-th target difference characterization value, and the product of the distance weight value corresponding to the h-th target difference characterization value and its corresponding first eigenvalue is used as the second eigenvalue corresponding to the h-th target difference characterization value; the sequence constructed by the second eigenvalues ​​corresponding to all target difference characterization values ​​in the historical sequence corresponding to the target difference characterization value g is recorded as the second eigenvalue sequence corresponding to the target difference characterization value g.

[0043] In addition, in this embodiment, the specific calculation expression of the difference degree representation value corresponding to the target difference representation value g is:

[0044]

[0045] in, is the difference degree representation value corresponding to the target difference representation value g in the target difference representation value sequence G, H is the total number of target difference representation values ​​in the historical sequence corresponding to the target difference representation value g, is the hth target difference representation value in the historical sequence corresponding to the target difference representation value g, is the target difference representation value g in the target difference representation value sequence G, is the first distance interval value corresponding to the hth target difference representation value in the historical sequence corresponding to the target difference representation value g, is the first eigenvalue corresponding to the hth target difference representation value in the historical sequence corresponding to the target difference representation value g, It is the distance weight value corresponding to the hth target difference representation value in the historical sequence corresponding to the target difference representation value g; and the closer the distance to the target difference representation value g is, the greater the reference significance is, that is, the greater the distance weight value is.

[0046] In addition, when The bigger and The smaller the time, the The larger the value of The larger the value is, the greater the difference between the target difference characterization value g and the historical target difference characterization value is. Analyzing the difference between the target difference characterization value g and the historical target difference characterization value is an important basis for subsequently determining the abnormality degree characterization value. The larger the difference degree characterization value is, the more likely that the sandblasting robot has a fault or abnormality. In addition, referring to the historical target difference characterization value of the target difference characterization value can avoid new changes in the working mode of the sandblasting robot, and can also monitor the working status of the sandblasting robot more accurately.

[0047] Therefore, this embodiment can obtain the difference degree representation value corresponding to each target difference representation value in each target difference representation value sequence through the above process.

[0048] Step S003, obtaining a set of abnormality degree characterization values ​​at the current moment according to the difference degree characterization values, and obtaining a result of monitoring the working state of the sandblasting equipment according to the abnormality degree characterization values ​​in the abnormality degree characterization value set.

[0049] After obtaining the difference degree representation value corresponding to each target difference representation value in each target difference representation value sequence, this embodiment will first obtain the abnormality degree representation value set at the current moment according to the difference degree representation value corresponding to each target difference representation value in the target difference representation value sequence. The abnormality degree representation value set is an important basis for subsequently obtaining accurate monitoring results. Therefore, the specific process of obtaining the abnormality degree representation value set at the current moment is:

[0050] First, according to the difference degree representation values ​​corresponding to each target difference representation value in each target difference representation value sequence, the abnormality degree representation value corresponding to each target difference representation value sequence is obtained; then the set constructed by the abnormality degree representation values ​​corresponding to all target difference representation value sequences is recorded as the abnormality degree representation value set; and the method for obtaining the abnormality degree representation value corresponding to the target difference representation value sequence is:

[0051] For any target difference characterization value sequence G: first, all difference degree characterization values ​​in the target difference characterization value sequence G are sorted in ascending order, and the sorted sequence is recorded as the first sequence corresponding to the target difference characterization value sequence G, and the median of the first sequence is used as the reference value of the target difference characterization value sequence G; then, the difference degree characterization value corresponding to the last target difference characterization value in the target difference characterization value sequence G and the difference degree characterization value corresponding to the second to last target difference characterization value are recorded as the first difference degree characterization value and the second difference degree characterization value, respectively; then, the ratio of the first difference degree characterization value to the second difference degree characterization value is obtained and recorded as the first ratio, and the ratio of the first difference degree characterization value to the reference value of the target difference characterization value sequence G is obtained and recorded as the second ratio; finally, the normalized value of the result obtained by multiplying the first ratio by the second ratio is recorded as the abnormality degree characterization value corresponding to the target difference characterization value sequence G; this embodiment uses the normalization function norm() to perform normalization processing.

[0052] And when the first ratio and the second ratio are larger, it indicates that within the time interval corresponding to the last target difference characterization value in the target difference characterization value sequence G, the possibility of abnormal changes or mutations in the motion data of the monitored sandblasting robot caused by a failure or abnormality of the sandblasting robot is greater, that is, the larger the abnormality degree characterization value, the more it indicates that the sandblasting robot has an abnormality or failure; in addition, the time interval corresponding to the target difference characterization value in the target difference characterization value sequence refers to, if any target difference characterization value sequence is a target difference characterization value sequence corresponding to the initial difference characterization value sequence corresponding to the differential sequence of the sandblasting flow sequence, then the vth target difference characterization value in the target difference characterization value sequence is the target difference characterization value corresponding to the initial difference characterization value between the vth window and the v+1th window on the differential sequence of the sandblasting flow sequence, then the data acquisition time interval in the vth window and the v+1th window is the time interval corresponding to the vth target difference characterization value.

[0053] The reason why the present embodiment mainly relies on the difference degree characterization value corresponding to the last target difference characterization value in the target difference characterization value sequence and the second to last target difference characterization value when calculating the abnormality degree characterization value is that, generally speaking, the above detection method can detect the failure or abnormality of the sandblasting robot in real time, and because the data collection time in the window corresponding to the difference degree characterization value corresponding to the last target difference characterization value in the target difference characterization value sequence and the second to last target difference characterization value is closest to the current moment, when obtaining the abnormality degree characterization value, it is sufficient to mainly refer to the change between the difference degree characterization value corresponding to the last target difference characterization value in the target difference characterization value sequence and the second to last target difference characterization value.

[0054] After obtaining the abnormality degree characterization value set at the current moment, and the number of abnormality degree characterization values ​​in the abnormality degree characterization value set is the same as the value of A, this embodiment will obtain the result of monitoring the working state of the sandblasting robot according to the abnormality degree characterization values ​​in the abnormality degree characterization value set. That is to say, according to the abnormality degree characterization values ​​in the abnormality degree characterization value set, the monitoring result of the working state of the sandblasting equipment can be obtained more accurately. The specific acquisition process is:

[0055] Determine whether there is an abnormality degree characterization value greater than the preset abnormality degree threshold in the abnormality degree characterization value set. If so, it is determined that the sandblasting robot at the current working moment has an abnormality or fault, that is, it is determined that the working state of the sandblasting machine at the current working moment is abnormal, and it is necessary to stop working at this time, and remind relevant staff to inspect the sandblasting robot, that is, the sandblasting equipment. Otherwise, it is determined that there is no abnormality in the sandblasting robot at the current moment, or it is determined that the working state of the sandblasting robot at the current moment is normal, then work can continue at this time.

[0056] In addition, in a specific application, the implementer needs to set the preset abnormality threshold according to the actual situation. For example, in this embodiment, the preset abnormality threshold can be set to 0.35.

[0057] So far, this embodiment has completed the monitoring of the working status of the sandblasting equipment.

[0058] In summary, the present embodiment first obtains a set of motion data sequences corresponding to the sandblasting equipment in the current working time period; then, according to the differential sequence of the motion data sequence, obtains an initial difference characterization value sequence corresponding to the differential sequence and a first mark value of each initial difference characterization value in the initial difference characterization value sequence; according to each initial difference characterization value in the initial difference characterization value sequence and the first mark value of each initial difference characterization value, obtains a target difference characterization value sequence corresponding to the initial difference characterization value sequence; according to the historical target difference characterization value of each target difference characterization value in the target difference characterization value sequence, obtains a difference degree characterization value corresponding to each target difference characterization value in the target difference characterization value sequence; then, according to the difference degree characterization value, obtains an abnormality degree characterization value set at the current moment; finally, according to the abnormality degree characterization value in the abnormality degree characterization value set, obtains a result of monitoring the working state of the sandblasting equipment; and the present embodiment can monitor the working state of the sandblasting equipment more accurately based on the abnormality degree characterization value in the abnormality degree characterization value set at the current moment, that is, can improve the accuracy of monitoring the working state of the sandblasting equipment.

[0059] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. An online monitoring system for motion data of sandblasting equipment based on artificial intelligence, comprising a processor and a memory, characterized in that: The processor executes the computer program stored in the memory to implement the following steps: Obtain a motion data sequence set corresponding to the sandblasting equipment in the current working time period, wherein the motion data sequence set consists of A different types of motion data sequences, where A is greater than 1; Acquire a differential sequence of the motion data sequence, and obtain, according to the differential sequence, an initial difference characterization value sequence corresponding to the differential sequence and a first label value of each initial difference characterization value in the initial difference characterization value sequence; Obtaining a target difference characterizing value sequence corresponding to the initial difference characterizing value sequence according to each initial difference characterizing value in the initial difference characterizing value sequence and a first label value of each initial difference characterizing value; Obtaining a difference degree representation value corresponding to each target difference representation value in the target difference representation value sequence according to a historical target difference representation value of each target difference representation value in the target difference representation value sequence; According to the difference degree characterization value, a set of abnormality degree characterization values ​​at the current moment is obtained, and according to the abnormality degree characterization values ​​in the set of abnormality degree characterization values, a result of monitoring the working state of the sandblasting equipment is obtained; The method for obtaining the initial difference characterization value sequence corresponding to the difference sequence includes: Sliding the difference sequence using a preset sliding window to obtain each window on the difference sequence; obtaining initial difference representation values ​​between adjacent windows on the difference sequence according to data in adjacent windows on the difference sequence; For any differential sequence, a sequence constructed by initial difference characterization values ​​between all adjacent windows on the differential sequence is recorded as an initial difference characterization value sequence corresponding to the differential sequence, and the a-th initial difference characterization value in the initial difference characterization value sequence corresponding to the differential sequence is the initial difference characterization value between the a-th window and the a+1-th window on the corresponding differential sequence; The method for obtaining the initial difference representation value between adjacent windows on the difference sequence includes: For the b-th window and the b+1-th window on the differential sequence of any motion data sequence: the sequences formed by the data in the b-th window and the b+1-th window are respectively recorded as the first sequence and the second sequence; the value obtained after positive correlation mapping of the absolute value of the difference between the coefficient of variation of the first sequence and the coefficient of variation of the second sequence is recorded as the coefficient of variation difference value; the value obtained after positive correlation mapping of the absolute value of the difference between the change trend factor of the first sequence and the change trend factor of the second sequence is recorded as the change trend factor difference value; the sum of the coefficient of variation difference value and the change trend factor difference value is used as the initial difference representation value between the b-th window and the b+1-th window, the change trend factor of the sequence is the cumulative sum of all adjacent difference values ​​in the adjacent difference value sequence corresponding to the corresponding sequence, and the k-th adjacent difference value in the adjacent difference value sequence corresponding to the sequence is the absolute value of the difference between the k-th data and the k+1-th data in the sequence; The first label value of the rth initial difference characterization value in the initial difference characterization value sequence is r; The method for obtaining the target difference characterization value sequence corresponding to the initial difference characterization value sequence includes: obtaining the target difference characterization value corresponding to each initial difference characterization value in the initial difference characterization value sequence according to each initial difference characterization value in the initial difference characterization value sequence and the first tag value of each initial difference characterization value; recording the sequence constructed by the target difference characterization value corresponding to each initial difference characterization value in the initial difference characterization value sequence as the target difference characterization value sequence corresponding to the corresponding initial difference characterization value sequence; The method for obtaining the target difference characterization value corresponding to each initial difference characterization value in the initial difference characterization value sequence includes: For the f-th initial difference characterization value in any initial difference characterization value sequence F: the remaining initial difference characterization value sequences except the initial difference characterization value sequence F are all recorded as sequences to be analyzed; the initial difference characterization values ​​in the sequence to be analyzed that have the same first label value as the f-th initial difference characterization value are all recorded as the characterization values ​​to be analyzed corresponding to the f-th initial difference characterization value; the cumulative sum of all the characterization values ​​to be analyzed corresponding to the f-th initial difference characterization value is recorded as the first cumulative characterization value; the result of subtracting the first cumulative characterization value from the f-th initial difference characterization value is recorded as the first index value; positive correlation mapping is performed on the first index value, and the mapping result is recorded as the target difference characterization value corresponding to the f-th initial difference characterization value; The method for obtaining the difference degree representation value corresponding to each target difference representation value in the target difference representation value sequence includes: Obtaining a second labeling value corresponding to each target difference characterizing value in the target difference characterizing value sequence, wherein the second labeling value corresponding to the d-th target difference characterizing value in the target difference characterizing value sequence is d; For any target difference characterization value g in any target difference characterization value sequence G: in the target difference characterization value sequence G, the sequence formed by all target difference characterization values ​​located before the target difference characterization value g is recorded as the historical sequence of the target difference characterization value g; according to the historical sequence, the second eigenvalue sequence of the target difference characterization value g is obtained, and the mean of the second eigenvalue sequence of the target difference characterization value g is recorded as the difference degree characterization value corresponding to the target difference characterization value g; the hth second eigenvalue in the second eigenvalue sequence is the product of the distance weight value corresponding to the hth target difference characterization value in the historical sequence and its corresponding first eigenvalue, the distance weight value corresponding to the hth target difference characterization value is the reciprocal of the absolute value of the difference between the second mark value of the hth target difference characterization value and the target difference characterization value g, and the first eigenvalue corresponding to the hth target difference characterization value is the absolute value of the difference between the hth target difference characterization value and the target difference characterization value g.

2. The online monitoring system for motion data of sandblasting equipment based on artificial intelligence as claimed in claim 1, characterized in that: The method for obtaining the abnormality degree characterization value set includes: According to the difference degree representation values ​​corresponding to the target difference representation values ​​in the target difference representation value sequence, the abnormality degree representation values ​​corresponding to the target difference representation value sequence are obtained; and a set constructed by the abnormality degree representation values ​​corresponding to all the target difference representation value sequences is recorded as the abnormality degree representation value set.

3. The online monitoring system for motion data of sandblasting equipment based on artificial intelligence as claimed in claim 2, characterized in that: The method for obtaining the abnormality degree representation value corresponding to the target difference representation value sequence includes: For any target difference characterization value sequence G: sort all the difference degree characterization values ​​in the target difference characterization value sequence G in ascending order, record the sorted sequence as the first sequence, and use the median of the first sequence as the reference value of the target difference characterization value sequence G; record the difference degree characterization value corresponding to the last target difference characterization value in the target difference characterization value sequence G and the difference degree characterization value corresponding to the second to last target difference characterization value as the first difference degree characterization value and the second difference degree characterization value, respectively; record the ratio of the first difference degree characterization value to the second difference degree characterization value as the first ratio; record the ratio of the first difference degree characterization value to the reference value as the second ratio; record the normalized value of the result obtained by multiplying the first ratio by the second ratio as the abnormality degree characterization value corresponding to the target difference characterization value sequence G.

4. The online monitoring system for motion data of sandblasting equipment based on artificial intelligence as claimed in claim 1, characterized in that: The method for obtaining a result of monitoring the working state of the sandblasting equipment according to the abnormality degree characterization value in the abnormality degree characterization value set includes: If it is determined that there is an abnormality degree characterization value greater than a preset abnormality degree threshold in the abnormality degree characterization value set, it is determined that the sandblasting equipment at the current moment is abnormal; otherwise, it is determined that the sandblasting equipment at the current moment is not abnormal.

Images