Multi-shaft motor drive fault real-time monitoring and early warning system

Abstract

The application discloses a multi-shaft motor driving fault real-time monitoring and early warning system and relates to the technical field of data processing. The multi-shaft motor driving fault real-time monitoring and early warning system comprises an acquisition module, a processing module and a control module. The acquisition module is used for acquiring target vibration data of a multi-shaft motor. The processing module is used for acquiring data abnormality features based on the target vibration data, acquiring initial significant values based on the data abnormality features and correcting the initial significant values based on feature data of vibration reasons of the multi-shaft motor to acquire corrected significant values. The control module is used for performing abnormality alarm based on the corrected significant values. The application is based on the distinguishing features of vibration data caused by motor load variation and motor driving faults, the distinguishing features are analyzed, the influence of interference factors is excluded, the vibration abnormality degree of monitoring data is acquired and the significant values are corrected. The application can avoid false detection and missed detection.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, in particular to a multi-axis motor drive fault real-time monitoring and early warning system. BACKGROUND

[0002] The multi-axis motor drive system is a key component of production, and its stability is directly related to the operation efficiency of the entire production line. Real-time monitoring of motor vibration data and early warning of abnormal conditions are of great significance for preventing faults, reducing downtime, and improving production efficiency. In the prior art, a Context-Aware (CA) algorithm, i.e., a CA algorithm, is used to extract abnormal values of motor vibration data and to warn of abnormal data conditions. However, a large load can cause the motor vibration amplitude to increase significantly, which may be identified as an abnormal state, thereby triggering a false alarm. It is easy to understand that if a false alarm is triggered, the operation efficiency of the production line will also be affected. SUMMARY

[0003] The present application aims to provide a multi-axis motor drive fault real-time monitoring and early warning system to solve the technical problem of false detection in the prior art using the CA algorithm.

[0004] To achieve the above-mentioned purpose, the present application provides the following technical solution:

[0005] A multi-axis motor drive fault real-time monitoring and early warning system comprises:

[0006] An acquisition module for acquiring target vibration data of a multi-axis motor;

[0007] A processing module for acquiring data anomaly features based on the target vibration data; the data anomaly features at least include feature data representing the vibration cause of the multi-axis motor;

[0008] and acquiring initial saliency values based on the data anomaly features;

[0009] and correcting the initial saliency values based on the feature data of the vibration cause of the multi-axis motor to obtain corrected saliency values;

[0010] A control module for performing an abnormal alarm based on the corrected saliency values.

[0011] As one specific solution in the technical solution of the present application, the method for acquiring vibration data of a multi-axis motor comprises:

[0012] Acquiring current vibration data of the multi-axis motor at a frequency that is an integer multiple of the highest vibration frequency;

[0013] and transmitting the current vibration data to a processing module to enable the processing module to preprocess the current vibration data and obtain the target vibration data.

[0014] As a specific solution in the technical scheme of the present application, the method for obtaining the data anomaly feature based on the target vibration data comprises:

[0015] segmenting the target vibration data to obtain a plurality of data segments;

[0016] obtaining the fluctuation degree of each data segment based on the plurality of data segments;

[0017] obtaining the data anomaly feature based on the fluctuation degree of each data segment.

[0018] As a specific solution in the technical scheme of the present application, the method for segmenting the target vibration data to obtain a plurality of data segments comprises:

[0019] converting the target vibration data from time domain data to frequency domain data;

[0020] segmenting the target vibration data according to the period of the frequency domain data to obtain a plurality of data segments.

[0021] As a specific solution in the technical scheme of the present application, the method for obtaining the fluctuation degree of each data segment based on the plurality of data segments comprises:

[0022] obtaining a first data segment based on the plurality of data segments; the first data segment is any data segment in the plurality of data segments for which the fluctuation degree has not been obtained;

[0023] obtaining a fluctuation parameter based on the first data segment; the fluctuation parameter at least includes the fluctuation range, the dispersion degree and the fluctuation times of the first data segment;

[0024] obtaining the fluctuation degree of the first data segment based on the fluctuation parameter of the first data segment.

[0025] As a specific solution in the technical scheme of the present application, the calculation formula for obtaining the fluctuation degree of the first data segment based on the fluctuation parameter of the first data segment is as follows:

[0026]

[0027] wherein, B i represents the fluctuation degree of the i-th data segment; N represents the number of vibration data in the i-th data segment; D i,n represents the n-th vibration data value in the i-th data segment; μ i represents the mean value of all vibration data values in the i-th data segment; Di,max represents the maximum value of the vibration data value in the i-th data segment; D i,min represents the minimum value of the vibration data value in the i-th data segment; represents the average value of the fluctuation factor of the i-th data segment in time domain analysis, that is, the average value of the ratio of the peak value of the vibration data to the root mean square value.

[0028] As one specific scheme in the technical scheme of the present application, the method for obtaining the data anomaly feature based on the fluctuation degree of each data segment comprises:

[0029] Based on the plurality of data segments, a second data segment and a third data segment are obtained; the second data segment and the third data segment are data segments that have not been subjected to similarity comparison among the plurality of data segments;

[0030] Based on the second data segment and the third data segment, a difference value of corresponding vibration data in the second data segment and the third data segment is obtained;

[0031] Based on the difference value of corresponding vibration data in the second data segment and the third data segment, the fluctuation degree of the second data segment, and the fluctuation degree of the third data segment, a similarity of the second data segment and the third data segment is obtained.

[0032] Based on the similarity of the second data segment and the third data segment, the data anomaly feature is obtained.

[0033] As one specific scheme in the technical scheme of the present application, the calculation formula for obtaining the similarity of the second data segment and the third data segment based on the difference value of corresponding vibration data in the second data segment and the third data segment, the fluctuation degree of the second data segment, and the fluctuation degree of the third data segment is as follows:

[0034]

[0035] wherein, R a,b represents the similarity of the fluctuation degree between the a-th data segment and the b-th data segment; N represents the number of vibration data in each data segment; D a,n represents the n-th vibration data value in the a-th data segment; D b,n represents the n-th vibration data value in the b-th data segment; B a represents the fluctuation degree of the a-th data segment; B b represents the fluctuation degree of the b-th data segment; exp represents the exponential function with the natural constant e as the base; sigmoid represents the sigmoid function for mapping data into the interval (0, 1).

[0036] As a specific scheme in the technical scheme of the application, the data anomaly feature includes the vibration anomaly degree of each data; the method for obtaining the data anomaly feature based on the fluctuation degree of each data segment includes:

[0037] Based on the plurality of data segments, a fourth data segment is obtained; the fourth data segment is any data segment in the plurality of data segments that does not have a vibration anomaly degree;

[0038] Based on the fourth data segment, a first similarity and a second similarity are obtained; the first similarity is the similarity between the fourth data segment and the previous data segment; the second similarity is the similarity between the previous two data segments of the fourth data segment;

[0039] Based on the fourth data segment, the number of data segments with a similarity greater than a first threshold value to the fourth data segment is obtained;

[0040] Based on the first similarity, the second similarity, and the number, the vibration anomaly degree of the fourth data segment is obtained.

[0041] As a specific scheme in the technical scheme of the application, the method for correcting the initial significant value to obtain a modified significant value based on the feature data of the vibration cause of the multi-axis motor includes:

[0042] Based on the target vibration data, a first vibration data is obtained; the first vibration data is any vibration data in the target vibration data;

[0043] Based on the first vibration data, the vibration anomaly degree of the data segment corresponding to the first vibration data is obtained;

[0044] Based on the initial significant value and the vibration anomaly degree, a modified significant value is obtained.

[0045] Compared with the prior art, the application has the following beneficial effects:

[0046] The application is based on the distinguishing features of vibration data caused by changes in motor load and motor drive faults. By analyzing the above distinguishing features, the influence of interference factors is excluded, the vibration anomaly degree of the monitoring data is obtained, and the significant value is corrected. As a result, the abnormal state caused by motor drive faults is extracted for early warning; the abnormal state caused by changes in motor load is not extracted. That is, the application can not only avoid false detection, but also avoid missed detection. BRIEF DESCRIPTION OF DRAWINGS

[0047] Figure 1 A flowchart of a multi-axis motor drive fault real-time monitoring and early warning method according to an embodiment of the application is shown;

[0048] Figure 2A structure diagram of a multi-shaft motor driving fault real-time monitoring and early warning system is provided in the embodiments of the present application. DETAILED DESCRIPTION

[0049] The technical solutions in the embodiments of the present application will be clearly and completely described with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative effort fall within the protection scope of the present application.

[0050] The technical solutions in the embodiments of the present application will be clearly and completely described with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative effort fall within the protection scope of the present application.

[0051] The terms "first", "second", etc. in the description and claims of the present application and the above drawings are used to distinguish similar objects, and do not necessarily indicate a specific order or sequence, for example, the first data segment and the second data segment proposed below belong to different data segments. It should be understood that the data segments used in this way can be interchanged under appropriate circumstances, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device that includes a series of steps or modules does not necessarily have to be limited to those steps or modules clearly listed, but can include other steps or modules that are not clearly listed or inherent to these processes, methods, products or devices. The division of modules in the embodiments of the present application is only a logical division, and in actual application, another division mode can be used, for example, a plurality of modules can be combined or integrated into another system, or some features can be ignored or not executed, in addition, the coupling or direct coupling or communication connection between the displayed or discussed modules can be through an interface, the indirect coupling or communication connection between the modules can be electrical or other similar forms, which are not limited in the embodiments of the present application. In addition, the modules or sub-modules described as separate components can or can not be physically separated, can or can not be physical modules, or can be distributed into a plurality of circuit modules, and part or all of the modules can be selected according to actual needs to achieve the purpose of the present application.

[0052] In order to solve the technical problem that false detection is easy to occur when the CA algorithm is used to monitor and early warn the multi-shaft motor driving fault in real time in the prior art, an embodiment of a multi-shaft motor driving fault real-time monitoring and early warning method is provided. Specifically, as shown in Figure 1 The method comprises steps S100 to S500.

[0053] Step S100: Obtain target vibration data of a multi-shaft motor.

[0054] It should be noted that in the embodiment of the present application, the target vibration data can be actual vibration data generated by the multi-shaft motor during use. Generally, the vibration sensor is installed near the motor bearing and is corrected so that it can capture stable actual vibration data of the multi-shaft motor. Since collecting vibration data of the multi-shaft motor is a mature technology, it will not be described in detail here.

[0055] In order to be able to obtain all vibration data of the multi-shaft motor, in the embodiment of the present application, the frequency at which the vibration sensor obtains the vibration data can be higher than the highest vibration frequency of the multi-shaft motor. For example, in an embodiment of the present application, the frequency at which the vibration sensor obtains the vibration data can be an integer multiple of the highest vibration frequency of the multi-shaft motor.

[0056] It should be noted that in most cases, the actual vibration data obtained based on the vibration sensor has a large amount of noise. In order to be able to denoise the obtained actual vibration data, in an embodiment of the present application, step S100 of obtaining target vibration data of a multi-shaft motor comprises steps S110 and S120.

[0057] Step S110: Obtain current vibration data of the multi-shaft motor at a frequency of twice the highest vibration frequency.

[0058] It should be noted that the current vibration data of the multi-shaft motor, i.e., the actual vibration data generated by the multi-shaft motor during use, is known from the foregoing. If the frequency at which the vibration sensor obtains the vibration data is lower than the highest vibration frequency of the multi-shaft motor, all vibration data of the multi-shaft motor cannot be obtained completely. If the frequency at which the vibration sensor obtains the vibration data is much higher than the highest vibration frequency of the multi-shaft motor, energy waste is easy to occur. Through the inventor's repeated research and attempts, the current vibration data of the multi-shaft motor is obtained at a frequency of twice the highest vibration frequency, which can reduce the energy consumption in the monitoring process to a minimum while obtaining most of the vibration data of the multi-shaft motor completely.

[0059] Step S120: Transmit the current vibration data to a processing module, so that the processing module can preprocess the current vibration data to obtain the target vibration data.

[0060] It needs to be clear that in the embodiments of the present application, the purpose of preprocessing the current vibration data is to denoise. In the embodiments of the present application, the current vibration data can be denoised in any manner. For example, standard deviation denoising, binning denoising, DBSCAN denoising or isolation forest denoising can be used. Among them, the standard deviation denoising is based on statistical principles, by calculating the mean and standard deviation of the data, and then setting a threshold (usually 3 times the standard deviation), the data beyond the threshold is considered as noise and removed. Binning denoising smooths data values by examining the "neighbors" of the data (i.e. the surrounding values). The specific method is to divide the data into multiple intervals (i.e. boxes) according to certain rules, and then replace all data in each box with the median (or other statistical quantity) of the box. DBSCAN denoising is a density-based clustering algorithm, which discovers clusters of arbitrary shape by calculating the density between data points, and considers data points not belonging to any cluster as noise. Isolation forest denoising is an anomaly detection method based on ensemble learning, which detects outliers in the data by constructing multiple decision trees. In the process of constructing the decision tree, the outliers are usually quickly isolated, so the path length of each data point in the decision tree can be used to judge whether it is an outlier (i.e. noise).

[0061] Step S200: obtaining data anomaly features based on the target vibration data.

[0062] It needs to be clear that the multi-axis motor drive system is an advanced automation control system that can accurately control the operation of multiple motors to achieve complex mechanical movements and operations. The multi-axis motor drive system is usually composed of multiple motors, respective drivers, a centralized controller, and feedback devices. Specifically, the controller receives operation instructions or program settings from the user interface, sensor feedback, or automation control system, the processor analyzes these instructions, and generates corresponding control signals according to the predetermined motion control logic. Further, the driver provides appropriate voltage and current to each motor according to the control signal, accurately controls the speed and torque of the motor. The motor then executes the instructions to perform the predetermined motion, such as rotation, acceleration, or deceleration, etc. At the same time, the encoder or other feedback devices in the multi-axis motor drive system monitor the motion state of the motor in real time, including position, speed, and acceleration, etc., and feed back these information to the controller. The controller processes the feedback signal, compares it with the instruction signal, and makes necessary corrections to ensure that the motion of the motor can reach the expected accuracy and synchronization. Throughout the process, the multi-axis motor drive system also needs to communicate with external devices or networks to achieve more extensive automation and integrated control. Through this series of precise coordination steps, the multi-axis motor drive system can achieve efficient automation production and precise control, meet the use requirements of various application scenarios such as industrial automation, robotics, CNC machine tools, aerospace, medical devices, 3D printing, etc., and is an indispensable key technology in the field of modern intelligent manufacturing and precise control.

[0063] To ensure the efficient and accurate operation of the multi-axis motor drive system, the motor drive state needs to be monitored in real time, and the change of motor vibration data is usually a direct reflection of internal problems or performance degradation of the motor. By monitoring motor vibration data, potential fault problems of the motor can be identified in advance, thereby preventing equipment failure, improving production efficiency, and reducing maintenance costs, etc.

[0064] Motor drive failure abnormal conditions are often reflected in abnormal vibration data within a short period of time, and there is no problem with only one vibration data value, so the motor vibration time series data needs to be segmented to extract features for analysis. The fluctuation of motor vibration data can reflect the stability of motor operation, and abnormal fluctuations may indicate that the motor condition has a problem. By monitoring the fluctuation degree of vibration, it can be roughly identified whether the motor has a fault. In other words, if the target vibration data has data anomaly features, it means that the motor operation has a problem (e.g., motor failure or abnormal load). That is, in the embodiments of the present application, data anomaly features need to be obtained based on the target vibration data to determine whether the motor operation has an abnormality.

[0065] It should be noted that in the embodiments of the present application, the data anomaly feature at least includes feature data for characterizing the vibration cause of the multi-axis motor. Any method can be used to obtain the data anomaly feature based on the target vibration data. For example, in the embodiments of the present application, the data anomaly feature can be obtained directly based on the target vibration data by the CA algorithm. In another embodiment of the present application, in order to accurately determine the cause of the data anomaly feature leading to the motor target vibration data, step S200, the method of obtaining the data anomaly feature based on the target vibration data includes steps S210 to S230.

[0066] Step S210: segmenting the target vibration data to obtain a plurality of data segments.

[0067] It is easy to understand that the main problem of using the saliency detection algorithm to obtain the data anomaly feature is that the vibration data fluctuation caused by too large or too small load may be misdetected as motor fault anomaly. That is, the saliency detection algorithm cannot stably distinguish between the vibration fluctuation generated by the motor in the normal use process and the vibration fluctuation generated by the fault.

[0068] It is easy to understand that the vibration fluctuation of the motor in the normal use process is generally repeatedly generated in large quantities, while the vibration fluctuation of the motor due to fault is sudden and has no premonition before and after. That is, in the embodiments of the present application, the target vibration data can be segmented to obtain a plurality of data segments. The similarity of the fluctuation between the data segment with obvious vibration and other data segments is used to assist in determining whether the data segment with obvious vibration is the vibration fluctuation in the normal use process or the vibration fluctuation generated by the fault. It is easy to understand that if the data segment with obvious vibration has high similarity with the data segments before and after it, it means that the abnormal vibration is caused by the change of motor load; if the data segment with obvious vibration has low similarity with the data segments before and after it, it means that the abnormal vibration is caused by motor fault.

[0069] It should be noted that in the embodiments of the present application, the target vibration data can be segmented in any way. For example, the target vibration data is segmented in equal time intervals. In order to accurately determine whether each data segment is similar, in one embodiment of the present application, step S210 of segmenting the target vibration data to obtain a plurality of data segments includes steps S211 and S212.

[0070] Step S211: converting the target vibration data from time domain data to frequency domain data.

[0071] It needs to be clear that by transforming the time domain data into the frequency domain data, the period of the motor vibration data can be clearly obtained, and then each data can be segmented according to the period. Segmenting the target vibration data according to the period of the motor vibration data is beneficial to the similarity comparison of each data segment obtained subsequently.

[0072] In the embodiments of the present application, the target vibration data can be converted from time domain data to frequency domain data in any way. For example, in the embodiments of the present application, Fourier transform or Laplace transform can be used to convert the target vibration data from time domain data to frequency domain data. Fourier transform is a mathematical method of expressing a certain function satisfying certain conditions as a linear combination of trigonometric functions (sine and / or cosine functions) or their integrals. Since Fourier transform and Laplace transform are mature technologies, they will not be described in detail here.

[0073] Step S212: segmenting the target vibration data according to the period of the frequency domain data to obtain a plurality of data segments.

[0074] It needs to be clear that in the embodiments of the present application, each data segment can contain one period of motor vibration data, or can include an integer multiple of other number of periods of motor vibration data. In order to accurately determine the cause of each abnormal vibration, each data segment contains only one period of motor vibration data.

[0075] Step S220: obtaining the fluctuation degree of each data segment based on a plurality of the data segments.

[0076] It needs to be clear that in the embodiments of the present application, the main purpose of obtaining the fluctuation degree of each data segment is to determine whether the fluctuations of two adjacent data segments are similar subsequently. In the embodiments of the present application, the fluctuation degree can be any data that can reflect the fluctuation of the data segment. For example, in one embodiment of the present application, the fluctuation degree of each data segment can be an image data of a waveform. In another embodiment of the present application, the fluctuation degree of each data segment can be the fluctuation range, dispersion degree and time interval of the vibration waveform of the data segment.

[0077] In one specific embodiment of the present application, step S220, the method for obtaining the fluctuation degree of each data segment based on a plurality of the data segments includes steps S221 to S223.

[0078] Step S221: obtaining a first data segment based on a plurality of the data segments.

[0079] In the embodiments of the present application, the fluctuation degree of each data segment in the plurality of data segments needs to be obtained in sequence or randomly. Therefore, in the present embodiment, the first data segment is any data segment in the plurality of data segments whose fluctuation degree has not been obtained.

[0080] Step S222: obtaining a fluctuation parameter based on the first data segment.

[0081] It should be noted that the fluctuation parameter at least includes a fluctuation range, a dispersion degree and a fluctuation frequency of the first data segment. It can be easily understood that if the similarity of two data segments is high, the fluctuation range, the dispersion degree and the fluctuation frequency of the two data segments are relatively close. Therefore, subsequently, whether two data segments are similar can be determined based on the fluctuation parameters of the data segments.

[0082] Step S223: obtaining a fluctuation degree of the first data segment based on the fluctuation parameter of the first data segment.

[0083] In the embodiments of the present application, the fluctuation degree of the first data segment can be obtained in any manner based on the fluctuation parameter. For example, in an embodiment of the present application, the fluctuation degree of the first data segment can be represented by only the dispersion degree and the fluctuation frequency of the two data segments. Alternatively, the fluctuation degree of the first data segment can be represented by only the fluctuation range and the dispersion degree.

[0084] In a specific embodiment of the present application, step S223, the calculation formula of the fluctuation degree of the first data segment based on the fluctuation parameter of the first data segment is as follows:

[0085]

[0086] wherein, B i represents the fluctuation degree of the i-th data segment (i.e. the first data segment); N represents the number of fluctuation data in the i-th data segment (i.e. the fluctuation frequency); D i,n represents the n-th fluctuation data value in the i-th data segment; μ i represents the mean value of all fluctuation data values in the i-th data segment; D i,max represents the maximum value of the fluctuation data values in the i-th data segment; D i,min represents the minimum value of the fluctuation data values in the i-th data segment; represents the mean value of the fluctuation factor in the i-th data segment in time domain analysis, i.e. the mean value of the ratio of the peak value of the fluctuation data to the root mean square (RMS) value.

[0087] It should be noted that in the above formula, B represents the center position of the fluctuation data in the i-th data segment (i.e. the first data segment); represents the deviation degree of each fluctuation data value from the mean value in the i-th data segment, reflecting the fluctuation range and the dispersion degree of the fluctuation data; |D i,max -D i,min| represents the difference between the peak and the valley of the i-th data segment, reflecting the extreme fluctuation range of the vibration. It is easy to understand that the more similar the waveforms of two data segments are, the closer the fluctuation degrees of the two data segments are.

[0088] Step S230: Obtain the data anomaly feature based on the fluctuation degree of each data segment.

[0089] It needs to be clear that in the embodiments of the present application, the main purpose of obtaining the data anomaly feature is to obtain a saliency value based on the CA algorithm. If the saliency value is greater than the preset alarm threshold, it means that the motor vibration is abnormal and an alarm needs to be sent. If the saliency value is less than the preset alarm threshold, it means that the motor vibration is normal and no alarm needs to be sent.

[0090] In the embodiments of the present application, the cause of the abnormal vibration of the motor (change of motor load or motor failure) also needs to be determined based on the data anomaly feature. As known from the foregoing, in the embodiments of the present application, the cause of the abnormal vibration of the motor can be determined based on the similarity between two data segments. Therefore, in the embodiments of the present application, the data anomaly feature can be the similarity between two data segments. It is easy to understand that if the similarity between the data segment with obvious vibration and other data segments (including the two data segments before and after the data segment with obvious vibration) is low, it can be considered that the cause of the vibration of the data segment is most likely to be motor failure. If the similarity between the data segment with obvious vibration and other data segments is low, it can be considered that the cause of the vibration of the data segment is most likely to be the change of motor load.

[0091] In the embodiments of the present application, if the fluctuation degree of each data segment is a data image of the waveform, the similarity between each data segment can be obtained through an artificial intelligence model. It needs to be clear that the similarity of data images obtained through an artificial intelligence model (such as a twin network model or a triplet network model) is a mature technology, which will not be described here. Of course, the similarity of data images of two data segments can also be directly obtained through an image similarity algorithm. For example, a histogram algorithm, a grayscale algorithm, a hash algorithm, etc.

[0092] In a specific embodiment of the present application, step S230, the method for obtaining the data anomaly feature based on the fluctuation degree of each data segment includes steps S231 to S234.

[0093] Step S231: Obtain a second data segment and a third data segment based on a plurality of data segments.

[0094] It should be noted that in the embodiments of the present application, the second data segment and the third data segment are data segments that are not subjected to similarity comparison among the plurality of data segments. That is, in the embodiments of the present application, the second data segment and the third data segment can be adjacent data segments or non-adjacent data segments.

[0095] Step S232: obtaining a difference value of corresponding vibration data in the second data segment and the third data segment based on the second data segment and the third data segment.

[0096] It should be noted that the higher the similarity of the second data segment and the third data segment, the more similar the vibration waveforms of the second data segment and the third data segment, and the more similar the number of vibration data in the second data segment and the third data segment; the lower the similarity of the second data segment and the third data segment, the greater the difference in the vibration waveforms of the second data segment and the third data segment, and the greater the difference in the number of vibration data in the second data segment and the third data segment. That is, in the embodiments of the present application, the similarity of the second data segment and the third data segment can be evaluated based on the difference value of corresponding vibration data in the second data segment and the third data segment.

[0097] Step S233: obtaining the similarity of the second data segment and the third data segment based on the difference value of corresponding vibration data in the second data segment and the third data segment, the fluctuation degree of the second data segment, and the fluctuation degree of the third data segment.

[0098] In the embodiments of the present application, the similarity of the second data segment and the third data segment can be obtained based on the difference value of corresponding vibration data in the second data segment and the third data segment, the fluctuation degree of the second data segment, and the fluctuation degree of the third data segment by any method.

[0099] In one specific embodiment of the present application, in step S233, the calculation formula for obtaining the similarity of the second data segment and the third data segment based on the difference value of corresponding vibration data in the second data segment and the third data segment, the fluctuation degree of the second data segment, and the fluctuation degree of the third data segment is as follows:

[0100]

[0101] wherein R a,b represents the similarity of the fluctuation degree between the a-th data segment (i.e., the second data segment) and the b-th data segment (i.e., the third data segment); N represents the number of vibration data in each data segment; D a,n represents the n-th vibration data value in the a-th data segment; D b,n represents the n-th vibration data value in the b-th data segment; B arepresents the fluctuation degree of the a-th data segment; B b represents the fluctuation degree of the b-th data segment; exp represents the exponential function with the natural constant e as the base; sigmoid represents the sigmoid function for mapping data into the interval (0, 1).

[0102] It is easily understood that, represents the mean value of the difference between each corresponding vibration data value between the a-th data segment and the b-th data segment, which reflects the difference degree between the vibration data in the two data segments. The smaller the value is, the smaller the difference degree between the a-th data segment and the b-th data segment is, that is, the more similar the two data segments are. a -B b represents the difference between the fluctuation degrees between the a-th data segment and the b-th data segment, which reflects the difference between the fluctuation degrees of the data in the two data segments. The smaller the value is, the higher the similarity of the fluctuation degrees of the data in the a-th data segment and the b-th data segment is.

[0103] Step S234: obtaining the data anomaly feature based on the similarity between the second data segment and the third data segment.

[0104] It should be noted that the data anomaly feature can be directly obtained based on the similarity between the second data segment and the third data segment. For example, if the similarity between the second data segment and the third data segment is greater than a preset threshold, it is determined that the abnormal vibration generated by the second data segment or the third data segment is the change of the motor load, and no alarm needs to be issued; if the similarity between the second data segment and the third data segment is less than the preset threshold, it is determined that the abnormal vibration generated by the second data segment or the third data segment is the motor fault, and an alarm needs to be issued.

[0105] It should be noted that in another embodiment of the present application, the data anomaly feature includes the vibration anomaly degree of each data. Step S230, the method for obtaining the data anomaly feature based on the fluctuation degree of each data segment further includes steps S235 to S238.

[0106] Step S235: obtaining a fourth data segment based on a plurality of data segments.

[0107] It should be noted that in the embodiment of the present application, the fourth data segment is any data segment in the plurality of data segments that does not have a vibration anomaly degree.

[0108] Step S236: obtaining a first similarity and a second similarity based on the fourth data segment.

[0109] In the embodiments of the present application, the first similarity is the similarity between the fourth data segment and the previous data segment; and the second similarity is the similarity between the previous two data segments of the fourth data segment. It can be easily understood that when the motor load changes greatly, the fluctuation of the vibration data will also change significantly accordingly. However, the fluctuation of the vibration data caused by the load exists excessively in the entire time sequence vibration data sequence, and is similar to the fluctuation degree of the adjacent vibration data segment. The abnormality caused by the motor fault occurs suddenly, and is relatively disconnected with the fluctuation of the adjacent vibration data segment, and the similarity is low. Therefore, according to the similarity between each vibration data segment and the adjacent vibration data segment, the vibration abnormality degree of the data can be obtained.

[0110] Step S237: based on the fourth data segment, the number of data segments with a similarity greater than a first threshold value to the fourth data segment is obtained.

[0111] It can be easily understood that the fluctuation of the vibration caused by the increase of the load is more common than the fluctuation of the vibration caused by the motor fault in the historical monitoring vibration data. Therefore, the number of the data segments with a similar fluctuation degree is counted from the historical monitoring vibration data. If the number accounts for a small proportion, the vibration data is likely to be an abnormal value; and if the number accounts for a large proportion, the vibration data is likely to be a normal value.

[0112] Step S238: based on the first similarity, the second similarity and the number, the vibration abnormality degree of the fourth data segment is obtained.

[0113] In the embodiments of the present application, the vibration abnormality degree of the fourth data segment can be obtained in any form with the first similarity, the second similarity and the number of data segments with a similarity greater than a first threshold value to the fourth data segment. In one specific embodiment of the present application, step S238, based on the first similarity, the second similarity and the number, the calculation formula of the vibration abnormality degree of the fourth data segment is as follows:

[0114]

[0115] wherein, Z i represents the vibration abnormality degree of the i-th data segment; R i,i-1 represents the fluctuation similarity between the i-th data segment and the adjacent i-1-th data segment; R i-1,i-2 represents the fluctuation similarity between the i-1-th data segment and the adjacent i-2-th data segment; m i represents the number of data segments with a fluctuation similarity greater than a first threshold value between the i-th data segment and each data segment; M represents the total number of the motor vibration data segments for analysis; and sigmoid represents a sigmoid function for mapping data into the interval (0, 1).

[0116] In the embodiments of the present application, the first threshold value is not limited, which can be an empirical value or an experimental value obtained through a large number of experiments. In a specific embodiment of the present application, the first threshold value can be any one of 0.90, 0.91, 0.92, 0.93, 0.94 and 0.95, or any value between any two adjacent values.

[0117] Specifically, in the embodiments of the present application, |R i,i-1 -R i-1,i-2 | represents the absolute value of the difference between the first similarity and the second similarity, the abnormal fluctuation caused by the motor drive fault is a sudden change, which has a small similarity with the previous segment, and a high similarity between the adjacent two segments. The greater the difference between the two fluctuation similarities, the greater the vibration abnormality degree of the i-th data segment. represents the proportion of the vibration data segment with high similarity to the i-th data segment in the overall motor vibration data segment. The smaller the formula, the less likely it is that similar fluctuation abnormalities have occurred before, and the higher the vibration abnormality degree of the motor vibration data segment.

[0118] Step S300: obtaining an initial significant value based on the data anomaly feature.

[0119] It should be clear that obtaining an initial significant value based on the data anomaly feature using the CA algorithm is a mature technology, and will not be described in detail here.

[0120] Step S400: correcting the initial significant value based on the feature data of the vibration cause of the multi-axis motor to obtain a corrected significant value.

[0121] It should be clear that in the prior art, whether the initial significant value is greater than the second threshold value is directly used to determine whether the motor has failed. If the initial significant value is greater than the second threshold value, it is determined that the motor has failed; if the initial significant value is less than the second threshold value, it is determined that the motor has not failed. As known from the foregoing, the direct determination using the initial significant value may also determine the vibration caused by the change in motor load as abnormal vibration. In order to avoid frequent false positives of the system, the second threshold value can only be increased; if the second threshold value is increased, even if the motor fails and produces abnormal vibration, it is difficult to be monitored.

[0122] In order to accurately monitor the motor failure after increasing the second threshold value, in an embodiment of the present application, the initial significant value can be corrected based on some data in the foregoing. It is easy to understand that if it is determined from the foregoing that the cause of the vibration of the vibration data in the data segment is motor failure, the significant value corresponding to the vibration data can be increased, thereby avoiding the vibration being undetectable, i.e., avoiding false negatives.

[0123] In a specific embodiment of the present application, step S400, the initial saliency value is corrected based on the characteristic data of the vibration cause of the multi-shaft motor, and the method for obtaining the corrected saliency value comprises steps S410 to S430.

[0124] Step S410: based on the target vibration data, first vibration data is obtained.

[0125] It should be clear that the first vibration data is any vibration data in the target vibration data.

[0126] Step S420: based on the first vibration data, the vibration abnormality degree of the data segment corresponding to the first vibration data is obtained.

[0127] It should be clear that, as known from the foregoing, the vibration abnormality degree of each data segment can be calculated based on step S238, which will not be repeated here.

[0128] Step S430: based on the initial saliency value and the vibration abnormality degree, a corrected saliency value is obtained.

[0129] As known from the foregoing, the cause of the abnormal vibration generated by the corresponding vibration data can be confirmed based on the vibration abnormality degree. If the cause of the abnormal vibration generated by the vibration data is motor failure, the value of the initial saliency value can be increased to obtain a corrected saliency value with a larger value, thereby avoiding false negatives caused by a small correction value.

[0130] In an embodiment of the present application, the corrected saliency value can be obtained based on the initial saliency value and the vibration abnormality degree in any way. In a specific embodiment of the present application, step S430, the corrected saliency value can be obtained based on the initial saliency value and the vibration abnormality degree by the following formula:

[0131] ZX j = sigmoid{X j ×(1+Z j,g )}

[0132] ZX j represents the corrected saliency value of the jth motor vibration data value; X j represents the initial saliency value of the jth motor vibration data value obtained by the saliency detection CA algorithm; Z j,g represents the vibration abnormality degree of the gth data segment in which the jth motor vibration data value is located. In the formula, X j ×(1+Z j,g ) represents that the saliency value is corrected according to the vibration abnormality degree of the data segment in which the vibration data value is located. The greater the vibration abnormality degree, the higher the corrected saliency value; sigmoid represents a sigmoid function for mapping data to the interval (0, 1).

[0133] It needs to be clear that the vibration fluctuation caused by the large change of the load amount may have a high significant value calculated in the significant detection CA algorithm, and the vibration data fluctuation anomaly caused by the motor fault also has a high significant value, and the two cannot be distinguished according to the initial significant value. In order to exclude the influence of interference factors and only keep the vibration data of the motor fault, the significant value is corrected by analyzing the vibration anomaly degree of the monitored data in the segment, so that the significant value of the vibration data caused by the motor drive fault is higher, and a higher second threshold is set, which can avoid false detection and also avoid missed detection.

[0134] Step S500: performing an anomaly alarm based on the corrected significant value.

[0135] It needs to be clear that the anomaly alarm based on the significant value and the second threshold is a mature technology, which will not be described here.

[0136] It needs to be clear that the multi-axis motor drive fault real-time monitoring and early warning method proposed in the present application is based on the distinguishing features of the vibration data caused by the change of the motor load amount and the motor drive fault, and by analyzing the above distinguishing features, the influence of interference factors is excluded, the vibration anomaly degree of the monitored data is obtained, and the significant value is corrected, so that the abnormal state caused by the motor drive fault is extracted and early warning is performed; the abnormal state caused by the change of the motor load amount is not extracted. That is, the multi-axis motor drive fault real-time monitoring and early warning method proposed in the present application can avoid false detection and also avoid missed detection.

[0137] After introducing the embodiment of the multi-axis motor drive fault real-time monitoring and early warning method proposed in the present application, the embodiment of a multi-axis motor drive fault real-time monitoring and early warning system proposed in the present application is introduced as follows. Figure 2 As shown in the figure, the multi-axis motor drive fault real-time monitoring and early warning system 10 includes:

[0138] The acquisition module 11 is configured to acquire target vibration data of a multi-axis motor.

[0139] The processing module 12 is configured to acquire data anomaly features based on the target vibration data, wherein the data anomaly features at least include feature data for representing vibration reasons of the multi-axis motor.

[0140] The processing module 12 is further configured to acquire an initial significant value based on the data anomaly features.

[0141] The processing module 12 is further configured to correct the initial significant value based on the feature data for representing the vibration reasons of the multi-axis motor, and acquire a corrected significant value.

[0142] The control module 13 is configured to perform an anomaly alarm based on the corrected significant value.

[0143] In one specific embodiment of the present application, the acquisition module 11 is further configured to acquire the current vibration data of the multi-axis motor at a frequency of an integer multiple of the highest vibration frequency;

[0144] and transmit the current vibration data to the processing module, so that the processing module can preprocess the current vibration data and acquire the target vibration data.

[0145] In one specific embodiment of the present application, the processing module 12 is further configured to segment the target vibration data to obtain a plurality of data segments;

[0146] and acquire the fluctuation degree of each data segment based on the plurality of data segments;

[0147] and acquire the data anomaly feature based on the fluctuation degree of each data segment.

[0148] In one specific embodiment of the present application, the processing module 12 is further configured to convert the target vibration data from time domain data to frequency domain data;

[0149] and segment the target vibration data according to the period of the frequency domain data to obtain a plurality of data segments.

[0150] In one specific embodiment of the present application, the processing module 12 is further configured to acquire a first data segment based on the plurality of data segments; the first data segment is any data segment in the plurality of data segments for which the fluctuation degree is not acquired;

[0151] and acquire a fluctuation parameter based on the first data segment; the fluctuation parameter at least includes the fluctuation range, the dispersion degree and the fluctuation times of the first data segment;

[0152] and acquire the fluctuation degree of the first data segment based on the fluctuation parameter of the first data segment.

[0153] In one specific embodiment of the present application, the processing module 12 is further configured to acquire the fluctuation degree of the first data segment by the following formula.

[0154]

[0155] wherein, B i represents the fluctuation degree of the i-th data segment (i.e. the first data segment) in each data segment; N represents the number of vibration data (i.e. the fluctuation times) in the i-th data segment; D i,n represents the n-th vibration data value in the i-th data segment; μ i represents the mean value of all vibration data values in the i-th data segment; D i,maxrepresents the maximum value of the vibration data value in the ith data segment; D i,min represents the minimum value of the vibration data value in the ith data segment; represents the average value of the fluctuation factor in the time domain analysis of the ith data segment, i.e., the average value of the ratio of the peak value of the vibration data to the root mean square value (RMS).

[0156] In an embodiment of the present application, the processing module 12 is further configured to obtain a second data segment and a third data segment based on the plurality of data segments; the second data segment and the third data segment are data segments that have not been subjected to similarity comparison among the plurality of data segments;

[0157] and obtain a difference value of corresponding vibration data in the second data segment and the third data segment based on the second data segment and the third data segment;

[0158] and obtain a similarity of the second data segment and the third data segment based on the difference value of corresponding vibration data in the second data segment and the third data segment, the fluctuation degree of the second data segment, and the fluctuation degree of the third data segment;

[0159] and obtain the data anomaly feature based on the similarity of the second data segment and the third data segment.

[0160] In an embodiment of the present application, the processing module 12 is further configured to obtain the similarity of the second data segment and the third data segment according to the following formula.

[0161]

[0162] wherein, R a,b represents the similarity of the fluctuation degree between the ath data segment (i.e., the second data segment) and the bth data segment (i.e., the third data segment); N represents the number of vibration data in each data segment; D a,n represents the nth vibration data value in the ath data segment; D b,n represents the nth vibration data value in the bth data segment; B a represents the fluctuation degree of the ath data segment; B b represents the fluctuation degree of the bth data segment; exp represents the exponential function with the natural constant e as the base; sigmoid represents the sigmoid function for mapping data to the interval (0, 1).

[0163] In an embodiment of the present application, the processing module 12 is further configured to obtain a fourth data segment based on the plurality of data segments; the fourth data segment is an arbitrary data segment that has not been subjected to vibration anomaly degree determination among the plurality of data segments;

[0164] and, based on the fourth data segment, obtaining a first similarity and a second similarity; the first similarity is a similarity between the fourth data segment and a previous data segment; the second similarity is a similarity between two previous data segments of the fourth data segment;

[0165] and, based on the fourth data segment, obtaining a number of data segments with a similarity greater than a first threshold value to the fourth data segment;

[0166] and, based on the first similarity, the second similarity and the number, obtaining a vibration abnormality degree of the fourth data segment.

[0167] In an embodiment of the present application, the processing module 12 is further configured to obtain first vibration data based on the target vibration data; the first vibration data is any vibration data in the target vibration data.

[0168] and, based on the first vibration data, obtain a vibration abnormality degree of a data segment corresponding to the first vibration data.

[0169] and, based on the initial significant value and the vibration abnormality degree, obtain a corrected significant value.

[0170] It should be noted that the multi-axis motor drive fault real-time monitoring and early warning system proposed in the present application is based on the difference between the vibration data caused by the motor load change and the motor drive fault, and by analyzing the above difference, the influence of interference factors is excluded, the vibration abnormality degree of the monitoring data is obtained, so that the significant value is corrected, so that the abnormal state caused by the motor drive fault is extracted and early warning is performed; the abnormal state caused by the motor load change is not extracted. That is, the multi-axis motor drive fault real-time monitoring and early warning system proposed in the present application can not only avoid false detection, but also avoid missed detection

[0171] In the above embodiments, the description of each embodiment has its own emphasis, and the parts not described in detail in a certain embodiment can be referred to the related description of other embodiments.

[0172] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described method, device and equipment can refer to the corresponding process in the foregoing method embodiment, which will not be repeated here.

[0173] In several embodiments provided in the embodiments of the present application, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other manners. For example, the embodiments of the apparatus described above are merely schematic. For example, the division of the modules is merely a logical function division. There can be another division manner for the actual implementation. For example, a plurality of modules or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections can be indirect couplings or communication connections through some interfaces, devices or modules, and can be in electrical, mechanical or other forms.

[0174] The modules described as separated components can or can not be physically separated, and the components displayed as modules can or can not be physical modules, i.e., can be located in one place, or can be distributed on a plurality of network modules. Part or all of the modules can be selected according to actual needs to achieve the purpose of the embodiments.

[0175] In addition, each functional module in each of the embodiments of the present application can be integrated in one processing module, or each module can be physically present alone, or two or more modules can be integrated in one module. The integrated module can be realized in the form of hardware or in the form of a software functional module. When the integrated module is realized in the form of a software functional module and sold or used as an independent product, it can be stored in a computer readable storage medium.

[0176] In the above embodiments, all or part of the embodiments can be realized by software, hardware, firmware or any combination thereof. When realized by software, all or part of the embodiments can be realized in the form of a computer program product.

[0177] The computer program product includes one or more computer instructions. When the computer program is loaded and executed on a computer, all or part of the processes or functions described in the embodiments of the present application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center through wired (such as coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can store or a data storage device such as a server, data center, etc. integrated with one or more available media. The available media can be a magnetic medium (for example, a floppy disk, a hard disk, a magnetic tape), an optical medium (for example, a DVD), or a semiconductor medium (for example, a solid state disk (SSD)), etc.

[0178] Although the embodiments of the present application have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to the embodiments without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A real-time monitoring and early warning system for multi-axis motor drive faults, characterized in that, include: The acquisition module is used to acquire target vibration data of the multi-axis motor; The processing module is configured to acquire data anomaly features based on the target vibration data, acquire an initial significance value based on the data anomaly features, and correct the initial significance value based on feature data of the vibration cause of the multi-axis motor to acquire a corrected significance value; the data anomaly features include at least feature data characterizing the vibration cause of the multi-axis motor; the data anomaly features are determined based on the similarity between a second data segment and a third data segment, the similarity being determined based on the difference between corresponding vibration data in the second data segment and the third data segment, the fluctuation degree of the second data segment, and the fluctuation degree of the third data segment, wherein the second data segment and the third data segment are two data segments among multiple data segments obtained by segmenting the target vibration data, and the similarity is used to reflect the cause of the abnormal vibration of the multi-axis motor, the cause being a change in motor load or a motor malfunction; The control module is used to issue anomaly alarms based on the corrected significant value.

2. The multi-axis motor drive fault real-time monitoring and early warning system according to claim 1, characterized in that, The method for acquiring vibration data of a multi-axis motor includes: The current vibration data of the multi-axis motor is obtained at frequencies that are integer multiples of the highest vibration frequency; In addition, the current vibration data is transmitted to the processing module so that the processing module can preprocess the current vibration data to obtain the target vibration data.

3. The multi-axis motor drive fault real-time monitoring and early warning system according to claim 1, characterized in that, The method for obtaining data anomaly characteristics based on the target vibration data includes: The target vibration data is segmented to obtain multiple data segments; Based on multiple data segments, the fluctuation level of each data segment is obtained; Based on the degree of fluctuation in each data segment, the abnormal characteristics of the data are obtained.

4. The multi-axis motor drive fault real-time monitoring and early warning system according to claim 3, characterized in that, The method for segmenting the target vibration data to obtain multiple data segments includes: The target vibration data is converted from time-domain data to frequency-domain data; The target vibration data is segmented according to the period of the frequency domain data to obtain multiple data segments.

5. The multi-axis motor drive fault real-time monitoring and early warning system according to claim 4, characterized in that, The method for obtaining the fluctuation level of each data segment based on multiple data segments includes: Based on the multiple data segments, a first data segment is obtained; the first data segment is any data segment among the multiple data segments for which the fluctuation level has not been obtained; Based on the first data segment, fluctuation parameters are obtained; the fluctuation parameters include at least the fluctuation range, dispersion, and number of fluctuations of the first data segment. Based on the fluctuation parameters of the first data segment, the fluctuation degree of the first data segment is obtained.

6. The multi-axis motor drive fault real-time monitoring and early warning system according to claim 5, characterized in that, The formula for calculating the degree of fluctuation of the first data segment based on the fluctuation parameters of the first data segment is as follows: ； in, This indicates the degree of fluctuation in the i-th data segment within each data segment; This represents the number of vibration data points in the i-th data segment; Indicates the i-th segment of data. One vibration data value; This represents the mean value of all vibration data in the i-th data segment; This represents the maximum value of the vibration data in the i-th data segment; This represents the minimum value of the vibration data in the i-th data segment; This represents the mean of the fluctuation factor in the time domain analysis of the i-th data segment, which is also the mean of the ratio of the peak value to the root mean square value of the vibration data.

7. The multi-axis motor drive fault real-time monitoring and early warning system according to claim 1, characterized in that, The formula for calculating the similarity is as follows: ； in, Indicates the first Segment data segment and the first The similarity of the degree of fluctuation between data segments; This indicates the number of vibration data points in each data segment; Indicates the first The first data segment One vibration data value; Indicates the first The first data segment One vibration data value; Indicates the first The degree of fluctuation in the data segment; Indicates the first The degree of fluctuation of a data segment; exp represents an exponential function with the natural constant e as the base; sigmoid represents the sigmoid function that maps data to the interval (0, 1).

8. The multi-axis motor drive fault real-time monitoring and early warning system according to claim 3, characterized in that, The data anomaly characteristics include the degree of vibration anomaly in each data point; The method for obtaining the data anomaly characteristics based on the fluctuation level of each data segment includes: Based on the multiple data segments, a fourth data segment is obtained; the fourth data segment is any data segment from the multiple data segments for which the degree of vibration abnormality was not obtained; Based on the fourth data segment, a first similarity and a second similarity are obtained; the first similarity is the similarity between the fourth data segment and the previous data segment; the second similarity is the similarity between the first two data segments of the fourth data segment. Based on the fourth data segment, the number of data segments whose similarity to the fourth data segment is greater than the first threshold is obtained; Based on the first similarity, the second similarity, and the quantity, the degree of vibration anomaly of the fourth data segment is obtained.

9. The multi-axis motor drive fault real-time monitoring and early warning system according to claim 8, characterized in that, The method for correcting the initial significant value based on the feature data of the vibration cause of the multi-axis motor, and obtaining the corrected significant value, includes: Based on the target vibration data, first vibration data is obtained; the first vibration data is any vibration data in the target vibration data. Based on the first vibration data, the degree of vibration anomaly in the corresponding data segment of the first vibration data is obtained; Based on the initial significance value and the degree of vibration anomaly, a corrected significance value is obtained.

Images