Industrial design cloud platform system based on artificial intelligence

Abstract

The invention relates to the technical field of data processing, in particular to an artificial intelligence-based industrial design cloud platform system, which comprises an acquisition module used for acquiring vibration monitoring data and operation parameters of a target part of target equipment within a monitoring duration; the decomposition module is used for performing independent component decomposition on the vibration monitoring data to obtain a plurality of vibration component and signal component sequences; the determination module is used for determining the average anomaly degree of the vibration monitoring data according to the amplitudes of the data points of the signal component sequences of the vibration components and the signal component sequences; the determining module is used for determining the vibration deviation degree of the vibration monitoring data of the target part according to the operation parameters; the determination module is also used for determining the processing demand degree of the vibration monitoring data according to the average abnormity degree and the vibration deviation degree; and the processing module is used for smoothing the vibration monitoring data based on the processing demand degree so as to remove noise interference in the vibration monitoring data. And the accuracy and the credibility of the monitoring data are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and particularly to an industrial design cloud platform system based on artificial intelligence. Background Art

[0002] The industrial design cloud platform system combines the powerful capabilities of the cloud with the complex requirements of industrial design, aiming to support the whole-process management and optimization of industrial products from concept to production. The industrial design cloud platform system usually includes key features such as integrated design tools, collaboration and communication, big data and analysis, security and reliability.

[0003] In some scenarios, in the industrial design cloud platform system based on artificial intelligence, the industrial design cloud platform system usually needs to integrate monitoring data from multiple different data sources, including various data types from each data source such as sensors, production equipment, and product testing. The monitoring data collected in real time from each data source vary in data quality. Some monitoring data have high data quality, but some have low data quality. The monitoring data with low quality may be due to large noise interference or abnormal data generated due to the equipment itself. Among them, the real abnormal data caused by the equipment itself can be retained for subsequent monitoring and prevention. The abnormal data caused by large noise interference will lead to low accuracy and credibility of the monitoring data, which is not conducive to the whole-process management and optimization of industrial products from concept to production. Summary of the Invention

[0004] In order to solve the technical problem of low accuracy and credibility of the monitoring data, the purpose of the present invention is to provide an industrial design cloud platform system based on artificial intelligence, and the specific technical solution adopted is as follows: An embodiment of the present application provides an industrial design cloud platform system based on artificial intelligence, including: an acquisition module, configured to acquire vibration monitoring data of a target part of a target device within a monitoring duration and operating parameters of the target part of the target device during operation; a decomposition module, configured to perform independent component analysis on the vibration monitoring data to obtain a plurality of vibration components and corresponding signal component sequences for each vibration component; a determination module, configured to determine an average abnormality degree of the vibration monitoring data according to the amplitudes of the data points in the signal component sequences of the vibration components and the signal component sequences, where the average abnormality degree indicates the authenticity of the vibration monitoring data, and the average abnormality degree is inversely proportional to the authenticity; a determination module, configured to determine a vibration deviation degree of the vibration monitoring data of the target part according to the operating parameters, where the vibration deviation degree indicates the possibility that the abnormal data points in the vibration monitoring data are caused by the target device, and the vibration deviation degree is directly proportional to the possibility that the abnormal data points in the vibration monitoring data are caused by the target device; the determination module is further configured to determine a processing requirement degree of the vibration monitoring data according to the average abnormality degree and the vibration deviation degree, where the processing requirement degree indicates the possibility that the abnormal data points in the vibration monitoring data are caused by noise interference, and the processing requirement degree is directly proportional to the possibility that the abnormal data points in the vibration monitoring data are caused by noise interference; a processing module, configured to perform smoothing processing on the vibration monitoring data based on the processing requirement degree to remove noise interference in the vibration monitoring data.

[0005] Optionally, the determination module is further configured to: determine the number of types of data points in the signal component sequence of the vibration component and the total number of data points of all data points, where data points with equal values are of the same type; determine the local complexity of the vibration component among all vibration components according to the distribution probability values, the number of types, and the total number of data points of each type of data points in the signal component sequence of the vibration component, where the local complexity indicates the periodicity of the vibration component, and the local complexity is inversely proportional to the periodicity of the vibration component; determine the vibration abnormality degree of each data point in the signal component sequence of the vibration component according to the local complexity and the amplitude of the data points in the signal component sequence, where the vibration abnormality degree indicates the possibility of the data point being abnormal, and the vibration abnormality degree is directly proportional to the possibility of the data point being abnormal; determine the average abnormality degree of the vibration monitoring data according to the vibration abnormality degrees of the data points of each vibration component.

[0006] Optionally, the determination module is further configured to: determine the absolute value of the difference between the distribution probability values of any two types of data points in the signal component sequence of the vibration component to obtain m absolute values of differences; add up the m absolute values of differences and divide by m to obtain a first ratio, where the first ratio indicates the number of similar data points in the vibration monitoring data, and the first ratio is inversely proportional to the number of similar data points in the vibration monitoring data, and m is an integer greater than 0; determine a second ratio between the number of types and the total number of data points in the signal component sequence of the vibration component; determine the product of the first ratio and the second ratio as the local complexity.

[0007] Optionally, the determining module is further configured to: determine the average amplitude of the amplitudes of all data points in the signal component sequence of the vibration component, and the first difference between the maximum amplitude and the minimum amplitude of the data points in the signal component sequence; determine the absolute value of the second difference between the amplitude of the data point and the average amplitude, and the absolute value of the third difference between the amplitude of the data point and the first difference; determine the third ratio of the absolute value of the second difference to the absolute value of the third difference; perform a functional operation on the local complexity using the hyperbolic tangent function to obtain an operation result; determine the product of the third ratio and the operation result, and use the product of the third ratio and the operation result as the vibration abnormality degree of the data points in the signal component sequence.

[0008] Optionally, the determining module is further configured to: determine the feed rate of the target part determined from the operating parameters during the monitoring duration; determine the error between the operating parameters and the vibration monitoring data; determine the parameter change complexity of the operating parameters according to the error and the feed rate, where the parameter change complexity indicates the possibility that the abnormal data points in the vibration monitoring data are caused by the operating parameters of the target part, and the parameter change complexity is directly proportional to the possibility that the abnormal data points in the vibration monitoring data are caused by the operating parameters of the target part; determine the load on the target part during the monitoring duration and the displacement generated by the target part under the load; determine the rigidity of the target part according to the load and the displacement, where the rigidity is inversely proportional to the probability of abnormal data points in the vibration monitoring data; perform a functional operation on the product of the rigidity and the parameter change complexity using the hyperbolic tangent function to obtain the vibration deviation degree.

[0009] Optionally, the determining module is further configured to: fit the operating parameters of the target part according to the monitoring duration of the vibration monitoring data to obtain a fitting curve; determine the mean square error between the fitting curve and the vibration monitoring curve of the vibration monitoring data as the error; determine the degree of fluctuation of the feed rate of the target part during the monitoring duration; determine the parameter change complexity of the operating parameters according to the mean square error and the degree of fluctuation.

[0010] Optionally, the determining module is further configured to: determine the ratio of the degree of fluctuation to the mean square error as the parameter change complexity.

[0011] Optionally, the load includes a radial load and an axial load, the displacement includes an axial displacement and a radial displacement, and the obtaining module is further configured to: obtain the radial load and the axial load of the target part during the monitoring duration, and the axial displacement of the target part in the axial direction and the radial displacement of the target part in the radial direction; the determining module is further configured to: determine the ratio of the degree of fluctuation to the mean square error as the parameter change complexity. Determine the fourth ratio between the axial load and the axial displacement, and the fifth ratio between the radial load and the radial displacement; determine the product of the fourth ratio and the fifth ratio as the rigidity of the target part.

[0012] Optionally, the determination module is further configured to: determine the sum of the square of the average abnormality degree and the square of the reciprocal of the vibration deviation degree; perform a square root operation on the sum value to obtain an operation result, and use the operation result as the processing requirement degree.

[0013] The processing module is further configured to: use the processing requirement degree as the standard deviation followed by the filtering kernel of the Gaussian filter to obtain a target Gaussian filter; perform smoothing processing on the vibration monitoring data by using the target Gaussian filter.

[0014] The present invention has the following beneficial effects: The acquisition module acquires the vibration monitoring data of the target part of the target device during the monitoring duration and the operating parameters of the target part of the target device during operation. Then, the decomposition module performs independent component decomposition on the vibration monitoring data to obtain a plurality of vibration components and corresponding signal component sequences. Next, the determination module determines the average abnormality degree of the vibration monitoring data according to the amplitudes of the data points of the signal component sequences of the respective vibration components and the signal component sequences, and determines the vibration deviation degree of the vibration monitoring data of the target part according to the operating parameters. The vibration deviation degree can reflect whether the abnormal data points in the vibration monitoring data are caused by the device, so that the possible faults of the target device in the production process can be predicted, and timely processing can be carried out, improving the production efficiency. Finally, the determination module determines the processing requirement degree of the vibration monitoring data according to the average abnormality degree and the vibration deviation degree. The processing requirement degree can reflect whether the abnormal data points in the vibration monitoring data are caused by noise interference, so that the processing module can perform smoothing processing on the vibration monitoring data based on the processing requirement degree to remove the noise interference in the vibration monitoring data, improving the accuracy and reliability of the monitoring data, and facilitating the whole-process management and optimization of industrial products from concept to production. Description of the Drawings

[0015] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0016] Figure 1 Schematic diagram of the module composition of an industrial design cloud platform system based on artificial intelligence provided by an embodiment of the present invention; Figure 2 Schematic diagram of a fitting curve and a curve of the intensity value of vibration monitoring data provided by an embodiment of the present invention; Figure 3 Schematic diagram of a radial load and an axial load acting on a main shaft provided by an embodiment of the present invention. Detailed implementation manners

[0017] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following combines the accompanying drawings and preferred embodiments to detail the specific implementation manners, structures, features and effects of an industrial design cloud platform system based on artificial intelligence proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

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

[0019] The following specifically describes the specific solution of an industrial design cloud platform system based on artificial intelligence provided by the present invention with reference to the accompanying drawings.

[0020] Please refer to Figure 1 , which shows a schematic diagram of the module composition of an industrial design cloud platform system based on artificial intelligence provided by an embodiment of the present invention. The industrial design cloud platform system 100 based on artificial intelligence includes: an acquisition module 101, a decomposition module 102, a determination module 103, and a processing module 104.

[0021] The acquisition module 101 is used to acquire vibration monitoring data of a target part of a target device within a monitoring duration and operating parameters of the target part of the target device during operation.

[0022] Specifically, the target device can be a production line device, an industrial robot, etc., such as a numerically controlled machine tool. The target part can be a key part of the target device. The vibration monitoring data can be collected after installing vibration sensors at the target part. For example, the key parts of a numerically controlled machine tool can be the spindle, spindle box, workbench, guide rail, etc. Vibration sensors can be installed at these key parts to collect vibration data at each position. The acquisition module 101 uses a wireless or wired network to acquire the vibration monitoring data collected by the vibration sensors, such as Ethernet, Wireless Fidelity (WIFI), etc. The monitoring duration can be set according to actual needs, and the embodiments of the present application do not limit this here.

[0023] Vibration monitoring data includes, but is not limited to, data such as acceleration, velocity, displacement, phase, etc. The vibration monitoring data of the target device can reflect the operating state and health condition of the device. By real-time monitoring of vibration data, abnormal vibration patterns of the target device can be identified, potential mechanical failures or component wear can be predicted, and maintenance and repairs can be carried out in a timely manner to avoid production interruptions and additional costs caused by sudden equipment shutdowns. In the vibration monitoring data of the target device, since the vibration monitoring data may contain some situations deviating from the normal vibration pattern, and such situations may be caused by parameter settings during the operation of the target device or the device itself, or may be caused by a large amount of interference in the vibration monitoring data. Therefore, it is necessary to conduct a detailed analysis of the reasons for the abnormalities in the vibration monitoring data.

[0024] The operating parameters of the target part of the target device during operation include, but are not limited to, data such as rotational speed, power, feed rate, spindle magnification, temperature, humidity, load, and displacement under load. In the industrial design cloud platform system based on artificial intelligence, during the real-time monitoring of the vibration condition of the target device and the processing of vibration monitoring data, the "abnormal" phenomenon different from the normal vibration pattern may also be an "abnormal" behavior caused by parameter changes during the operation of the device or the operation duration of the device, and this abnormal behavior is inevitable, that is, the allowable error during operation, so as to better distinguish the real noise interference situation. Therefore, the operating parameters of the target device can be used to distinguish whether the abnormality in the vibration monitoring data is due to the device itself or the existence of interference noise.

[0025] The decomposition module 102 is used to perform independent component decomposition on the vibration monitoring data to obtain multiple vibration components and the signal component sequences corresponding to each vibration component.

[0026] Specifically, the decomposition module 102 can use a decomposition algorithm to decompose the vibration monitoring data. There are certain decomposition rules in the process of decomposing signal data by the decomposition algorithm, that is, due to the existence of different types of vibration components in the vibration monitoring data, it will be decomposed into different vibration components, such as fundamental frequency components, harmonic components, sub-harmonic components, natural frequencies, etc. The signal component sequence corresponding to each vibration component includes multiple data points.

[0027] The determination module 103 is used to determine the average abnormality degree of the vibration monitoring data according to the amplitudes of the data points of the signal component sequences of each vibration component and the signal component sequences. The average abnormality degree indicates the authenticity of the vibration monitoring data, and the average abnormality degree is inversely proportional to the authenticity.

[0028] Furthermore, the authenticity indicates the number of data points with abnormalities in the vibration monitoring data, and the authenticity is inversely proportional to the number of data points with abnormalities in the vibration monitoring data.

[0029] Specifically, there are multiple data points in the signal component sequence, and each data point corresponds to a specific amplitude (value). The data points with equal amplitudes in the multiple data points are regarded as one type of data points. For example, in a certain signal component sequence of vibration monitoring data, the data points with a value of 5 are one type of data points, and the data points with a value of 2 are one type of data points. Among them, the local complexity of the vibration component can be calculated from the signal component sequence of each vibration component, and then the vibration abnormality of each data point in each vibration component can be determined according to the local complexity. Finally, the average abnormality of the vibration monitoring data can be determined according to the local complexity and the vibration abnormality.

[0030] Among them, the process for the determination module 103 to determine the local complexity is as follows: Determine the number of types of data points and the total number of all data points in the signal component sequence of the vibration component, where the data points with equal values are of the same type; Determine the local complexity of the vibration component among all vibration components according to the distribution probability values, the number of types, and the total number of data points of each type of data points in the signal component sequence of the vibration component. The local complexity indicates the periodicity of the vibration component, and the local complexity is inversely proportional to the periodicity of the vibration component.

[0031] Specifically, the amplitudes of the data points in the signal component sequence may have multiple different values, and the number of types of data points can be determined from the number of different values. For example, if there are multiple data points with values of 1, 2, 4, and 6 in the signal component sequence, then the number of types of data points is 4. The total number of data points is the number of all data points in the signal component sequence. The greater the local complexity, the worse the periodicity of the vibration component, and there are fewer equal data points among all the data points in the signal component sequence of the vibration component, indicating that the local complexity of the data points of this vibration component among all vibration components is relatively high. This local complexity provides data support for subsequent evaluation of abnormalities in the monitoring data. The distribution probability value refers to the probability of the data in the signal component sequence of the vibration component in a certain interval or a specific value. The content related to the distribution probability value in the prior art also falls within the scope protected by the embodiments of the present application, and they can be referred to each other. The embodiments of the present application will not elaborate herein.

[0032] When the determination module 103 determines the local complexity, the process is as follows: Determine the absolute value of the difference between the distribution probability values of any two types of data points in the signal component sequence of the vibration component, and obtain m absolute values of differences; After superimposing the m absolute values of differences and dividing by m, obtain a first ratio, which indicates the number of similar data points in the vibration monitoring data. m is an integer greater than 0, and the first ratio is inversely proportional to the number of similar data points in the vibration monitoring data; Determine the second ratio between the number of types and the total number of data points in the signal component sequence of the vibration component; Determine the product of the first ratio and the second ratio as the local complexity.

[0033] Specifically, the local complexity can be expressed by the following formula: In the formula, represents the local complexity, represents the distribution probability value of the i-th type of data point in the signal component sequence of the vibration component of the vibration monitoring data of the target device (data points with equal values are recorded as one type of data point); represents the distribution probability value of the j-th type of data point in the signal component sequence of the vibration component of the vibration monitoring data of the target device; there is a set of differences between the distribution probability values of any two types of data points, and there are a total of m groups of differences; g represents the number of types of data points in the signal component sequence of the vibration monitoring data; G represents the total number of all data points.

[0034] represents the average value of the differences between the distribution probability values of various types of data points in the signal component sequence of the vibration component of the vibration monitoring data. The smaller the average value, the more similar data points exist in the vibration component of the vibration monitoring data, that is, it indicates that there are strong periodic characteristics in the data corresponding to the vibration component of the vibration monitoring data of the target device; The smaller the ratio of, the more it indicates that the number of types of data points is less than the total number of data points, that is, it indicates that there must be many data points with equal values among the G data points. represents the local complexity of the data of the vibration component of the vibration monitoring data of the target device among all vibration components. The greater the local complexity, the worse the periodicity of the vibration component of the vibration monitoring data of the target device, and there are fewer equal data points among all data points of the vibration component of the vibration monitoring data of the target device, indicating that the local complexity of the vibration component of the vibration monitoring data of the target device is relatively high, and this local complexity provides data support for subsequent evaluation of anomalies in the monitoring data.

[0035] The process by which the determination module 103 determines the vibration anomaly degree is as follows: Determine the vibration anomaly degree of each data point in the signal component sequence of the vibration component according to the local complexity and the amplitude of the data points in the signal component sequence. The vibration anomaly degree indicates the possibility of a data point appearing abnormal, and the vibration anomaly degree is directly proportional to the possibility of a data point appearing abnormal.

[0036] Specifically, during the independent component analysis of vibration monitoring data, under normal circumstances, there should be no local abnormal behavior in the data points of the signal component sequence of a single vibration component, that is, in this vibration component, the average level of the data points should be relatively stable; if there are relatively "prominent" data points in the signal component sequence of the vibration component, they can be considered relatively abnormal data points. The prominent data points can be determined by the vibration abnormality degree. The greater the vibration abnormality degree of the data point, the greater the possibility that the data point is abnormal.

[0037] More specifically, the more specific process for the determination module 103 to determine the vibration abnormality degree is as follows: Determine the average amplitude of the amplitudes of all data points in the signal component sequence of the vibration component, and the first difference between the maximum amplitude and the minimum amplitude of the data points in the signal component sequence; Determine the absolute value of the second difference between the amplitude of the data point and the average amplitude, and the absolute value of the third difference between the amplitude of the data point and the first difference; Determine the third ratio of the absolute value of the second difference to the absolute value of the third difference; Perform a functional operation on the local complexity using the hyperbolic tangent function to obtain the operation result; Determine the product of the third ratio and the operation result, and use the product of the third ratio and the operation result as the vibration abnormality degree of the data point in the signal component sequence.

[0038] Among them, the vibration abnormality degree can be calculated by the following formula: In the formula, represents the vibration abnormality degree, represents the local complexity of the u-th vibration component, represents the amplitude of the e-th data point in the vibration component of the vibration monitoring data of the target device; represents the average amplitude of the amplitudes of all data points in the signal component sequence; represents the first difference (range) between the maximum amplitude and the minimum amplitude of the data points in the signal component sequence, represents the hyperbolic tangent function.

[0039] Among them, represents the initial vibration abnormality degree of a single data point. The larger the ratio, the greater the difference between the data point and the average level of the data of the vibration component of the vibration monitoring data of the target device, and the closer it is to the first difference, that is, the more obvious the "abnormal" behavior of the data point in the distribution. Adjusts the "initial vibration abnormality degree" of a single data point as the overall weight of this vibration component; The larger the product of

[0040] Furthermore, in order to avoid Regarding the problem that the vibration abnormality degree cannot be calculated due to being zero, the absolute value of the third difference can be added to a constant and then compared with the absolute value of the second difference, and a non-zero value can be taken according to the actual situation. The embodiments of the present application do not limit this here. That is, the calculation formula of the vibration abnormality degree can also adopt the following formula according to the actual situation: In the formula, represents the vibration abnormality degree, represents the local complexity of the u-th vibration component, represents the amplitude of the e-th data point in the vibration component of the vibration monitoring data of the target device; represents the average amplitude of the amplitudes of all data points in the signal component sequence; represents the first difference (range) between the maximum amplitude and the minimum amplitude of the data points in the signal component sequence, represents the hyperbolic tangent function, is a constant, and the value of k is as small as possible. As an embodiment of the present invention, for example, the value of k can be set to 0.1.

[0041] Among them, represents the initial vibration abnormality degree of a single data point. The larger the ratio, the greater the difference between the data point and the average level of the data of the vibration component of the vibration monitoring data of the target device, and the closer it is to the first difference, that is, the more obvious the "abnormal" behavior of the data point in the distribution. Adjusts the "initial vibration abnormality degree" of a single data point as the overall weight of this vibration component; The larger the product of, the greater the vibration abnormality degree of the data point. This value mainly provides data support for the subsequent screening of abnormal data points.

[0042] By obtaining the vibration abnormality degree of a single data point in a single vibration component as described above, then arranging the vibration abnormality degrees of all data points in this vibration component in ascending order, then obtaining the difference between adjacent two vibration abnormality degrees, traversing all the differences to find the first node with the largest difference, and then dividing the sorted sequence into two parts. The data points on the right side of the largest difference are considered as data points with large vibration abnormality degrees. Or directly select the vibration abnormality degree corresponding to the maximum value as the data point with a large vibration abnormality degree. Based on the abnormality degrees of these data points, the abnormal components of the vibration monitoring data in the industrial design cloud platform system based on artificial intelligence can be evaluated.

[0043] Finally, the process of the determination module 103 for determining the average abnormality degree is as follows: determining the average abnormality degree of the vibration monitoring data according to the vibration abnormality degrees of the data points of each vibration component.

[0044] Specifically, in an industrial design cloud platform system based on artificial intelligence, vibration monitoring data of various devices and environments is usually integrated, and the collected vibration monitoring data is effectively processed. The interference degree in the vibration monitoring data will submerge the expression ability of the real data, that is, reduce the authenticity of the vibration monitoring data of the system. Among them, the average abnormality can indicate the authenticity of the vibration monitoring data, and it can be calculated by the following formula: In the above formula, represents the average abnormality. Assuming that E data points are selected from the vibration components of the vibration monitoring data of a single target device, represents the vibration abnormality of the e-th abnormal data point in the signal component sequence of the u-th vibration component of the vibration monitoring data. The vibration monitoring data has R vibration components.

[0045] Among them, represents the average abnormality of the vibration monitoring data of all vibration components of the vibration monitoring data of the target device. The larger the average abnormality, the more "abnormal" data points and the higher the vibration abnormality in the vibration monitoring data; it reflects from the side that the data authenticity of the vibration monitoring data is worse. Subsequently, it is necessary to verify and adjust the data abnormality in the vibration monitoring data from the target device itself.

[0046] The determination module 103 is configured to determine the vibration deviation degree of the vibration monitoring data of the target part according to the operating parameters.

[0047] The vibration deviation degree indicates the possibility that the abnormal data points in the vibration monitoring data are caused by the target device, and the vibration deviation degree is directly proportional to the possibility that the abnormal data points in the vibration monitoring data are caused by the target device.

[0048] Specifically, in the industrial design cloud platform system based on artificial intelligence, during the process of processing the vibration monitoring data detected from the target device, the "abnormal" phenomenon different from the normal vibration mode of the target device may also be an "abnormal" behavior caused by the change of operating parameters or the running time of the device during operation. This abnormal behavior is caused by the device itself and is inevitable, that is, the allowable error during operation. Another situation is that there are interference components in the vibration monitoring data, which affect the evaluation of the target device. By determining the real cause of the abnormal data points in the vibration monitoring data through the operating parameters, if it is caused by interference noise, the interference components can be purified to more accurately screen out the interference components. Therefore, the vibration deviation degree of the target part of the target device can be determined through the operating parameters, so as to determine the real cause of the abnormal data points that cause the vibration monitoring data, and then process it specifically to improve the accuracy and effectiveness of the vibration monitoring data.

[0049] Among them, the process by which the determination module 103 determines the vibration deviation degree is as follows: First, determine the parameter change complexity of the operating parameters; second, determine the rigidity of the target part of the target device; finally, determine the vibration deviation degree according to the rigidity and the parameter change complexity.

[0050] First, the process by which the determination module 103 determines the parameter change complexity is specifically as follows: Determine the feed rate of the target part within the monitoring duration determined from the operating parameters; determine the error between the operating parameters and the vibration monitoring data; determine the parameter change complexity of the operating parameters according to the error and the feed rate. The parameter change complexity indicates the possibility that the abnormal data points in the vibration monitoring data are caused by the operating parameters of the target part, and the parameter change complexity is proportional to the possibility that the abnormal data points in the vibration monitoring data are caused by the operating parameters of the target part.

[0051] Specifically, taking the target device as a numerically controlled machine tool as an example to illustrate the embodiments of the present application, the target part can be the main shaft of the numerically controlled machine tool, and the operating parameters can be the feed rate of the main shaft and the rotational speed of the main shaft. The error between the operating parameters and the vibration monitoring data can be the mean square error. The parameter change complexity indicates the possibility that the cause of the abnormal data points in the vibration monitoring data is the change of the operating parameters of the target part. The greater the parameter change complexity, the more it shows that the abnormal cause of the abnormal data points in the vibration monitoring data is mainly caused by the change of the operating parameters, rather than the external noise interference during the data acquisition process. In this way, the real cause of the abnormality of the abnormal data points in the vibration monitoring data can be distinguished, and then processed specifically, improving the data processing efficiency and ensuring the reliability of the vibration monitoring data.

[0052] When determining the error, the determining module 103 fits the operating parameters of the target part according to the monitoring duration of the vibration monitoring data to obtain a fitting curve; and determines the mean square error between the fitting curve and the vibration monitoring curve of the vibration monitoring data as the error. As Figure 2 shown, the determining module 103 performs least squares fitting on the rotational speed of the main shaft according to the monitoring duration of the vibration monitoring data to obtain a rotational speed curve; performs least squares fitting on the vibration monitoring data of the main shaft to obtain a vibration curve; normalizes the ordinates of the rotational speed curve and the vibration monitoring curve respectively to ensure the comparability of the two types of monitoring data in one coordinate system, and then keeps the time points of the abscissa unchanged, thereby obtaining a new coordinate system. In the new coordinate system, the abscissa represents time, and the ordinate represents the normalized data of the rotational speed data and the normalized data of the vibration monitoring data, which can be understood as the intensity values of the rotational speed and the vibration monitoring data. The obtained curve is as Figure 2 shown, Figure 2 in which, is the fitting curve of the normalized value of the rotational speed of the main shaft, that is, p is the fitting curve of the intensity value of the rotational speed, is the curve of the normalized value of the vibration monitoring data, that is, q is the fitting curve of the intensity value of the vibration monitoring data; Figure 2 The abscissa of is time, and the ordinate is the intensity values of the rotational speed data and the vibration monitoring data, where the value range of the ordinate is 0 - 1. Then determine the mean square error between the fitting curve and the vibration monitoring curve as the error between the rotational speed of the main shaft and the vibration monitoring curve. It should be noted that, Figure 2 the ordinate in is the normalized value, and its value range is between 0 - 1, without unit; Figure 2 the unit of the abscissa in is second (s). Mapping the vibration monitoring data and the rotational speed data to one coordinate system is to determine whether there is a changing response relationship between the data in two dimensions. Because the mean square error needs to be calculated, it is necessary to ensure that the subtracted data are in the same dimension. Therefore, the two types of data are normalized and then fitted into the same coordinate.

[0053] When determining the complexity of the parameter change, the determining module 103 determines the fluctuation degree of the feed rate of the target part within the monitoring duration; determines the complexity of the parameter change of the operating parameter according to the mean square error and the fluctuation degree. Specifically, the determining module 103 determines the ratio of the fluctuation degree to the mean square error as the complexity of the parameter change.

[0054] Specifically speaking, the complexity of the parameter change can be calculated by the following formula: In the formula, represents the complexity of the parameter change, represents the mean square error between the fitting curve of the rotational speed of the target device (such as a numerically controlled machine tool) in the cloud platform system and the curve of the vibration monitoring data of the main shaft; It represents the degree of fluctuation of the feed rate of the spindle during the monitoring duration.

[0055] Among them, It represents the complexity of the parameter changes of the target device being monitored during the data monitoring duration of the cloud platform system. The larger the product, the higher the responsiveness of the change in the spindle speed and the vibration monitoring data and the greater the fluctuation of the feed rate, indicating that the reason for the "abnormal" vibration of the target device during this monitoring process may be due to the change in the feed rate of the spindle or the change in the spindle speed, rather than the abnormality caused by external interference during the data acquisition process.

[0056] Secondly, the process by which the determination module 103 determines rigidity is as follows: Determine the load on the target part during the monitoring duration and the displacement generated by the target part under the load; Determine the rigidity of the target part based on the load and displacement. The rigidity is inversely proportional to the probability of data points in the vibration monitoring data appearing abnormally.

[0057] Specifically, in the industrial design cloud platform system based on artificial intelligence, during the vibration monitoring of a numerically controlled machine tool, the abnormality of the vibration monitoring data of the spindle may be due to the axial and radial rigidity of the spindle during the machining process. If the spindle deforms or is affected externally during operation, resulting in insufficient rigidity, it will cause the vibration mode to deviate from the normal state, increasing the probability of data points in the vibration monitoring data appearing abnormally. Therefore, the probability of data points in the vibration monitoring data appearing abnormally can be determined through rigidity. The greater the rigidity, the smaller the probability of data points in the vibration monitoring data appearing abnormally due to rigidity, and the greater the probability of data points in the vibration monitoring data appearing abnormally due to external interference noise; The smaller the rigidity, the greater the probability of data points in the vibration monitoring data appearing abnormally due to rigidity, and the smaller the probability of data points in the vibration monitoring data appearing abnormally due to external interference noise. The load on the spindle during the monitoring duration includes radial load and axial load, and the displacement generated by the spindle under the action of the load includes axial displacement and radial displacement.

[0058] First, install electromagnetic induction sensors at the radial position and the axis position of the spindle to detect the radial displacement and axial displacement generated when the spindle is stressed radially, as Figure 3 shown. The forces (loads) applied to the axial position and radial position of the spindle can be directly measured using force sensors to obtain the axial load and radial load.

[0059] When the determination module 103 determines the load on the target part during the monitoring period and the displacement of the target part under the load, the acquisition module 101 is first required to obtain the radial load and axial load of the target part during the monitoring period, and the axial displacement of the target part in the axial direction and the radial displacement of the target part in the radial direction. Specifically, the acquisition module 101 can obtain the radial load, axial load, axial displacement and radial displacement from the electromagnetic induction sensor and the force sensor. Then the determination module 103 determines the fourth ratio between the axial load and the axial displacement, and the fifth ratio between the radial load and the radial displacement; the product of the fourth ratio and the fifth ratio is determined to be the rigidity of the target part.

[0060] Specifically, the determination module 103 may determine the rigidity by using the following formula: In the formula, Represents rigidity, Indicates the axial load of the spindle during the vibration monitoring period, Indicates the radial load of the spindle during the vibration monitoring period; It represents the axial displacement of the spindle measured in the axial direction. Represents the radial displacement of the spindle measured in the radial direction.

[0061] in, It represents the rigidity (axial rigidity and radial rigidity) of the spindle of the target device in the cloud platform system. The larger the product, the stronger the rigidity. Stronger rigidity means that the deformation of the spindle is smaller when it is subjected to force, which is more helpful in maintaining machining accuracy and reducing vibration. Because the spindle is deformed during operation or lacks rigidity due to external influences, it will be easier for the vibration to exhibit "abnormal" behavior, resulting in abnormal data points in the vibration monitoring data due to rigidity.

[0062] The process of finally determining the vibration deviation by the module 103 is specifically as follows: a hyperbolic tangent function is used to perform a function operation on the product of the rigidity and the complexity of parameter variation to obtain the vibration deviation.

[0063] Specifically, the vibration deviation can be calculated using the following formula: In the formula, Indicates the vibration deviation, Indicates the rigidity of the target device in the cloud platform system; Indicates the complexity of parameter changes of the target device during the data monitoring period of the cloud platform system. represents the hyperbolic tangent function.

[0064] in, It represents the vibration deviation degree of the main shaft vibration of the target device in the cloud platform system. The larger the product, the greater the vibration deviation degree of the vibration monitoring data in the cloud platform system, that is, the higher the possibility that the abnormal cause of the "abnormal" data points in the vibration monitoring data belongs to equipment abnormality, indicating that there may be less external noise interference (i.e., noise intensity) in the vibration monitoring data. The smaller the product, the smaller the vibration deviation degree of the vibration monitoring data in the cloud platform system, that is, the higher the possibility that the abnormal cause of the "abnormal" data points in the vibration monitoring data belongs to external noise interference, indicating that there may be more external interference in the vibration monitoring data. Thus, the abnormality in the vibration monitoring data is adjusted and purified.

[0065] The determination module 103 is further configured to determine the processing requirement degree of the vibration monitoring data according to the average abnormality degree and the vibration deviation degree.

[0066] The processing requirement degree indicates the possibility that the abnormal data points in the vibration monitoring data are caused by noise interference, and the processing requirement degree is directly proportional to the possibility that the abnormal data points in the vibration monitoring data are caused by noise interference.

[0067] Specifically, in the industrial design cloud platform system based on artificial intelligence, a large amount of vibration monitoring data needs to be integrated and processed to achieve effective monitoring and management of the cloud platform system. The processing of vibration monitoring data requires effectively removing the interference components in the vibration monitoring data and retaining the real abnormal data caused by the device itself in the vibration monitoring data for subsequent monitoring and prevention. Therefore, the processing requirement degree can be used to determine whether the cause of data abnormality in the vibration monitoring data is caused by noise interference. The greater the processing requirement degree, the greater the possibility that the abnormal cause of the abnormal data points in the vibration monitoring data is caused by noise interference. The smaller the processing requirement degree, the smaller the possibility that the abnormal cause of the abnormal data points in the vibration monitoring data is caused by noise interference.

[0068] Among them, the process for the determination module 103 to determine the processing requirement degree is as follows: Determine the sum value of the square of the average abnormality degree and the square of the reciprocal of the vibration deviation degree; perform a square root operation on the sum value to obtain the operation result, and use the operation result as the processing requirement degree.

[0069] Further, the calculation formula for the processing requirement degree is as follows: In the formula, represents the processing requirement degree, W represents the average abnormality degree of the vibration monitoring data in the cloud platform system, and T represents the vibration deviation degree of the main shaft vibration of the target device in the cloud platform system.

[0070] Among them, It represents the Euclidean norm of the average anomaly degree of the vibration monitoring data in the cloud platform system and the vibration deviation degree of the main shaft vibration of the target device in the cloud platform system. The purpose of the Euclidean norm is to output a comprehensive index based on the anomalies in the vibration monitoring data and the operating state of the target device. Moreover, the larger the value of the Euclidean norm, the greater the possibility that the cause of the abnormal data in the vibration monitoring data belongs to external noise interference, that is, there may be a relatively high degree of noise interference in the vibration monitoring data, which means that the processing requirement degree of the vibration monitoring data in the cloud platform system is higher, and it is necessary to denoise and purify the vibration monitoring data.

[0071] The processing module 104 is used to perform smoothing processing on the vibration monitoring data based on the processing requirement degree to remove the noise interference in the vibration monitoring data.

[0072] Specifically, after obtaining the processing requirement degree, the processing module 104 can denoise the vibration monitoring data according to the value of the processing requirement degree. For example, when the processing requirement degree is greater than the threshold, the processing module 104 can determine that it is necessary to remove the noise interference in the vibration monitoring data. When the processing requirement degree is not greater than the threshold, it means that the abnormal data points in the vibration monitoring data are caused by the device itself and need to be retained. The threshold can be determined according to the actual situation, and the embodiments of the present application do not limit it here.

[0073] Alternatively, the processing module 104 uses the processing requirement degree as the standard deviation followed by the filtering kernel of the Gaussian filter to obtain the target Gaussian filter; and uses the target Gaussian filter to perform smoothing processing on the vibration monitoring data.

[0074] Specifically, according to the calculated processing requirement degree of the vibration monitoring data collected in the cloud platform system, that is, using the processing requirement degree as the intensity value of the noise intensity included in the vibration monitoring data center, and using the processing requirement degree as the standard deviation followed by the filtering kernel of the Gaussian filter, that is, using the adjusted filter (target Gaussian filter) to smooth the vibration monitoring data collected in the cloud platform system, removing the noise interference in the vibration monitoring data, and then obtaining the monitoring data with high data quality. According to the vibration monitoring data after removing the noise interference, the safety of the industrial environment and the production efficiency can be effectively improved, and then more efficient production operation and product design optimization can be realized.

[0075] It should be noted that in the above embodiments, the target device is a numerically controlled machine tool, the target part is the main shaft, and the operating parameters are the operating parameters of the main shaft as an example for illustration. According to the actual situation, the embodiments of the present application can also be applied to any type of target device and the corresponding target part of the target device. The embodiments of the present application do not make any limitations on the target device and the target part here.

[0076] The present invention has the following beneficial effects: The acquisition module acquires the vibration monitoring data of the target part of the target device during the monitoring period and the operating parameters of the target part of the target device during operation. Then, the decomposition module performs independent component analysis on the vibration monitoring data to obtain a plurality of vibration components and corresponding signal component sequences. Next, the determination module determines the average abnormality degree of the vibration monitoring data based on the amplitudes of the data points of the signal component sequences of the respective vibration components and the signal component sequences, and determines the vibration deviation degree of the vibration monitoring data of the target part based on the operating parameters. This vibration deviation degree can reflect whether the abnormal data points in the vibration monitoring data are caused by the device, so that possible faults of the target device during the production process can be predicted and processed in a timely manner, improving production efficiency. Finally, the determination module determines the processing requirement degree of the vibration monitoring data according to the average abnormality degree and the vibration deviation degree. This processing requirement degree can reflect whether the abnormal data points in the vibration monitoring data are caused by noise interference, so that the processing module can perform smoothing processing on the vibration monitoring data based on the processing requirement degree to remove the noise interference in the vibration monitoring data, improving the accuracy and reliability of the monitoring data, and facilitating the whole-process management and optimization of industrial products from concept to production.

[0077] Each embodiment in this specification is described in a progressive manner. For the same or similar parts among the embodiments, reference can be made to each other. Each embodiment focuses on the differences from other embodiments.

Claims

1. An industrial design cloud platform system based on artificial intelligence, characterized in that, The industrial design cloud platform system based on artificial intelligence includes: An acquisition module, configured to acquire vibration monitoring data of a target part of a target device within a monitoring duration and operating parameters of the target part of the target device during operation; A decomposition module, configured to perform independent component analysis on the vibration monitoring data to obtain a plurality of vibration components and corresponding signal component sequences for each of the vibration components; A determination module, configured to determine an average abnormality degree of the vibration monitoring data according to amplitudes of data points of the signal component sequences of the vibration components and the signal component sequences, where the average abnormality degree indicates the authenticity of the vibration monitoring data, and the average abnormality degree is inversely proportional to the authenticity; The determination module is configured to determine a vibration deviation degree of the vibration monitoring data of the target part according to the operating parameters, where the vibration deviation degree indicates the possibility that abnormal data points of the vibration monitoring data are caused by the target device, and the vibration deviation degree is directly proportional to the possibility that abnormal data points of the vibration monitoring data are caused by the target device; The determination module is further configured to determine a processing requirement degree of the vibration monitoring data according to the average abnormality degree and the vibration deviation degree, where the processing requirement degree indicates the possibility that abnormal data points in the vibration monitoring data are caused by noise interference, and the processing requirement degree is directly proportional to the possibility that abnormal data points in the vibration monitoring data are caused by noise interference; A processing module, configured to perform smoothing processing on the vibration monitoring data based on the processing requirement degree to remove noise interference in the vibration monitoring data.

2. The industrial design cloud platform system based on artificial intelligence according to claim 1, wherein The determination module is further configured to: Determine the number of types of data points and the total number of data points in the signal component sequence of the vibration component, where data points with equal values are of the same type; Determine the local complexity of the vibration component among all vibration components according to the distribution probability values of various types of data points, the number of types, and the total number of data points in the signal component sequence of the vibration component, where the local complexity indicates the periodicity of the vibration component, and the local complexity is inversely proportional to the periodicity of the vibration component; Determine the vibration abnormality degree of each data point in the signal component sequence of the vibration component according to the local complexity and the amplitude of the data point in the signal component sequence, where the vibration abnormality degree indicates the possibility of the data point being abnormal, and the vibration abnormality degree is directly proportional to the possibility of the data point being abnormal; Determine the average abnormality degree of the vibration monitoring data according to the vibration abnormality degrees of the data points of each vibration component.

3. The industrial design cloud platform system based on artificial intelligence according to claim 2, wherein The determination module is further configured to: Determine the absolute value of the difference between the distribution probability values of any two types of data points in the signal component sequence of the vibration component to obtain m absolute values of differences; Perform superposition on the m absolute values of differences and then divide by m to obtain a first ratio, where the first ratio indicates the number of similar data points existing in the vibration monitoring data, and the first ratio is inversely proportional to the number of similar data points existing in the vibration monitoring data, and m is an integer greater than 0; Determine a second ratio between the number of types of the signal component sequence of the vibration component and the total number of data points; Determine the product of the first ratio and the second ratio as the local complexity.

4. The industrial design cloud platform system based on artificial intelligence according to claim 2, characterized in that, The determining module is further configured to: Determine the average amplitude of the amplitudes of all data points in the signal component sequence of the vibration component, and a first difference between the maximum amplitude and the minimum amplitude of the data points in the signal component sequence; Determine the absolute value of a second difference between the amplitude of the data point and the average amplitude, and the absolute value of a third difference between the amplitude of the data point and the first difference; Determine a third ratio of the absolute value of the second difference to the absolute value of the third difference; Perform a function operation on the local complexity using the hyperbolic tangent function to obtain an operation result; Determine the product of the third ratio and the operation result, and use the product of the third ratio and the operation result as the vibration abnormality degree of the data points in the signal component sequence.

5. The industrial design cloud platform system based on artificial intelligence according to any one of claims 1-4, characterized in that, The determining module is further configured to: Determine the feed rate of the target part determined from the operating parameters during the monitoring period; Determine the error between the operating parameters and the vibration monitoring data; Determine the parameter change complexity of the operating parameters according to the error and the feed rate, where the parameter change complexity indicates the possibility that the abnormal data points in the vibration monitoring data are caused by the operating parameters of the target part, and the parameter change complexity is proportional to the possibility that the abnormal data points in the vibration monitoring data are caused by the operating parameters of the target part; Determine the load on the target part during the monitoring period and the displacement generated by the target part under the load; Determine the rigidity of the target part according to the load and the displacement, where the rigidity is inversely proportional to the probability that the data points in the vibration monitoring data are abnormal; Perform a function operation on the product of the rigidity and the parameter change complexity using the hyperbolic tangent function to obtain the vibration deviation degree.

6. The industrial design cloud platform system based on artificial intelligence according to claim 5, characterized in that The determining module is further configured to: Fit the operating parameters of the target part according to the monitoring period of the vibration monitoring data to obtain a fitting curve; Determine the mean square error between the fitting curve and the vibration monitoring curve of the vibration monitoring data as the error; Determine the degree of fluctuation of the feed rate of the target part during the monitoring period; Determine the parameter change complexity of the operating parameters according to the mean square error and the degree of fluctuation.

7. The industrial design cloud platform system based on artificial intelligence according to claim 6, characterized in that The determining module is further configured to: Determine the ratio of the degree of fluctuation to the mean square error as the parameter change complexity.

8. The industrial design cloud platform system based on artificial intelligence according to claim 5, characterized in that, The load includes a radial load and an axial load, and the displacement includes an axial displacement and a radial displacement. The obtaining module is further configured to: Obtain the radial load and the axial load of the target part during the monitoring period, and the axial displacement of the target part in the axial direction and the radial displacement of the target part in the radial direction; The determining module is further configured to: Determine a fourth ratio between the axial load and the axial displacement, and a fifth ratio between the radial load and the radial displacement; Determine the product of the fourth ratio and the fifth ratio as the rigidity of the target part.

9. The industrial design cloud platform system based on artificial intelligence according to claim 1, characterized in that, The determining module is further configured to: Determine the sum of the square of the average abnormality degree and the square of the reciprocal of the vibration deviation degree; Perform a square root operation on the sum value to obtain an operation result, and use the operation result as the processing requirement degree.

10. The industrial design cloud platform system based on artificial intelligence according to claim 1, characterized in that The processing module is further configured to: Use the processing requirement degree as the standard deviation followed by the filter kernel of the Gaussian filter to obtain a target Gaussian filter; Perform smoothing processing on the vibration monitoring data by using the target Gaussian filter.

Images