Data collection method, system and storage medium based on industrial Internet of Things

Abstract

The present application relates to the field of data collection, and specifically discloses a data collection method, system and storage medium based on the industrial Internet of Things, which includes industrial equipment, an equipment access module, a protocol conversion module, an edge data processing module and an industrial Internet platform. When pre-processing data, by obtaining current COV data and a historical data set of COV data, and using a deep learning-based data analysis technology to perform time series correlation analysis on the current COV data, and performing time series correlation extraction and feature force field aggregation on the historical data set of COV data, it is automatically determined whether to delete the current COV data based on the semantic difference measure between the historical COV time series correlation aggregation feature and the current COV data time series correlation feature, which can effectively identify patterns and trends between current and historical data, thereby improving the overall performance and reliability of the industrial Internet of Things system.

Description

Technical Field

[0001] The present application relates to the field of data collection, and more specifically, to a data collection method, system and storage medium based on industrial Internet of Things. Background Art

[0002] With the development of Industry 4.0, the Industrial Internet of Things has become an indispensable part of modern manufacturing. In the Industrial Internet of Things, a large number of sensors and devices continuously generate massive amounts of data, which need to be efficiently collected, processed and analyzed to support real-time monitoring, predictive maintenance, optimized production and decision support.

[0003] Chinese patent CN118138614A proposes an industrial Internet of Things data acquisition system, which converts heterogeneous data from different industrial equipment into a standardized COV format through a custom protocol library. Then, the converted data is preprocessed and preliminarily analyzed on the edge side, abnormal data is removed and locally cached. The processed data is then transmitted to the industrial Internet platform, and finally the platform classifies and saves the data according to time nodes.

[0004] The above patents use desensitization, smoothing filtering and eigenvalue iteration techniques to pre-process the converted COV data, although these techniques can reduce duplicate, redundant, missing or erroneous data. However, traditional desensitization, smoothing filtering and eigenvalue iteration techniques mainly focus on the processing of single data and fail to fully consider the temporal correlation of data. During the processing, the relationship between current data and historical data may be ignored, making it impossible to effectively identify patterns and trends between data. In addition, although smoothing filtering can reduce noise, over-smoothing may lead to the loss of useful information, especially information that reflects sudden changes in equipment status.

[0005] Therefore, an optimized data acquisition system based on industrial Internet of Things is desired. Summary of the invention

[0006] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a data collection method, system and storage medium based on the industrial Internet of Things, which includes industrial equipment, equipment access module, protocol conversion module, edge data processing module and industrial Internet platform, wherein when processing data, by obtaining the current COV data, and obtaining the historical data set of COV data, and using the data analysis and processing technology based on deep learning to perform time series correlation analysis on the current COV data, and perform time series correlation extraction and feature force field aggregation on the historical data set of the COV data, so as to automatically determine whether to delete the current COV data according to the semantic difference metric between the historical COV time series correlation aggregation feature and the current COV data time series correlation feature, which can effectively identify the patterns and trends between the current and historical data, make up for the deficiency of traditional technology that only focuses on single data processing, and ensure the time series correlation in the data processing process. In addition, by retaining important information reflecting the equipment data through time series correlation analysis, the loss of useful information is avoided, and more efficient data processing and intelligent decision support are achieved, thereby improving the overall performance and reliability of the industrial Internet of Things system.

[0007] According to one aspect of the present application, a data acquisition system based on industrial Internet of Things is provided, which includes: industrial equipment, equipment access module, protocol conversion module, edge data processing module and industrial Internet platform.

[0008] The device access module is used to access the industrial device; the protocol conversion module is used to convert the data collected by the industrial device into COV data in a standardized format;

[0009] The edge data processing module is used to preprocess, store and analyze the COV data, and transmit the analyzed COV data to the industrial Internet platform; wherein the edge data processing module includes: a COV data preprocessing unit, used to preprocess the COV data to obtain preprocessed COV data; a COV data storage and analysis unit, used to store and analyze the preprocessed COV data to obtain analyzed COV data; a COV data transmission unit, used to transmit the analyzed COV data to the industrial Internet platform through a wireless communication module;

[0010] The industrial Internet platform is used to number and classify the data uploaded by the edge data processing module;

[0011] Among them, the COV data preprocessing unit includes: a data acquisition subunit, which is used to acquire the current COV data and the historical data set of COV data; a data feature extraction subunit, which is used to perform time series feature extraction on the current COV data and the historical data set of COV data to obtain a set of current COV time series associated implicit features and historical COV time series associated implicit features; a historical data association aggregation subunit, which is used to perform dynamic feature force field aggregation on the set of historical COV time series associated implicit features to obtain a historical COV time series associated significant aggregation representation; a data deletion subunit, which is used to determine whether to delete the current COV data based on the semantic difference measure between the historical COV time series associated significant aggregation representation and the current COV time series associated implicit features.

[0012] According to another aspect of the present application, a data collection method based on industrial Internet of Things is provided, which includes:

[0013] Get the current COV data and get the historical data set of COV data;

[0014] Performing time series feature extraction on the current COV data and the historical data set of the COV data to obtain a set of current COV time series associated implicit features and historical COV time series associated implicit features;

[0015] Performing dynamic feature force field aggregation on the set of historical COV time series association implicit features to obtain a significant aggregation representation of historical COV time series association;

[0016] Based on the semantic difference measure between the historical COV time series association significant aggregation representation and the current COV time series association implicit feature, determine whether to delete the current COV data.

[0017] According to another aspect of the present application, a storage medium is also provided, comprising: computer program instructions are stored in the storage medium, and when the computer program instructions are executed by a processor, the processor executes the data acquisition system based on the industrial Internet of Things as described.

[0018] This application has at least the following technical effects:

[0019] Compared with the prior art, the present application provides a data collection method, system and storage medium based on the industrial Internet of Things, which includes industrial equipment, equipment access module, protocol conversion module, edge data processing module and industrial Internet platform, wherein when processing data, by obtaining the current COV data, and obtaining the historical data set of COV data, and using the data analysis and processing technology based on deep learning to perform time series association analysis on the current COV data, and perform time series association extraction and feature field aggregation on the historical data set of the COV data, so as to automatically determine whether to delete the current COV data according to the semantic difference metric between the historical COV time series association aggregation feature and the current COV data time series association feature. In this way, the patterns and trends between the current and historical data can be effectively identified, which makes up for the deficiency of traditional technology that only focuses on single data processing, and ensures the time series correlation in the data processing process. In addition, by retaining important information reflecting the equipment data through time series association analysis, the loss of useful information is avoided, and more efficient data processing and intelligent decision support are achieved, thereby improving the overall performance and reliability of the industrial Internet of Things system. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0021] Figure 1 It is a block diagram of a data acquisition system based on industrial Internet of Things according to an embodiment of the present application.

[0022] Figure 2 It is a block diagram of an edge data processing module in a data acquisition system based on industrial Internet of Things according to an embodiment of the present application.

[0023] Figure 3 It is a block diagram of a COV data preprocessing unit in a data acquisition system based on industrial Internet of Things according to an embodiment of the present application.

[0024] Figure 4 This is a schematic diagram of data flow of a COV data preprocessing unit in a data acquisition system based on the industrial Internet of Things according to an embodiment of the present application.

[0025] Figure 5 It is a block diagram of a historical data association aggregation subunit in a data acquisition system based on the industrial Internet of Things according to an embodiment of the present application.

[0026] Figure 6The present invention is a flowchart of a data collection method based on industrial Internet of Things according to an embodiment of the present application. DETAILED DESCRIPTION

[0027] The embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments described herein, but rather these embodiments are provided for a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0028] It should be understood that the various steps described in the method embodiments of the present disclosure may be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present disclosure is not limited in this respect.

[0029] In the description of the embodiments of the present disclosure, the term "including" and similar terms should be understood as open inclusion, that is, "including but not limited to". The term "based on" should be understood as "based at least in part on". The term "one embodiment" or "the embodiment" should be understood as "at least one embodiment". The terms "first", "second", etc. may refer to different or the same objects. Other explicit and implicit definitions may also be included below.

[0030] It should be noted that the modifications of "one" and "plurality" mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0031] It should also be noted that in this application, all actions to obtain signals, information or data are carried out in compliance with the relevant data protection laws and policies of the country where they are located and with the authorization given by the owner of the corresponding device.

[0032] Specifically, Figure 1 FIG. 1 is a block diagram of a data acquisition system based on the industrial Internet of Things according to an embodiment of the present application. Figure 1As shown, according to an embodiment of the present application, the data acquisition system 100 based on the industrial Internet of Things includes: industrial equipment 110, an equipment access module 120, a protocol conversion module 130, an edge data processing module 140 and an industrial Internet platform 150; specifically, the equipment access module is used to access the industrial equipment; the protocol conversion module is used to convert the data collected by the industrial equipment into COV data in a standardized format; the edge data processing module is used to preprocess and store and analyze the COV data, and transmit the analyzed COV data to the industrial Internet platform; wherein, the edge data processing module includes: a COV data preprocessing unit, which is used to preprocess the COV data to obtain preprocessed COV data; a COV data storage and analysis unit, which is used to store and analyze the preprocessed COV data to obtain analyzed COV data; a COV data transmission unit, which is used to transmit the analyzed COV data to the industrial Internet platform through a wireless communication module; the industrial Internet platform is used to number and classify the data uploaded by the edge data processing module.

[0033] Specifically, in the embodiment of the present application, the industrial equipment 110 includes industrial field equipment and intelligent products. Specifically, in a specific embodiment of the present application, the industrial field equipment includes transmission equipment, data acquisition devices, sensors, programmable logic controllers (PLCs), remote terminal units (RTUs), industrial personal computers (IPCs), automated robots, and computer numerical control (CNC) machine tools. At the same time, the fields covered by intelligent products include engineering machinery equipment, household appliances with intelligent functions, and intelligent manufacturing equipment, etc. These equipment and products play an important role in intelligent industrial applications.

[0034] Specifically, in the embodiment of the present application, the device access module 120 is used to access the industrial device 110. Specifically, the device access module is an important component of the industrial Internet of Things data acquisition system, and its main function is to connect various industrial devices and transmit the data generated by the devices to subsequent data processing links.

[0035] Specifically, in an embodiment of the present application, the protocol conversion module 130 is used to convert the data collected by the industrial equipment into COV data in a standardized format. Accordingly, the protocol conversion module parses the data packets of different devices, and then converts these data into a unified COV format and identifies them. In this way, the consistency and processability of the data are ensured, which facilitates subsequent preprocessing, storage and analysis. In particular, the protocol conversion module also supports multiple communication protocols and ensures the security and integrity of the data during the conversion process.

[0036] Specifically, in the embodiment of the present application, the edge data processing module 140 is used to preprocess, store and analyze the COV data, and transmit the analyzed COV data to the industrial Internet platform. Figure 2 is a block diagram of an edge data processing module in a data acquisition system based on the industrial Internet of Things according to an embodiment of the present application. More specifically, in an embodiment of the present application, Figure 2 As shown, the edge data processing module 140 includes: a COV data preprocessing unit 141, used to preprocess the COV data to obtain preprocessed COV data; a COV data storage and analysis unit 142, used to store and analyze the preprocessed COV data to obtain analyzed COV data; a COV data transmission unit 143, used to transmit the analyzed COV data to the industrial Internet platform through a wireless communication module.

[0037] Correspondingly, in the COV data preprocessing unit, the technical concept of the present application is to obtain the current COV data and the historical data set of COV data, and use the data analysis and processing technology based on deep learning to perform time series correlation analysis on the current COV data, and perform time series correlation extraction and feature force field aggregation on the historical data set of COV data, so as to automatically determine whether to delete the current COV data according to the semantic difference metric between the historical COV time series correlation aggregation feature and the current COV data time series correlation feature, which can effectively identify the patterns and trends between the current and historical data, make up for the deficiency of traditional technology that only focuses on single data processing, and ensure the time series correlation in the data processing process. In addition, by retaining important information reflecting the equipment data through time series correlation analysis, the loss of useful information is avoided, and more efficient data processing and intelligent decision support are achieved, thereby improving the overall performance and reliability of the industrial Internet of Things system.

[0038] Figure 3 It is a block diagram of a COV data preprocessing unit in a data acquisition system based on industrial Internet of Things according to an embodiment of the present application. Figure 4 This is a data flow diagram of the COV data preprocessing unit in the data acquisition system based on the industrial Internet of Things according to an embodiment of the present application. Figure 3 and Figure 4As shown, the COV data preprocessing unit 141 includes: a data acquisition subunit 1411, which is used to acquire current COV data and acquire a historical data set of COV data; a data feature extraction subunit 1412, which is used to perform time series feature extraction on the current COV data and the historical data set of COV data to obtain a set of current COV time series associated implicit features and historical COV time series associated implicit features; a historical data association aggregation subunit 1413, which is used to perform dynamic feature force field aggregation on the set of historical COV time series associated implicit features to obtain a historical COV time series associated significant aggregation representation; a data deletion subunit 1414, which is used to determine whether to delete the current COV data based on a semantic difference measure between the historical COV time series associated significant aggregation representation and the current COV time series associated implicit features.

[0039] In an embodiment of the present application, the data acquisition subunit 1411 is used to acquire current COV data and acquire a historical data set of COV data. Specifically, considering that the current COV data refers to data collected from industrial equipment at a certain point in time, these data reflect the operating status of the equipment at that point in time, such as temperature, pressure, vibration and other sensor data. The historical data set of COV data refers to a collection of all COV data collected in the past period of time, which records the operating conditions of the equipment at different time points. Based on this, in the technical solution of the present application, the current COV data is acquired, and a historical data set of COV data is acquired. That is, in combination with the current COV data and the historical data set, it can be determined whether the current data needs to be retained by comparing the difference between the current data and the historical pattern to ensure the quality and integrity of data storage.

[0040] In the embodiment of the present application, the data feature extraction subunit 1412 is used to extract time series features from the current COV data and the historical data set of the COV data to obtain a set of current COV time series associated implicit features and historical COV time series associated implicit features. Specifically, in the embodiment of the present application, the data feature extraction subunit includes: a current COV data acquisition time series associated secondary subunit, which is used to extract the current COV time series associated implicit feature vector from the current COV data; a historical COV data acquisition time series associated secondary subunit, which is used to extract time series features from each COV data in the historical data set of the COV data to obtain a set of historical COV time series associated implicit feature vectors;

[0041] More specifically, in an embodiment of the present application, the current COV data obtains a time-series-associated secondary subunit, which is used to process the current COV data using a time-series extractor based on an RNN model to obtain the current COV time-series-associated implicit feature vector. The historical COV data obtains a time-series-associated secondary subunit, which is used to process each COV data in the historical data set of the COV data using the time-series extractor based on the RNN model to obtain a set of the historical COV time-series-associated implicit feature vectors. Accordingly, considering that the data of the current COV data and each COV data in the historical data set of the COV data changes with the passage of time, at the same time, the current COV data and each COV data have different time-series-associated characteristics on different time scales. Therefore, in order to more accurately analyze the behavior patterns at different time points, in the technical solution of the present application, the current COV time series associated implicit feature vector is extracted from the current COV data, and the time series feature extraction is performed on each COV data in the historical data set of the COV data to obtain a set of historical COV time series associated implicit feature vectors. In this way, the current and historical COV time series features can be more clearly understood, providing better support for subsequent data decisions. In particular, in a specific embodiment of the present application, the current COV data is processed by using a time series extractor based on the RNN model to obtain the current COV time series associated implicit feature vector, and the time series extractor based on the RNN model is used to process each COV data in the historical data set of the COV data to obtain the set of historical COV time series associated implicit feature vectors. It can be understood that the RNN model can capture the long-term dependencies in these data and the implicit associations between data in different time spans, which is crucial for understanding the dynamic behavior of complex industrial systems.

[0042] In the embodiment of the present application, the historical data association aggregation subunit 1413 is used to perform dynamic feature force field aggregation on the set of historical COV time series association implicit features to obtain a historical COV time series association significant aggregation representation. Figure 5 is a block diagram of a historical data association aggregation subunit in a data acquisition system based on the industrial Internet of Things according to an embodiment of the present application. Specifically, in the embodiment of the present application, Figure 5As shown, the historical data association aggregation subunit 1413 includes: a historical COV data static factor calculation clustering secondary subunit 14131, which is used to calculate the static energy factor of each historical COV time series association implicit feature vector in the set of historical COV time series association implicit feature vectors to obtain a set of historical COV static energy factors, and based on the set of historical COV static energy factors, obtain the historical COV clustering initial center vector; a historical COV data energy force field aggregation secondary subunit 14132, which is used to perform energy force field-based feature aggregation on the set of historical COV time series association implicit feature vectors based on the set of historical COV static energy factors and the historical COV clustering initial center vector to obtain the historical COV time series association significant aggregation representation. Specifically, considering that different features in the set of historical COV time series association implicit feature vectors have different influences and significances over the entire time range, in order to extract more significant and meaningful feature representations from historical data and obtain a comprehensive and comprehensive aggregate representation feature, in the technical solution of the present application, the set of historical COV time series association implicit feature vectors is subjected to dynamic feature force field aggregation to obtain a historical COV time series association significant aggregate representation vector as a historical COV time series association significant aggregate representation.

[0043] In detail, first, the static energy factor of each historical COV time series associated implicit feature vector in the set of the historical COV time series associated implicit feature vector is calculated to obtain a set of historical COV static energy factors, and based on the set of historical COV static energy factors, the historical COV clustering initial center vector is obtained, which can characterize different features based on the contribution of each feature and find the most representative clustering feature. Specifically, the importance or influence of each historical COV time series associated implicit feature vector can be evaluated by calculating the static energy factor of the historical COV time series associated implicit feature vector. In particular, the static energy factor can be regarded as an indicator of the importance of the feature vector in a specific context. Then, the historical COV time series associated implicit feature vector corresponding to the largest historical COV static energy factor in the set of historical COV static energy factors is selected as the clustering initial center vector. That is, the historical COV time series associated implicit feature vector corresponding to the maximum static energy factor usually represents the most representative or influential data, and using it as the initial center vector can ensure that the clustering algorithm starts from the most representative data point, thereby optimizing the clustering effect.

[0044] Next, based on the set of historical COV static energy factors and the initial center vector of the historical COV clustering, the set of historical COV time series associated implicit feature vectors is subjected to feature aggregation based on the energy force field to obtain the historical COV time series associated significant aggregation representation vector. In particular, the concept of energy force field draws on the force field theory in physics. By simulating the interaction force between particles, the distribution of features in multidimensional space can be optimized, wherein the static energy factor can be regarded as the "quality" of the feature vector, that is, it represents the importance of the feature vector. Feature vectors with larger static energy factors have greater "gravitational force" or importance in the simulated force field, making the original features more compact and representative. Specifically, the set of historical COV time series dynamic aggregation energy factors is calculated based on the static energy factors of each historical COV time series associated implicit feature vector and the initial center vector, and the spatial span between each historical COV time series associated implicit feature vector and the initial center vector, so as to more accurately reflect the real situation of the interaction between different data. Next, in order to dynamically and adaptively adjust the weights of each factor, it is converted into weights, and the set of dynamic aggregation energy factors is input into the gated mask unit for processing to obtain a set of dynamic aggregation weight factors of the historical COV time series. Finally, each value in the set of weight factors is used as a weight to weight and calculate the historical COV time series associated implicit feature vectors to ensure that those more representative and influential feature vectors occupy a larger proportion in the final aggregate representation, so that the final historical COV time series associated significant aggregate representation vector can better reflect the core features of the historical data.

[0045] Specifically, in an embodiment of the present application, the historical COV data static factor calculation clustering secondary subunit 14131 includes: a historical COV data static energy factor calculation tertiary subunit, which is used to calculate the static energy factor of each historical COV time series associated implicit feature vector in the set of historical COV time series associated implicit feature vectors to obtain the set of historical COV static energy factors; a historical COV data clustering tertiary subunit, which is used to select the historical COV time series associated implicit feature vector corresponding to the largest historical COV static energy factor in the set of historical COV static energy factors as the initial center vector of the historical COV clustering.

[0046] More specifically, in an embodiment of the present application, the three-level sub-unit for calculating the static energy factor of historical COV data is used to: calculate the mean and variance of the implicit feature vector associated with the historical COV time series to obtain the historical COV time series mean and the historical COV time series variance; subtract the implicit feature vector associated with the historical COV time series from the historical COV time series mean by position, and calculate the fourth power of each position of the feature vector after subtraction to obtain the historical COV time series differential modulation vector; calculate the expected value of the historical COV time series differential modulation vector to obtain the historical COV time series expected value; divide the historical COV time series expected value by the square of the historical COV time series variance, and input the divided value into the sigmoid function to obtain the historical COV static energy factor.

[0047] Specifically, in the embodiment of the present application, the historical COV data energy force field aggregation secondary subunit 14132 includes: a historical COV data time series dynamic aggregation energy factor calculation tertiary subunit, which is used to calculate each historical COV time series associated implicit feature vector in the set of the historical COV time series associated implicit feature vectors based on the spatial span between each historical COV time series associated implicit feature vector and the historical COV cluster initial center vector and the static energy factor of each historical COV time series associated implicit feature vector and the static energy factor of the historical COV cluster initial center vector. The dynamic aggregation energy factor of the implicit feature vector is used to obtain a set of historical COV time series dynamic aggregation energy factors; the historical COV time series dynamic aggregation energy factor gated mask three-level sub-unit is used to input the set of historical COV time series dynamic aggregation energy factors into the gated mask unit to obtain a set of historical COV time series dynamic aggregation weight factors; the historical COV data aggregation three-level sub-unit is used to calculate the weighted sum of the set of historical COV time series associated implicit feature vectors with the set of historical COV time series dynamic aggregation weight factors to obtain the historical COV time series associated significant aggregation representation vector as the historical COV time series associated significant aggregation representation.

[0048] More specifically, in an embodiment of the present application, the three-level sub-unit for calculating the dynamic aggregation energy factor of the historical COV data time series is used to: multiply the static energy factor of the implicit feature vector associated with the historical COV time series and the static energy factor of the initial center vector of the historical COV cluster by a first weighting parameter to obtain a first historical COV time series dynamic aggregation energy factor; multiply the square of the spatial span between the implicit feature vector associated with the historical COV time series and the initial center vector of the historical COV cluster by a second weighting parameter to obtain a second historical COV time series dynamic aggregation energy factor; divide the first historical COV time series dynamic aggregation energy factor by the second historical COV time series dynamic aggregation energy factor to obtain the historical COV time series dynamic aggregation energy factor.

[0049] More specifically, in an embodiment of the present application, the historical COV timing dynamic aggregation energy factor gated mask three-level sub-unit is used to: use a sigmoid function to normalize the set of historical COV timing dynamic aggregation energy factors to obtain a set of normalized historical COV timing dynamic aggregation energy factors; mask the set of normalized historical COV timing dynamic aggregation energy factors to obtain a set of historical COV timing dynamic aggregation weight factors, wherein, in response to each normalized historical COV timing dynamic aggregation energy factor in the set of normalized historical COV timing dynamic aggregation energy factors being greater than a predetermined threshold, the normalized historical COV timing dynamic aggregation energy factor is set to the original value, and the rest are set to zero, so as to obtain the set of historical COV timing dynamic aggregation weight factors.

[0050] In the embodiment of the present application, specifically, the dynamic feature force field aggregation is performed on the set of the historical COV time series associated implicit feature vectors, which can be expressed by the formula:

[0051]

[0052]

[0053] in, is the set of implicit feature vectors associated with the historical COV time series, are the first, second, ..., and third implicit feature vectors in the historical COV time series association set. ,..., The historical COV time series associated implicit feature vector, It is The eigenvalues ​​of each position in the historical COV time series associated implicit eigenvector, To calculate the expected value, and They are The mean and variance of is the sigmoid function, yes The corresponding historical COV static energy factor, To select the maximum value corresponding to value, is the maximum matching value, is the initial center vector of historical COV clustering, yes The corresponding historical COV static energy factor, express and The space span between and is the weighting parameter, for The corresponding historical COV time series dynamic aggregation energy factor, yes The corresponding normalized historical COV time series dynamic aggregation energy factor, For masking, is a predetermined threshold, yes The corresponding historical COV time series dynamic aggregation weight factor, is the number of vectors in the set of the historical COV time series associated implicit feature vectors, is the significant aggregate representation vector of the historical COV temporal association.

[0054] In an embodiment of the present application, the data deletion subunit 1414 is used to determine whether to delete the current COV data based on the semantic difference measurement value between the historical COV time series association significant aggregation representation and the current COV time series association implicit feature. Specifically, in an embodiment of the present application, the data deletion subunit is used to: input the historical COV time series association significant aggregation representation vector and the current COV time series association implicit feature vector into the semantic difference measurement network to obtain the semantic difference measurement value; based on the comparison between the semantic difference measurement value and the preset threshold, determine whether to delete the current COV data. Accordingly, in order to more carefully and accurately compare and measure the difference or similarity between the historical COV time series association significant aggregation representation vector and the current COV time series association implicit feature vector to provide data assistance for subsequent decision-making, in the technical solution of the present application, the historical COV time series association significant aggregation representation vector and the current COV time series association implicit feature vector are input into the semantic difference measurement network to obtain the semantic difference measurement value. In particular, the semantic difference measurement network is usually a neural network model designed to calculate the similarity or difference between two vectors. It can be based on distance measurement (such as Euclidean distance, Manhattan distance) or similarity measurement (such as cosine similarity). Then, based on the comparison between the semantic difference measurement value and the preset threshold, it is determined whether to delete the current COV data. That is, the semantic difference measurement value obtained by semantic measurement using the historical COV time series association significant aggregation representation vector and the current COV time series association implicit feature vector is compared with the preset threshold to automatically determine whether to delete the current COV data. This can effectively identify the patterns and trends between current and historical data, make up for the deficiency of traditional technology that only focuses on single data processing, and ensure the time series correlation in the data processing process. In addition, by retaining important information reflecting the status of the equipment through time series correlation analysis, the loss of useful information is avoided, and more efficient data processing and intelligent decision support are achieved, thereby improving the overall performance and reliability of the industrial Internet of Things system.

[0055] Specifically, in the technical solution of the present application, the historical COV temporal association significant aggregation representation vector and the current COV temporal association implicit feature vector represent the dynamic feature force field aggregation features of the historical data set of COV data and the temporal association features of the current COV data, respectively. Considering the difference in sample density, the current COV temporal association implicit feature vector will have sparse feature distribution relative to the historical COV temporal association significant aggregation representation vector, which affects the decoding accuracy of the semantic difference measurement value obtained by inputting the historical COV temporal association significant aggregation representation vector and the current COV temporal association implicit feature vector into the semantic difference measurement network.

[0056] Based on this, in a preferred embodiment, the current COV time series associated implicit feature vector is feature enhanced and optimized to obtain an optimized current COV time series associated implicit feature vector, including: performing L random sampling on the current COV time series associated implicit feature vector to obtain L COV data random perturbation feature values; arranging the L COV data random perturbation feature values ​​in order from large to small to obtain a COV data random perturbation ordered vector; calculating the self-inner product of the current COV time series associated implicit feature vector to obtain the COV data energy modulation feature value; based on The COV data energy modulation eigenvalue performs semantic intensity scaling on the COV data randomly perturbed ordered vector to obtain a semantically enhanced COV data randomly perturbed ordered vector; the product between the semantically enhanced COV data randomly perturbed ordered vector and the transposed vector of the current COV time series associated implicit feature vector is calculated to obtain the COV data best sample prompt orthogonal modulation matrix; the matrix product between the COV data best sample prompt orthogonal modulation matrix and the current COV time series associated implicit feature vector is calculated to obtain the optimized current COV time series associated implicit feature vector.

[0057] The above optimization process is expressed as follows:

[0058]

[0059] in, and Indicates that the current COV time series associated implicit feature vector is The random sampling COV data randomly perturbed eigenvalues, represents L random samplings, represents the Gaussian density probability function, represents the mean value of the implicit feature vector associated with the current COV time series, represents the current COV time series associated implicit feature vector, It means to calculate the mean of the vector. represents the variance of the implicit feature vector associated with the current COV time series, is the first implicit feature vector associated with the current COV time series The eigenvalues ​​at the positions, is the length of the implicit feature vector associated with the current COV time series, Express Arrange them in order from large to small. represents the randomly disturbed ordered vector of the semantically enhanced COV data, represents the square of the vector's two-norm, represents the energy modulation characteristic value of the COV data, It means point multiplication by position. represents matrix multiplication, represents the transposed vector of the current COV time series associated implicit feature vector, The COV data best example suggests an orthogonal modulation matrix, Represents the optimized current COV time series associated implicit feature vector.

[0060] Accordingly, by performing L random sampling on the current COV time series associated implicit feature vector to obtain a set of new feature values ​​based on the original features but containing slight changes, randomness is introduced to explore different aspects of the feature space, and by simulating the noise or variation in the actual data, the model can be better generalized to unseen data. Then, these COV data randomly perturbed feature values ​​are arranged in order from large to small to form a COV data randomly perturbed ordered vector to identify which perturbation directions may be the most important for the current problem. Such an ordering not only helps the algorithm focus on the feature differences that can best distinguish different categories, but also helps to reveal hidden data structures or patterns. Subsequently, the inner product of the current COV time series associated implicit feature vector is calculated to obtain the COV data energy modulation feature value, which reflects the overall strength or "energy" of the original feature vector and provides a benchmark for measuring the importance of the feature. Using this COV data energy modulation feature value as a reference point, the previously formed COV data random perturbation ordered vector is semantically scaled to construct a semantically enhanced COV data random perturbation ordered vector. In this way, those disturbances that appear to be more important in the original feature space are amplified, thereby enhancing the model's sensitivity to changes in key features, while also retaining relatively small but potentially useful disturbance information.

[0061] Then, the product of the randomly disturbed ordered vector of semantically enhanced COV data and the transpose of the implicit feature vector associated with the current COV time series is calculated to obtain the COV data best sample prompt orthogonal modulation matrix, the purpose of which is to find the way in which the features interact with each other from two different perspectives. The formed matrix can be regarded as a conversion tool for capturing and representing the complex linear relationship between features. Finally, the final feature structure modulation is completed by applying the COV data best sample prompt orthogonal modulation matrix constructed above to the original current COV time series associated implicit feature vector, so that the features are more in line with the requirements of the objective function, reducing redundant information, while maximizing the retention of useful information, so as to improve learning efficiency and prediction performance.

[0062] In summary, the feature manifold optimization process of the current COV temporal association implicit feature vector is explained, which extracts the best prompt sample based on the fine-grained target retrieval of random perturbations, and then guides the adversarial robustness learning based on the best prompt sample to complete the structure of the high-dimensional manifold of the feature vector in the feature space, so as to improve the detail and structure of the feature expression, improve the feature distribution consistency between the historical COV temporal association significant aggregation representation vector and the current COV temporal association implicit feature vector, so as to improve the decoding accuracy of the semantic difference metric value obtained by inputting the two into the semantic difference metric network, and can effectively identify the patterns and trends between current and historical data, making up for the deficiency of traditional technology that only focuses on single data processing, and ensuring the temporal correlation in the data processing process. In addition, by retaining important information reflecting the status of the equipment through temporal correlation analysis, the loss of useful information is avoided, and more efficient data processing and intelligent decision support are achieved, thereby improving the overall performance and reliability of the industrial Internet of Things system.

[0063] Specifically, in an embodiment of the present application, the COV data storage and analysis unit 142 is used to store and analyze the preprocessed COV data to obtain analyzed COV data. Specifically, the COV data storage and analysis unit first stores the preprocessed data for subsequent calls and queries. Subsequently, the stored data is deeply analyzed to extract useful information and identify patterns and trends in the data. Through these analyses, important conclusions about the operating status of the equipment can be drawn, providing data support for predictive maintenance, fault diagnosis, and optimization of production processes.

[0064] Specifically, in the embodiment of the present application, the COV data transmission unit 143 is used to transmit the analyzed COV data to the industrial Internet platform through the wireless communication module. Accordingly, the COV data transmission unit is responsible for transmitting the analyzed COV data to the industrial Internet platform through the wireless communication module to realize real-time updating and sharing of data, and support real-time monitoring and decision support systems on the platform. Through wireless transmission, data delays can be reduced, ensuring that data arrives quickly, facilitating timely adjustment of production plans and maintenance work, and ensuring the security of data transmission.

[0065] Specifically, in the embodiment of the present application, the industrial Internet platform 150 is used to number and classify the data uploaded by the edge data processing module. Specifically, in order to better organize and manage these data, ensure the orderly storage and rapid retrieval of data. By numbering and classifying the uploaded data, the industrial Internet platform can improve the traceability and utilization efficiency of the data, support subsequent data analysis, fault diagnosis and predictive maintenance, etc., thereby improving the management level and operation efficiency of the overall system.

[0066] In summary, the data acquisition system 100 based on the industrial Internet of Things according to the embodiment of the present application is explained, including industrial equipment, equipment access module, protocol conversion module, edge data processing module and industrial Internet platform, wherein when processing data, by obtaining the current COV data and the historical data set of COV data, and using the data analysis and processing technology based on deep learning to perform time series correlation analysis on the current COV data, and perform time series correlation extraction and feature field aggregation on the historical data set of COV data, so as to automatically determine whether to delete the current COV data according to the semantic difference metric between the historical COV time series correlation aggregation feature and the current COV data time series correlation feature, it can effectively identify the patterns and trends between the current and historical data, make up for the deficiency of traditional technology that only focuses on single data processing, and ensure the time series correlation in the data processing process. In addition, by retaining important information reflecting the equipment data through time series correlation analysis, the loss of useful information is avoided, and more efficient data processing and intelligent decision support are achieved, thereby improving the overall performance and reliability of the industrial Internet of Things system.

[0067] As described above, the data acquisition system 100 based on the industrial Internet of Things according to the embodiment of the present application can be implemented in various wireless terminals, such as a server with a data acquisition algorithm based on the industrial Internet of Things. In a possible implementation, the data acquisition system 100 based on the industrial Internet of Things according to the embodiment of the present application can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the data acquisition system 100 based on the industrial Internet of Things can be a software module in the operating system of the wireless terminal, or can be an application developed for the wireless terminal; of course, the data acquisition system 100 based on the industrial Internet of Things can also be one of the many hardware modules of the wireless terminal.

[0068] Alternatively, in another example, the industrial Internet of Things-based data acquisition system 100 and the wireless terminal may also be separate devices, and the industrial Internet of Things-based data acquisition system 100 may be connected to the wireless terminal via a wired and / or wireless network, and transmit interactive information in accordance with an agreed data format.

[0069] Figure 6 Flowchart of the data collection method based on industrial Internet of Things according to an embodiment of the present application. Figure 6As shown, according to the data collection method based on industrial Internet of Things according to the embodiment of the present application, it includes: S110, obtaining current COV data, and obtaining a historical data set of COV data; S120, performing time series feature extraction on the current COV data and the historical data set of COV data to obtain a set of current COV time series associated implicit features and historical COV time series associated implicit features; S130, performing dynamic feature force field aggregation on the set of historical COV time series associated implicit features to obtain a historical COV time series associated significant aggregation representation; S140, determining whether to delete the current COV data based on the semantic difference measure between the historical COV time series associated significant aggregation representation and the current COV time series associated implicit features.

[0070] The specific operations of each step in the above-mentioned data collection method based on industrial Internet of Things have been referenced above. Figures 1 to 5 The invention has been introduced in detail in the description of the industrial Internet of Things based data acquisition system, and therefore, its repeated description will be omitted.

[0071] The embodiments of the present invention also relate to a storage medium storing a computer program. The computer program is executed by a processor to implement the above-mentioned system embodiment. That is, those skilled in the art can understand that all or part of the implementation of the above-mentioned embodiment system can be completed by instructing the relevant hardware through a program, and the program is stored in a storage medium, including a number of instructions to enable a device (which can be a single-chip microcomputer, chip, etc.) or a processor (processor) to execute all or part of the steps of each embodiment method of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program code.

[0072] The above descriptions of various implementations of the present disclosure are exemplary and not exhaustive. The disclosure is not limited to the disclosed implementations, and many modifications and variations are obvious to those of ordinary skill in the art without departing from the scope and spirit of the described implementations. The terms used herein are selected to best explain the principles of the implementations, practical applications, or improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the various practical methods disclosed herein.

Claims

1. A data acquisition system based on industrial Internet of Things, comprising: Industrial equipment, equipment access module, protocol conversion module, edge data processing module and industrial Internet platform, characterized in that: The device access module is used to access the industrial device; the protocol conversion module is used to convert the data collected by the industrial device into COV data in a standardized format; The edge data processing module is used to preprocess, store and analyze the COV data, and transmit the analyzed COV data to the industrial Internet platform; wherein the edge data processing module includes: a COV data preprocessing unit, used to preprocess the COV data to obtain preprocessed COV data; a COV data storage and analysis unit, used to store and analyze the preprocessed COV data to obtain analyzed COV data; a COV data transmission unit, used to transmit the analyzed COV data to the industrial Internet platform through a wireless communication module; The industrial Internet platform is used to number and classify the data uploaded by the edge data processing module; Wherein, the COV data preprocessing unit includes: a data acquisition subunit, which is used to acquire the current COV data and the historical data set of COV data; a data feature extraction subunit, which is used to perform time series feature extraction on the current COV data and the historical data set of COV data to obtain a set of current COV time series associated implicit features and historical COV time series associated implicit features; a historical data association aggregation subunit, which is used to perform dynamic feature force field aggregation on the set of historical COV time series associated implicit features to obtain a historical COV time series associated significant aggregation representation; a data deletion subunit, which is used to determine whether to delete the current COV data based on the semantic difference measure between the historical COV time series associated significant aggregation representation and the current COV time series associated implicit features; The data feature extraction subunit comprises: A current COV data obtaining time series association secondary subunit is used to extract a current COV time series association implicit feature vector from the current COV data; A historical COV data temporal association secondary subunit is used to extract temporal features from each COV data in the historical data set of the COV data to obtain a set of historical COV temporal association implicit feature vectors; The current COV data obtains a time series associated secondary subunit, which is used to process the current COV data using a time series extractor based on an RNN model to obtain the current COV time series associated implicit feature vector; The historical COV data obtaining time series association secondary subunit is used to process each COV data in the historical data set of the COV data using the RNN model-based time series extractor to obtain a set of the historical COV time series association implicit feature vectors; The historical data association aggregation subunit includes: The historical COV data static factor calculation clustering secondary subunit is used to calculate the static energy factor of each historical COV time series associated implicit feature vector in the set of historical COV time series associated implicit feature vectors to obtain a set of historical COV static energy factors, and based on the set of historical COV static energy factors, obtain the historical COV clustering initial center vector; The secondary subunit of energy force field aggregation of historical COV data is used to perform energy force field-based feature aggregation on the set of historical COV temporal association implicit feature vectors based on the set of historical COV static energy factors and the historical COV clustering initial center vector to obtain a significant aggregation representation of the historical COV temporal association; The historical COV data static factor calculation clustering secondary subunit includes: The third-level sub-unit for calculating the static energy factor of historical COV data is used to calculate the static energy factor of each historical COV time series associated implicit feature vector in the set of the historical COV time series associated implicit feature vectors to obtain the set of the historical COV static energy factors; The historical COV data clustering third-level sub-unit is used to select the historical COV time series associated implicit feature vector corresponding to the largest historical COV static energy factor in the set of historical COV static energy factors as the initial center vector of the historical COV clustering; The historical COV data static energy factor calculation three-level subunit is used to: Calculate the mean and variance of the historical COV time series associated implicit feature vector to obtain the historical COV time series mean and the historical COV time series variance; Subtracting the historical COV time series associated implicit feature vector from the historical COV time series mean by position, and calculating the fourth power of each position of the feature vector after subtraction to obtain the historical COV time series differential modulation vector; Calculating the expected value of the historical COV timing differential modulation vector to obtain the historical COV timing expected value; Dividing the historical COV time series expected value by the square of the historical COV time series variance, and inputting the value obtained by the division into a sigmoid function to obtain the historical COV static energy factor; The historical COV data energy field aggregation secondary subunit includes: The third-level sub-unit for calculating the dynamic aggregation energy factor of the historical COV data time series is used to calculate the dynamic aggregation energy factor of each historical COV time series associated implicit feature vector in the set of the historical COV time series associated implicit feature vectors based on the spatial span between each historical COV time series associated implicit feature vector in the set of the historical COV time series associated implicit feature vectors and the historical COV cluster initial center vector and the static energy factor of each historical COV time series associated implicit feature vector and the static energy factor of the historical COV cluster initial center vector to obtain a set of historical COV time series dynamic aggregation energy factors; The historical COV time series dynamic aggregation energy factor gated mask tertiary subunit is used to input the set of historical COV time series dynamic aggregation energy factors into the gated mask unit to obtain a set of historical COV time series dynamic aggregation weight factors; The historical COV data aggregation tertiary subunit is used to calculate the weighted sum of the set of historical COV time series associated implicit feature vectors with the set of historical COV time series dynamic aggregation weight factors to obtain a historical COV time series associated significant aggregation representation vector as the historical COV time series associated significant aggregation representation; The three-level subunit for calculating the historical COV data time series dynamic aggregation energy factor is used to: Multiplying the static energy factor of the historical COV time series associated implicit feature vector and the static energy factor of the historical COV cluster initial center vector by the first weighting parameter to obtain a first historical COV time series dynamic aggregation energy factor; Multiplying the square of the spatial span between the historical COV time series associated implicit feature vector and the historical COV cluster initial center vector by a second weighting parameter to obtain a second historical COV time series dynamic aggregation energy factor; The first historical COV time series dynamic aggregation energy factor is divided by the second historical COV time series dynamic aggregation energy factor to obtain the historical COV time series dynamic aggregation energy factor.

2. The data acquisition system based on industrial Internet of Things according to claim 1 is characterized in that: The data deletion subunit is used for: Inputting the historical COV time series association significant aggregation representation vector and the current COV time series association implicit feature vector into a semantic difference measurement network to obtain the semantic difference measurement value; Based on the comparison between the semantic difference metric and a preset threshold, it is determined whether to delete the current COV data.

3. A data collection method based on industrial Internet of Things, performed by the data collection system based on industrial Internet of Things according to claim 1, characterized in that: include: Get the current COV data and get the historical data set of COV data; Performing time series feature extraction on the current COV data and the historical data set of the COV data to obtain a set of current COV time series associated implicit features and historical COV time series associated implicit features; Performing dynamic feature force field aggregation on the set of historical COV time series association implicit features to obtain a significant aggregation representation of historical COV time series association; Based on the semantic difference measure between the historical COV time series association significant aggregation representation and the current COV time series association implicit feature, determine whether to delete the current COV data.

4. A storage medium, characterized in that: include: Computer program instructions are stored in the storage medium, and when the computer program instructions are executed by the processor, the processor executes the data acquisition system based on the industrial Internet of Things according to any one of claims 1 to 2.

Images