A data acquisition method based on the Internet of Things

Abstract

This invention provides a data acquisition method based on the Internet of Things (IoT), relating to the field of IoT technology. The method includes: acquiring IoT data; performing a first processing on the IoT data based on an IoT data identification model to obtain a first processing result; wherein the first processing includes at least: data standardization and intelligent labeling; storing the first processing result; performing a second processing on the first processing result based on the IoT data identification model to obtain a second processing result; wherein the second processing includes at least: intelligent data desensitization, data hardening, and data identifiability processing; and performing data usage analysis and exchange on the second processing result. This invention utilizes an IoT data identification model to automate data processing, effectively identifying the meaning and status of data, improving the efficiency and accuracy of data acquisition, labeling, analysis, and processing, reducing manual intervention, lowering labor costs, reducing labeling errors and omissions, and enhancing the automation and intelligence level of data processing.

Description

Technical Field

[0001] This invention relates to the field of Internet of Things (IoT) technology, and in particular to a data acquisition method based on IoT. Background Technology

[0002] Currently, IoT technology can connect various sensors, devices, and machines to the network, collecting, transmitting, and storing large amounts of operational data in real time. With the rapid development of industrial internet technology, more and more factories and production lines are beginning to leverage IoT technology for data collection.

[0003] However, the following problems exist in the traditional data collection process:

[0004] 1. The data volume is huge, and the manual analysis and annotation work is labor-intensive and inefficient.

[0005] 2. The data annotation process lacks automation, which can easily lead to annotation errors or omissions.

[0006] 3. The inability to effectively identify the meaning and status of data leads to insufficient accuracy in subsequent data analysis and decision-making.

[0007] To address the aforementioned problems, this invention proposes a data acquisition method based on the Internet of Things. Summary of the Invention

[0008] One of the objectives of this invention is to provide a data acquisition method based on the Internet of Things (IoT). This method utilizes IoT data identification large-scale models to automatically process data, effectively identifying the meaning and status of the data. This improves the efficiency and accuracy of data acquisition, labeling, analysis, and processing, reduces manual intervention, lowers labor costs, reduces labeling errors and omissions, effectively solves the problems of inefficiency and low accuracy in traditional data acquisition, and enhances the automation and intelligence level of data processing.

[0009] This invention provides a data acquisition method based on the Internet of Things, comprising:

[0010] Collect IoT data;

[0011] Based on the large-scale model for IoT data identification, the IoT data undergoes a first processing step to obtain a first processing result; the first processing step includes at least: data standardization and intelligent labeling.

[0012] Store the first processing result;

[0013] Based on the IoT data identification big data model, the first processing result is subjected to a second processing to obtain a second processing result; wherein, the second processing includes at least: intelligent data desensitization, hardening, and identifiable processing;

[0014] The second processing result is used for data analysis and exchange.

[0015] Optionally, the collection of IoT data includes:

[0016] Collect IoT data from multiple devices using a data acquisition device;

[0017] The plurality of devices include at least: sensors, device monitors, and smart meters;

[0018] IoT data includes at least: temperature, humidity, pressure, current, rotation speed, device status, and running time.

[0019] Optionally, the data standardization steps include at least: noise removal, missing value filling, and data format standardization.

[0020] Optionally, the intelligent labeling steps include at least: labeling each data point with a corresponding data type label; labeling each data point with a corresponding status label based on the range to which the data value belongs or the judgment result; and labeling each data point with a collection timestamp label.

[0021] Optionally, storing the first processing result includes:

[0022] The first processing result is stored in a structured database.

[0023] Optionally, the data intelligent desensitization step includes at least: transforming or replacing sensitive data.

[0024] Optionally, the hardening step includes at least: enhancing the quality of the data.

[0025] Optionally, the identifiable processing steps include at least: identifying the type of data; determining whether the data is within a standard range; and analyzing whether the data exceeds the normal operating range of the device to identify risks.

[0026] Optionally, the data usage analysis and exchange of the second processing result includes:

[0027] Based on the second processing result, the factory can perform real-time analysis and decision support for multiple dimensions such as factory equipment status, production efficiency, and energy consumption.

[0028] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings.

[0029] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0030] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0031] Figure 1 This is a flowchart illustrating a data acquisition method based on the Internet of Things (IoT) in an embodiment of the present invention.

[0032] Figure 2 This is another flowchart illustrating a data acquisition method based on the Internet of Things in an embodiment of the present invention. Detailed Implementation

[0033] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0034] This invention provides a data acquisition method based on the Internet of Things, such as... Figure 1 As shown, it includes:

[0035] S1. Collect IoT data;

[0036] S2. Based on the IoT data identification model, perform first processing on the IoT data to obtain first processing results; wherein, the first processing includes at least: data standardization and intelligent labeling;

[0037] S3. Store the first processing result;

[0038] S4. Based on the IoT data identification big model, perform a second processing on the first processing result to obtain a second processing result; wherein, the second processing includes at least: intelligent data desensitization, hardening, and identifiable processing;

[0039] S5. Perform data usage analysis and exchange on the second processing result.

[0040] The collection of IoT data includes:

[0041] Collect IoT data from multiple devices using a data acquisition device;

[0042] The plurality of devices include at least: sensors, device monitors, and smart meters;

[0043] IoT data includes at least: temperature, humidity, pressure, current, rotation speed, device status, and running time.

[0044] The data standardization steps include at least: noise removal, missing value filling, and data format standardization.

[0045] The intelligent labeling steps include at least: labeling each data point with a corresponding data type label; labeling each data point with a corresponding status label based on the range to which the data value belongs or the judgment result; and labeling each data point with a collection timestamp label.

[0046] The steps of intelligent data desensitization include at least: converting or replacing sensitive data.

[0047] The reinforcement steps include at least: enhancing the quality of the data.

[0048] The steps for identifiable processing include at least: identifying the type of data; determining whether the data is within a standard range; and analyzing whether the data exceeds the normal operating range of the device to identify risks.

[0049] The working principle and beneficial effects of the above technical solution are as follows:

[0050] like Figure 2 As shown, IoT data is collected from multiple devices via data acquisition devices. These devices can include various sensors, equipment monitors, and smart meters, which are used to monitor equipment status and environmental parameters and collect data in real time. Among the collected IoT data, temperature, humidity, and pressure can be used for environmental monitoring to help determine whether the equipment is operating in an optimal environment. Current, speed, equipment status, and operating time reflect the operating status of the equipment itself and can help evaluate the equipment's performance and work cycle.

[0051] The large-scale IoT data identification model is trained using machine learning algorithms and a large number of data standardization rules (which convert raw data into a unified standard format to ensure consistency in numerical range, units, etc.) as training samples. These include: labeled test IoT data (historically collected IoT data is used as test IoT data and labeled by data experts); data labeling rules (labeling data type, device status reflected in the data, data collection time, etc.); data anonymization rules (removing or encrypting sensitive information to ensure the privacy of users or devices is not leaked, e.g., hashing rules); data hardening rules (improving model generalization ability by enhancing data quality, avoiding overfitting or inefficient learning due to data quality issues, e.g., data cleaning, outlier detection and removal); and data identifiability processing rules (identifying data type, data range, etc.). Therefore, the large-scale IoT data identification model can autonomously and automatically complete the first and second processing steps mentioned above.

[0052] When removing noise, filtering techniques (such as mean filtering and median filtering) can be used to remove irrelevant noise and ensure data accuracy. When filling in missing values, interpolation and regression methods can be used to fill in missing data and ensure data integrity. When standardizing data formats, unified data standards and format normalization processes ensure that the data can be used consistently in subsequent steps.

[0053] When labeling data types, assign a type (e.g., temperature, humidity, equipment status) to each data point based on its data source. When labeling status, assign a status label to each data point based on whether the data value falls within a specific range; for example, label "high temperature" when the temperature exceeds a set threshold. When labeling timestamps, all data points must be stamped with their collection timestamps to ensure the data has a temporal sequence and facilitates maintaining its timeliness for subsequent analysis.

[0054] When intelligently desensitizing data, encryption, replacement, or anonymization can be used to avoid privacy leaks. For example, a device's unique identifier can be replaced with a hash value or encrypted data to protect the device's identity from direct exposure. During hardening, techniques such as data cleaning can be used to ensure high reliability of the final data. Improved data quality ensures more accurate analysis results and reduces misjudgments in subsequent analyses. During identifiable processing, the type of each data item is automatically identified (e.g., device status data, environmental monitoring data, etc.). Based on standard ranges, it is determined whether the data conforms to expectations; if the data exceeds the standard range, it may indicate an anomaly or malfunction in the device. Analysis of whether the data exceeds the device's normal operating range is performed, and potential risks are assessed through data pattern recognition. For example, abnormal fluctuations in current, speed, etc., may indicate a risk of device failure.

[0055] This invention utilizes IoT data identification large-scale models to automate data processing, effectively identifying the meaning and status of data, improving the efficiency and accuracy of data collection, labeling, analysis and processing, reducing manual intervention, lowering labor costs, reducing labeling errors and omissions, effectively solving the problems of inefficiency and low accuracy in traditional data collection, and improving the level of automation and intelligence in data processing.

[0056] In one embodiment, storing the first processing result includes:

[0057] The first processing result is stored in a structured database.

[0058] The initial processing results (standardized and labeled data) are stored in a structured database. Structured databases use relational databases (such as MySQL and PostgreSQL) or NoSQL databases (such as MongoDB) to store this pre-processed IoT data. Relational databases are well-suited for processing structured data, enabling efficient data querying and updating. Each data point is stored as a record based on its data type, status tag, timestamp, and other information. Structured databases allow for fast data querying, indexing, and updating, ensuring data persistence and accessibility.

[0059] In one embodiment, the data usage analysis and exchange of the second processing result includes:

[0060] Based on the second processing result, the factory can perform real-time analysis and decision support for multiple dimensions such as factory equipment status, production efficiency, and energy consumption.

[0061] The results of the second processing step are used for data analysis and exchange to support decision-making processes in the factory or other business systems. For example, the results of the second processing step can be used to support the factory in performing the following multi-dimensional analyses:

[0062] Equipment status analysis: Real-time monitoring of equipment operating status, such as temperature, speed, current, etc., to determine the health status of the equipment and perform fault prediction or early warning.

[0063] Production efficiency analysis: By analyzing data such as equipment uptime and production cycle, production processes can be optimized to improve production efficiency.

[0064] Energy consumption analysis: Based on the power consumption data of equipment, it helps enterprises monitor energy usage, optimize energy management strategies, and reduce operating costs.

[0065] For example, the data exchange can be performed using the results of the second processing step:

[0066] Exchange the analysis results with other systems or platforms, such as sharing data with enterprise management systems, cloud platforms, and decision support systems.

[0067] In specific IoT data acquisition scenarios (e.g., factories), the large area often leads to incomplete and insufficient data collection. Therefore, more comprehensive and in-depth data acquisition is needed. To replace manual labor, mobile data acquisition devices (e.g., drones equipped with cameras, RFID readers, Bluetooth readers, etc.) can be used for supplementary mobile data collection. Prior to this, to ensure proper implementation of supplementary mobile data collection and to guarantee the rapid acquisition of relevant data (to ensure quick understanding of device status), managers need to quickly plan how to conduct supplementary mobile data collection. However, IoT data acquisition scenarios are very large, and the workload for managers is enormous, making it impossible to meet this requirement and potentially leading to incomplete or untimely planning. Therefore, in one embodiment, the IoT-based data acquisition method further includes:

[0068] Based on the second processing result, a data interaction map is generated;

[0069] Based on the data interaction map, assist user groups in making decisions on mobile supplementary data collection needs;

[0070] Based on the mobile data collection needs, a control auxiliary network is built in the mobile data collection control model;

[0071] Based on the control-assisted network, the user group is assisted in making decisions on mobile supplementary sampling strategies in the mobile supplementary sampling control model;

[0072] Implement a mobile data collection strategy to collect new IoT data on mobile devices.

[0073] The user group includes multiple administrators. The mobile supplementary data collection requirement represents the initial need for mobile supplementary data collection in the IoT data acquisition scenario, and the data interaction graph can assist users in making this decision. The mobile supplementary data collection model is a three-dimensional simulation model of the IoT data acquisition scenario. The mobile supplementary data collection strategy is the final strategy to be executed by the user group to conduct mobile supplementary data collection in the IoT data acquisition scenario, and the control auxiliary network can assist users in making this strategy decision. Finally, the mobile supplementary data collection strategy is executed to collect new IoT data mobilely.

[0074] By generating a data interaction graph, the system assists user groups in making decisions about mobile supplementary data collection needs. Based on this graph, a control assistance network is built to further assist user groups. The system then assists user groups in making decisions about mobile supplementary data collection strategies within the mobile supplementary data collection control model. This greatly improves the efficiency of managers in arranging how to conduct mobile supplementary data collection, thus meeting the need for managers to arrange how to conduct mobile supplementary data collection as quickly as possible.

[0075] In one embodiment, generating a data interaction map based on the second processing result includes:

[0076] The second processing result is then subjected to clustering to obtain multiple first clusters;

[0077] Based on the importance analysis library, the importance of each first cluster is determined;

[0078] Sort each first cluster in descending order of importance to obtain the first cluster sequence;

[0079] Adaptive content is added to each i-th first cluster in the first cluster sequence to obtain the second cluster sequence; where K≤i≤N, the sum of the importance of the first K-1 first clusters in the first cluster sequence is closest to the importance and threshold;

[0080] The second cluster sequence is divided into multiple local sequences; among them, there are standard association relationships between the adaptive contents added to each pair of second clusters in the same local sequence.

[0081] Based on the display configuration rules, the template graph is configured to obtain a data interaction graph; wherein, the display configuration rules include: the j-th graph level in the template graph displays the contents of all second clusters in the j-th local sequence of the second cluster sequence;

[0082] The steps for determining adaptive content are as follows:

[0083] The clustering category of the i-th first cluster is used as the association target;

[0084] Use the importance of the i-th first cluster as the target threshold;

[0085] A third cluster is determined from the first cluster sequence; wherein, there is a correlation between the cluster category of the third cluster and the associated target, and the difference between the importance of the third cluster and the target threshold does not exceed a first difference threshold;

[0086] Feature extraction is performed on the i-th first cluster and the third cluster to obtain the feature set;

[0087] Based on the feature set, generate effective content selection rules;

[0088] Based on the effective content selection rules, effective content is selected from the multimodal information of user groups;

[0089] Based on the effective content, determine the adaptive content.

[0090] When performing clustering on the second processing result, clustering is performed according to multiple clustering categories to obtain multiple first clusters. The first clusters contain data of the same clustering category in the second processing result, i.e., the cluster content. The clustering categories include at least: data belonging to the same device, data belonging to the same device area, data belonging to the same device with the same environmental conditions, and data generated in the same time period.

[0091] The importance analysis library contains the importance of different first clusters. The importance represents the degree to which the content in the first cluster needs to be viewed by the user group. For example, if the content in the cluster belongs to the same device, it is convenient for the user group to comprehensively analyze the situation of the device, and the corresponding importance is 8. The importance corresponding to different first clusters can also be set in advance by technical personnel.

[0092] The first clusters are sorted from highest to lowest importance to form a first cluster sequence, ensuring that content from clusters prioritized by the user group is displayed at the top of the list. The importance and threshold can be set to 30, constraining the sum of the importance of the first K-1 first clusters in the first cluster sequence to be as close as possible to the importance and threshold. This ensures that K has a unique value. The fact that the sum of the importance of the first K-1 first clusters is closest to the importance and threshold indicates that displaying the content from each of the first K-1 first clusters to the user group has essentially conveyed the important information. Starting from the Kth first cluster, personalized adaptive content can be added.

[0093] The standard association between adaptive content refers to the association between adaptive content and each other, such as: content of interest belonging to the same expert in the same expert group, or content belonging to the management authority scope of different experts in the same expert group; setting the constraint that there is a standard association between the adaptive content added to each pair of second clusters in the same local sequence divides the second cluster sequence into multiple local sequences.

[0094] The template graph serves as a template for generating the data interaction graph. It contains subgraphs at multiple graph levels. Subgraphs at different graph levels can display different content separately, with smaller subgraphs displaying their content first. After generating the data interaction graph in this way, user groups can view the content in the corresponding clusters sequentially from lower graph levels to higher graph levels.

[0095] When determining adaptive content, it is necessary to consider the common suitability between the current adaptive content output and the adaptive content output shortly thereafter. Therefore, the association target and the target threshold are determined separately to identify the third cluster. The association between the cluster category and the management target means that the cluster categories are similar and the data from different cluster categories can work together to play a synergistic role in the user group's decision on mobile supplementary data collection needs. The first difference threshold can be set to 5 to constrain the difference between the importance of the third cluster and the target threshold to not exceed the first difference threshold, so that the third cluster is the first cluster not far after the i-th first cluster.

[0096] The feature sets of the i-th first cluster and the third cluster are extracted. The features in the feature set include at least: the content type in each cluster and the data relationship between the content in each cluster. The generated effective content selection rule refers to the rule of selecting effective content reflected by the features in the feature set that can be used as the basis for determining adaptive content. For example, if the feature in the feature set is the content type of a device to be managed in a cluster, then the corresponding effective content selection rule is to select the historical managed situation of the user group as the effective content and add it to the i-th first cluster. When the user group sees the content in the cluster at this map level, they can compare it with their own historical managed situation and decide whether they want to learn from or refer to it in order to determine the mobile supplementary collection needs.

[0097] Through multi-level clustering and adaptive content supplementation strategies, a data interaction graph tailored to user needs can be efficiently generated based on the clustering categories and importance of the data. First, the data is clustered into multiple primary clusters, and then sorted according to the importance of the content within each cluster, ensuring that the most important content is displayed first. This not only optimizes the data display order but also, through adaptive content supplementation, allows users to obtain personalized recommendations based on historical behavior and management needs when viewing data. The combination of adaptive content determination and clustering analysis not only improves the relevance of information but also enhances the interactivity of the graph. Third clusters are accurately selected, generating feature sets. These feature sets are then used to generate effective content selection rules, providing users with more accurate and personalized information. By dividing the second cluster sequence into local sequences and displaying them at different graph levels, the system ensures the hierarchy and organization of information, facilitating on-demand browsing for users.

[0098] Data interaction graphs not only improve users' analytical efficiency but also enhance the intelligence of decision support systems, enabling users to extract information that is highly relevant to their needs from massive amounts of data more effectively. This improves the overall quality of decision-making and response speed, and greatly enhances the accuracy of their decision-making and mobile data acquisition needs.

[0099] In one embodiment, based on mobile data acquisition needs, a control-aided network is built in the mobile data acquisition control model, including:

[0100] Fuzzy logic parsing is performed on the mobile supplementary data collection demand to obtain a fuzzy logic set;

[0101] Based on fuzzy logic sets, multiple mobile sampling routes are marked in the mobile sampling control model;

[0102] Generate line-of-sight preprocessing rules and route triggering rules for each mobile sampling route;

[0103] Based on the line-of-sight preprocessing rules and route triggering rules for each mobile sampling route, the control auxiliary network is determined;

[0104] The gaze preprocessing rules include:

[0105] If it is determined whether more than a certain number of first trajectory points are continuously generated on the target trajectory to be preprocessed, and the straight-line distance between the starting position of the moving supplementary sampling route and the first trajectory point does not exceed a certain distance threshold, then the part of the trajectory before the last first trajectory point on the target trajectory is removed.

[0106] The route triggering rules include:

[0107] Analyze the multiple sets of one-to-one corresponding second trajectory points and their generation times after preprocessing the target trajectory;

[0108] The vertical spacing between the second trajectory point and the moving and supplementary sampling route at each generation time and at different generation times is mapped into the spacing-time curve to obtain the target curve;

[0109] Extract multiple trough and peak values ​​from the target curve;

[0110] When the difference between each pair of trough and peak values ​​does not exceed the second difference threshold, the mobile replenishment route will be used as the basis for mobile replenishment strategy decision-making.

[0111] Mobile supplementary data collection needs reflect fuzzy logic. Fuzzy logic refers to the arrangement logic for mobile supplementary data collection initially represented by the mobile supplementary data collection needs. For example, if the mobile supplementary data collection needs refer to collecting the on-site working images of a certain type of equipment in a certain area, then the fuzzy logic is to collect the working images of each of these devices in sequence. Fuzzy logic parsing of mobile supplementary data collection needs forms a fuzzy logic set.

[0112] When marking the mobile data collection route, determine the mutation logic of the fuzzy logic in the fuzzy logic set (the mutation logic is the arrangement logic after the user group will suddenly change from the fuzzy logic, such as: moving to take pictures of the on-site work of the staff around them), determine the corresponding data collection area of ​​the fuzzy logic and mutation logic in the mobile data collection control model, that is, the on-site area that needs to be data collection indicated by each logic, and finally, connect the centers of the two on-site areas to obtain the mobile data collection route.

[0113] The line-of-sight preprocessing rules are rules for preprocessing the viewing trajectories generated by each user in the user group within a recently preset time period, which can be 100 seconds. The route triggering rules are rules for analyzing whether the preprocessed viewing trajectories trigger mobile supplementary acquisition routes. Ultimately, the line-of-sight preprocessing rules and route triggering rules for each mobile supplementary acquisition route form a control auxiliary network.

[0114] Accordingly, the method of assisting user groups in making decisions on mobile data acquisition strategies in the mobile acquisition control model based on the control-aided network includes:

[0115] Display the mobile data acquisition control model after the control-assisted network is built to the user group;

[0116] Get the viewing trajectory of each user in the user group within the most recent preset time period;

[0117] Based on the line-of-sight preprocessing rules, the viewing line-of-sight trajectory is preprocessed; wherein, the viewing line-of-sight trajectory is used as the target trajectory to be preprocessed.

[0118] Based on the route triggering rules, determine whether to trigger the mobile sampling route according to the preprocessed viewing line trajectory;

[0119] When triggered, the triggered mobile sampling route is simultaneously displayed to each user in the user group, allowing each user group to decide which mobile sampling route to ultimately adopt as the mobile sampling strategy.

[0120] The number threshold can be 3; the spacing threshold can be 5 cm; when more than the number of first trajectory points with a straight-line spacing of no more than the spacing threshold are continuously generated on the target trajectory to be preprocessed, it indicates that the user wants to trigger the mobile sampling route and remove the part of the trajectory before the last first trajectory point on the target trajectory.

[0121] The horizontal axis of the spacing-time curve represents time, and the vertical axis represents the spacing distance. The vertical spacing between the corresponding second trajectory point and the moving replenishment route at each generation time and at different generation times is mapped into the spacing-time curve to obtain the target curve.

[0122] The second difference threshold can be 3 centimeters. When the difference between any two troughs and peaks does not exceed the second difference threshold, it indicates that the user intends to trigger a mobile data supplementation route, using this route as the basis for mobile data supplementation strategy decisions. In practical applications, users can view the control auxiliary network. When scheduling mobile data supplementation, users should focus their gaze on the starting position of a mobile data supplementation route, and then look towards its end. The system will then automatically trigger the corresponding mobile data supplementation route.

[0123] By constructing a control-assisted network based on fuzzy logic parsing, intelligent support is provided for mobile data acquisition strategy decision-making. First, fuzzy logic parsing accurately identifies user data acquisition needs and marks multiple mobile data acquisition routes accordingly, ensuring precise matching between needs and strategies. Line-of-sight preprocessing rules and route triggering rules, through intelligent analysis of user viewing trajectories, enable the system to respond to user behavior in real time and automatically trigger relevant data acquisition routes, greatly improving the system's automation level and user experience. Precise mapping of the spacing and time between trajectory points ensures high reliability of route triggering, while peak and trough analysis further optimizes the decision-making basis for determining triggering conditions. In practical applications, users only need to focus on the starting position of a specific data acquisition route, and the system will automatically identify and trigger the corresponding mobile data acquisition route based on their line-of-sight trajectory, reducing human intervention and improving work efficiency and accuracy. Overall, this system not only effectively optimizes the selection of data acquisition routes but also dynamically adjusts data acquisition strategies through intelligent analysis, exhibiting high adaptability and foresight, providing strong technical support for the implementation of mobile data acquisition.

[0124] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A data acquisition method based on the Internet of Things, characterized in that, include: Collect IoT data; Based on the large-scale model for IoT data identification, the IoT data undergoes a first processing step to obtain a first processing result; the first processing step includes at least: data standardization and intelligent labeling. Store the first processing result; Based on the IoT data identification big data model, the first processing result is subjected to a second processing to obtain a second processing result; wherein, the second processing includes at least: intelligent data desensitization, hardening, and identifiable processing; The second processing result is used for data analysis and exchange; Based on the second processing result, a data interaction map is generated; Based on the data interaction map, assist user groups in making decisions on mobile supplementary data collection needs; Based on the mobile data collection needs, a control auxiliary network is built in the mobile data collection control model; Based on the control-assisted network, the user group is assisted in making decisions on mobile supplementary sampling strategies in the mobile supplementary sampling control model; Implement a mobile supplementary data collection strategy to collect new IoT data on a mobile basis; The generation of the data interaction map based on the second processing result includes: The second processing result is then subjected to clustering to obtain multiple first clusters; Based on the importance analysis library, the importance of each first cluster is determined; Sort each first cluster in descending order of importance to obtain the first cluster sequence; Adaptive content is added to each i-th first cluster in the first cluster sequence to obtain the second cluster sequence; where K≤i≤N, the sum of the importance of the first K-1 first clusters in the first cluster sequence is closest to the importance and threshold; The second cluster sequence is divided into multiple local sequences; among them, there are standard association relationships between the adaptive contents added to each pair of second clusters in the same local sequence. Based on the display configuration rules, the template graph is configured to obtain a data interaction graph; wherein, the display configuration rules include: the j-th graph level in the template graph displays the contents of all second clusters in the j-th local sequence of the second cluster sequence; The steps for determining adaptive content are as follows: The clustering category of the i-th first cluster is used as the association target; Use the importance of the i-th first cluster as the target threshold; A third cluster is determined from the first cluster sequence; wherein, there is a correlation between the cluster category of the third cluster and the associated target, and the difference between the importance of the third cluster and the target threshold does not exceed a first difference threshold; Feature extraction is performed on the i-th first cluster and the third cluster to obtain the feature set; Based on the feature set, generate effective content selection rules; Based on the effective content selection rules, effective content is selected from the multimodal information of user groups; Based on the effective content, determine the adaptive content.

2. The data acquisition method based on the Internet of Things as described in claim 1, characterized in that, The collection of IoT data includes: Collect IoT data from multiple devices using a data acquisition device; The plurality of devices include at least: sensors, device monitors, and smart meters; IoT data includes at least: temperature, humidity, pressure, current, rotation speed, device status, and running time.

3. The data acquisition method based on the Internet of Things as described in claim 1, characterized in that, The data standardization steps include at least: noise removal, missing value filling, and data format standardization.

4. The data acquisition method based on the Internet of Things as described in claim 1, characterized in that, The intelligent labeling steps include at least: labeling each data point with a corresponding data type label; labeling each data point with a corresponding status label based on the range to which the data value belongs or the judgment result; and labeling each data point with a collection timestamp label.

5. The data acquisition method based on the Internet of Things as described in claim 1, characterized in that, The storage of the first processing result includes: The first processing result is stored in a structured database.

6. The data acquisition method based on the Internet of Things as described in claim 1, characterized in that, The steps of intelligent data desensitization include at least: converting or replacing sensitive data.

7. The data acquisition method based on the Internet of Things as described in claim 1, characterized in that, The reinforcement steps include at least: enhancing the quality of the data.

8. The data acquisition method based on the Internet of Things as described in claim 1, characterized in that, The steps for identifiable processing include at least: identifying the type of data; determining whether the data is within a standard range; and analyzing whether the data exceeds the normal operating range of the device to identify risks.

9. The data acquisition method based on the Internet of Things as described in claim 1, characterized in that, The data usage analysis and exchange of the second processing result includes: Based on the second processing result, the factory can perform real-time analysis and decision support for multiple dimensions such as factory equipment status, production efficiency, and energy consumption.

Images