Internet of things asset positioning system based on identification analysis system

Abstract

The invention discloses an Internet of Things asset positioning system based on an identifier analysis system, relates to the technical field of network communication and positioning, and solves the problem that an existing system does not support accurate positioning and management in cross-enterprise and cross-region circulation and exchange processes of Internet of Things assets. The asset positioning system comprises an equipment communication end and an asset positioning end. The equipment communication end comprises an active identification carrier module and an information transmission module; the asset positioning end comprises an identification analysis docking module, a data analysis processing module, an asset positioning service module and a display early warning module; detecting and backing up the acquired real-time data of the equipment through an anomaly detection model, and positioning all the equipment through an asset positioning service module; acquiring real-time data of equipment and generating an equipment identification code through the active identification carrier module; the data operation load of the identification analysis server is reduced, the efficient and stable operation of the asset positioning system is ensured, and the safety and reliability of all equipment data in an enterprise are ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of network communication and positioning, and in particular, to an Internet of Things asset positioning system based on an identification and resolution system. Background Art

[0002] At present, with the rapid development of the industrial chain and supply chain of enterprise production, the exchange of assets between enterprises is becoming increasingly frequent. There are technical problems and data interoperability problems in the existing Internet of Things asset positioning systems, and the effective positioning of Internet of Things assets cannot be achieved. Currently, the mainstream Internet of Things asset positioning systems are dominated by foreign companies (such as Google and IBM), lacking the in-depth application of China's independent and controllable identification and resolution system. In order to meet the positioning and security management requirements of Internet of Things assets across enterprises and regions, and at the same time promote the application of the identification and resolution system in the field of positioning systems, it is necessary to develop an Internet of Things asset positioning system based on the identification and resolution system.

[0003] Patent No. CN202110060408.8 discloses an automatic positioning device electronic identification plate based on Internet of Things technology and its use method. The former includes a hardware device and a software processing system. The hardware device includes an RFID module or an Internet of Things card module, a power module, a display device, and an ARM processing unit. The power module is connected to the ARM processing unit, and the RFID module or the Internet of Things card module is connected to the ARM processing unit. When using the above-mentioned electronic identification plate, during the power asset inventory, the electronic tag information on the device can be quickly read, and the read tag information is sent to the background server for processing through networks such as 3G and 4G; in case of emergencies, the RFID handheld device sends a signal of abnormal conditions to the background in a timely manner according to the device positioning information, so as to be able to respond quickly and process in a timely manner.

[0004] Patent No. CN202411031122.7 discloses a high-precision Internet of Things positioning method and device based on the Beidou satellite system. The method includes: receiving a positioning request from an Internet of Things terminal to understand information such as the terminal identifier, request accuracy, and request frequency; querying a preset terminal configuration database according to the terminal identifier to determine the positioning strategy template corresponding to the Internet of Things terminal, providing a basis for customized Internet of Things positioning; determining the positioning algorithm and parameter configuration according to the positioning strategy template to achieve customized positioning of the Internet of Things terminal, improving flexibility and practicality; performing customized positioning on the Internet of Things terminal according to the request accuracy, request frequency, and the selected positioning algorithm and parameter configuration. It can improve the problem that in the process of positioning Internet of Things terminals, due to ignoring the different requirements of different Internet of Things terminals for positioning accuracy and frequency, the positioning service cannot accurately match the actual needs, and the positioning flexibility is not high.

[0005] However, the above patents and existing systems do not support the accurate positioning and management of IoT assets during the cross-enterprise and cross-regional transfer and exchange processes, cannot achieve the transfer and sharing of IoT assets, cannot achieve the traceability and transparency of the supply chain, the management efficiency of IoT assets is relatively low, enterprises cannot detail the location and status of IoT assets at any time, cannot achieve the fine monitoring and scheduling of IoT assets, resulting in obvious idle and waste phenomena of assets, which is not conducive to cooperation and collaboration among enterprises. Summary of the Invention

[0006] The purpose of the present invention is to provide an IoT asset positioning system based on an identification and resolution system, which can carry identification codes and actively establish communication connection capabilities and active information sharing for different types of devices through an active identification carrier module and an asset positioning service module, and collect and abnormally locate the real-time data of the devices, so as to reduce the data operation load of the identification and resolution server, ensure the efficient and stable operation of the asset positioning system, and guarantee the security and reliability of all device data in the enterprise.

[0007] The present invention utilizes the following technical solutions:

[0008] An IoT asset positioning system based on an identification and resolution system, comprising a device communication end and an asset positioning end; wherein,

[0009] The device communication end is used to collect real-time device data and generate device identification codes through an active identification carrier module, and use different communication protocols to connect to and transmit data with the asset positioning end and several devices through an information transmission module;

[0010] The asset positioning end is used to parse the device identification codes by combining with the identification and resolution system through an identification and resolution docking module, detect and back up the collected real-time device data through a data analysis and processing module using an anomaly detection model, locate all devices through an asset positioning service module, and display the device status and anomaly alarm notifications through a display and warning module.

[0011] Preferably, the information transmission module connects to the asset positioning terminal and several devices using different communication protocols; the active identification carrier module applies for an enterprise node prefix based on enterprise authentication information; the enterprise authentication information includes the company organization code, enterprise name, enterprise address, enterprise type, and legal representative; at the same time, the factory list of all enterprises is obtained by combining the enterprise name and address with the enterprise type; and the devices in all factories of the enterprise are clustered to obtain the device list of all factories; the devices are parsed to obtain all parts of each device, resulting in a device part table; at the same time, the operating data of the devices is collected using a sensor network to obtain device real-time data; and the device real-time data is subjected to asymmetric encryption and decryption operations, and then loaded into the information transmission module for transmission to the asset positioning terminal; the devices include Internet of Things devices and positioning devices; the device real-time data includes the deployment location, security status, operation log, and remote control data.

[0012] Preferably, the identification and resolution docking module generates a factory code by combining an enterprise node with the factory list and enterprise node prefix according to the identification and resolution system; at the same time, the device category code and device code are sequentially generated according to the device list and factory code; at the same time, the code point codes in each part of the device are obtained by combining the device part table with the device real-time data; finally, the enterprise node prefix, company organization code, factory code, device category code, device code, part code, and code point code are combined to obtain the device identification code, and then the identification code is issued for all devices; subsequently, the identification and resolution docking module aggregates the device real-time data of the same company according to the device identification code to obtain a device operation overview table, and at the same time, the device operation overview table is combined with the device real-time data and re-partitioned and aggregated according to the factory code to obtain a device management table. At the same time, the device management table of the same device is extended according to the device attributes and device code to obtain a device detailed list, thus completing the identification and resolution of the device real-time data; finally, the device operation overview table, device management table, and device detailed list are hierarchically nested using a knowledge graph to obtain an identification code graph, thus completing the identification registration of all devices and transmitting them to the recursive resolution node.

[0013] Preferably, the identification and resolution system includes a top-level node, a second-level node, an enterprise node, and a recursive resolution node; the operating mechanism of the identification and resolution system is as follows:

[0014] According to the enterprise node prefix in the recursive resolution node combined with the enterprise address, all device identification codes in the enterprise node are associated and allocated to obtain a regional identification list, which is classified as a comprehensive second-level node in the second-level node;

[0015] According to the industry to which the enterprise belongs and the enterprise scale, the regional identification list is divided to obtain an industry identification sub-list, which is classified as an industry-type second-level node in the second-level node;

[0016] The top-level node connects and parses the secondary nodes according to the docking application information and in combination with the access application protocol to obtain identification statistical data, metadata, and master data;

[0017] The top-level node divides the metadata and master data of the comprehensive secondary nodes and industry-type secondary nodes into several batches of synchronized data sets according to the identification statistical data in combination with the data synchronization channel;

[0018] The top-level node performs synchronous expansion and feedback update on the metadata in each batch of synchronized data sets according to the data attributes, and at the same time fills in the master data in a preset format according to the data type for full-volume and incremental synchronization;

[0019] The top-level node sends monitoring requests to all enterprise nodes and secondary nodes according to the preset parsing ports, and at the same time receives the security status and operation logs of all enterprise nodes and secondary nodes.

[0020] Preferably, the data analysis and processing module hierarchically expands the real-time device data of all devices according to the device identification code in combination with the device management table to obtain a device part association data set;

[0021] At the same time, according to the preset device part threshold set in combination with the device deployment environment information, the real-time device data is detected, judged, and stored and backed up:

[0022] If the acquisition interval of adjacent device real-time data is equal to the preset collection time difference, it is determined that the current device real-time data is valid data, and it is input into the anomaly detection model to perform data quality detection in combination with the device operation overview table to obtain an abnormal device data set and a normal device data set;

[0023] If the acquisition interval of adjacent device real-time data is not equal to the preset collection time difference, it is determined that the current device real-time data is suspicious data, and at the same time, the operation state of the device is checked:

[0024] If the operation state of the device is shutdown or standby state, the current suspicious data is marked and converted into valid data, and at the same time added to the normal device data set;

[0025] If the operation state of the device is normal operation state, the current suspicious data is marked and converted into invalid data, and at the same time, the invalid data is cleaned using a data cleaning algorithm, and a suspicious data set is constructed in combination with the acquisition time;

[0026] If the operation state of the device is abnormal operation state, the current suspicious data is marked and converted into invalid data, and at the same time added to the abnormal device data set.

[0027] Preferably, the working process of the anomaly detection model is as follows:

[0028] A: The environmental analysis layer of the anomaly detection model uses the reflection mapping algorithm to decompose and sample the device deployment environment information, obtaining an environmental stress mapping diagram;

[0029] B: The feature extraction layer of the anomaly detection model uses an improved ant colony algorithm to extract features from the abnormal device dataset and the normal device dataset according to the preset information extraction mechanism, obtaining an abnormal feature matrix and a normal feature matrix;

[0030] C: The feature association layer of the anomaly detection model uses the frequent pattern growth algorithm to respectively perform information association on the abnormal feature matrix and the normal feature matrix in combination with the environmental stress mapping diagram, obtaining an environment-device feature matrix;

[0031] D: The iterative training layer of the anomaly detection model uses the inertia-compressed particle swarm algorithm to perform several iterative trainings on the environment-device feature matrix in combination with the preset device part threshold set, obtaining a comprehensive monitoring weight matrix;

[0032] E: The prediction and update layer of the anomaly detection model uses an energy-non-saturated loss function to perform feedback optimization on the comprehensive monitoring weight matrix in combination with the device historical data and the device operation overview table, and then predicts and outputs abnormal device data and normal device data.

[0033] Preferably, the asset positioning service module hierarchically splits the abnormal device dataset and the normal device dataset according to the device identification code in accordance with the identification resolution coding rule and in combination with the device deployment environment information, thereby determining the abnormal parts of all devices. At the same time, it synchronously updates the metadata and master data in the enterprise node in combination with the device management table and the device details list, and fills the abnormal device dataset and the normal device dataset in the device management table and the device details list to obtain a device status table, and then completes the deployment location positioning and abnormal part positioning of all devices in combination with the positioning correction model.

[0034] Preferably, the working process of the positioning correction model is as follows:

[0035] The communication parsing layer of the positioning correction model parses the communication protocols of all devices in combination with the device distance and the signal attenuation value according to the device identification code in combination with the security status, obtaining a protocol-environment-distance communication table;

[0036] The positioning analysis layer of the positioning correction model analyzes the deployment locations of all devices according to the preset positioning accuracy threshold and the signal anti-interference value, obtaining a device deployment accuracy table;

[0037] The data judgment layer of the positioning correction model performs an extended judgment on the comprehensive monitoring weight matrix according to the protocol-environment-distance communication table in combination with the device deployment accuracy table, obtaining a positioning data expansion matrix;

[0038] The positioning correction layer of the positioning correction model corrects the deployment location and abnormal parts of the device according to the positioning data expansion matrix and the abnormal device data set, combines the identification coding map and the environmental impact factor, and obtains the positioning data of the optimal deployment location and abnormal parts.

[0039] Preferably, the display and warning module uses a web page to display the completed device management table and device details to the administrator in real time. At the same time, it uses notifications to alarm the abnormal parts of the device combined with the abnormal device data, and stores them in the asset positioning database in combination with the device status table.

[0040] The present invention detects and backs up the real-time data of the device collected by the anomaly detection model, and locates all devices through the asset positioning service module; collects the real-time data of the device and generates the device identification code through the active identification carrier module; combines the identification resolution system through the identification resolution docking module to resolve the device identification code; reduces the data operation load of the identification resolution server, ensures the efficient and stable operation of the asset positioning system, and guarantees the security and reliability of all device data in the enterprise. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or related technologies, the following will briefly introduce the drawings required for use in the description of the embodiments or related technologies. Obviously, the drawings in the following description are only embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained according to the provided drawings.

[0042] Figure 1 is the schematic diagram of the asset positioning system;

[0043] Figure 2 is the schematic diagram of the principle of the anomaly detection model;

[0044] Figure 3 is the flowchart of the asset positioning system;

[0045] Figure 4 is the schematic diagram of the principle of the identification resolution system docking module;

[0046] Figure 5 is the example diagram of the identification resolution coding rule. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0047] The following will describe the present invention in detail with reference to the drawings and embodiments:

[0048] As Figures 1 to 5 shown, an Internet of Things asset positioning system based on an identification resolution system described in the present invention includes a device communication end and an asset positioning end; wherein,

[0049] The device communication end is used to collect real-time device data through the active identification carrier module and generate device identification codes, and connect and transmit data with the asset positioning end and several devices through the information transmission module using different communication protocols;

[0050] The asset positioning end is used to parse the device identification code by combining the identification resolution system through the identification resolution docking module, detect and back up the collected real-time device data using the anomaly detection model through the data analysis and processing module, locate all devices through the asset positioning service module, and display the device status and anomaly alarm notifications through the display and warning module.

[0051] In this embodiment, the asset positioning end processes and stores the information actively sent by the device in real time. The collected device data is stored in the Greenplum database using message queues or real-time data processing technologies. By monitoring and analyzing the location and status information of the device in real time, the operation status of the Internet of Things device can be understood in a timely manner:

[0052] By deploying sensors and data collection devices on the device, the data is sent to the Greenplum database through real-time stream processing technologies (such as Apache Kafka, Apache Flink, etc.) to achieve real-time writing and storage of the data;

[0053] Using real-time stream processing technologies, the device data is processed and analyzed in real time. By setting up a real-time monitoring system, the data is aggregated, calculated, and analyzed in real time to timely detect changes and anomalies in the device status;

[0054] Through data visualization technologies, the real-time monitored device location and status information are presented to relevant personnel in an intuitive manner. At the same time, an alarm system is established to send alarm notifications to relevant personnel in a timely manner when abnormal device status is detected, so as to take actions in a timely manner;

[0055] Ensure real-time synchronization and backup of device data. The multi-copy replication mechanism of the Greenplum database is adopted to enable the data to be synchronously and backed up in real time in a distributed environment to ensure the high availability and reliability of the data.

[0056] In the present invention, the information transmission module is connected to the asset positioning terminal and several devices by using different communication protocols; the active identification carrier module applies for the enterprise node prefix according to the enterprise authentication information; the enterprise authentication information includes the company organization code, enterprise name, enterprise address, enterprise type, and legal representative; at the same time, the factory list of all enterprises is obtained according to the enterprise name and enterprise address combined with the enterprise type; and the devices in all factories of the enterprise are clustered to obtain the device list of all factories; the devices are parsed to obtain all parts of each device, and the device part table is obtained; at the same time, the operation data of the devices is collected by using the sensor network to obtain the device real-time data; and the asymmetric encryption and decryption operation is performed on the device real-time data, and then it is loaded into the information transmission module for transmission to the asset positioning terminal.

[0057] In this embodiment, the devices include Internet of Things devices and positioning devices;

[0058] The device real-time data includes the deployment location, security status, operation log, and remote control data;

[0059] The deployment location includes geographical location information such as longitude and latitude to accurately understand the location of the device;

[0060] The security status includes real-time monitoring data such as the operation status, working hours, output, temperature, pressure, and humidity of the device to timely understand the operation status and performance indicators of the device;

[0061] The operation log includes the power-on and power-off records, operation logs, alarm information, etc. of the device to track the operation and abnormal conditions of the device;

[0062] The remote control data includes remote control instructions and feedback information for the device to perform remote operation and regulation;

[0063] The active identification carrier module also collects the device real-time data through devices such as sensors, PLCs (programmable logic controllers), and SCADA systems (supervisory control and data acquisition systems);

[0064] The communication protocols include RFID, Beidou, 5G, WiFi, Bluetooth, and UWB, etc.;

[0065] In the present invention, the identification resolution docking module generates factory codes according to the identification resolution system by using enterprise nodes in combination with a factory list and an enterprise node prefix; at the same time, it generates equipment category codes and equipment codes in sequence according to the equipment list in combination with the factory codes; at the same time, it obtains the code point codes in each part of the equipment according to the equipment part table in combination with the real-time equipment data; finally, it combines the enterprise node prefix, company organization code, factory code, equipment category code, equipment code, part code, and code point code to obtain the equipment identification code, thereby completing the identification code issuance for all equipment; subsequently, the identification resolution docking module aggregates the real-time equipment data of the same company according to the equipment identification code to obtain an equipment operation overview table, and at the same time, it re-divides and aggregates the equipment operation overview table in combination with the real-time equipment data according to the factory code to obtain an equipment management table, and at the same time, it expands the equipment management table of the same equipment according to the equipment attributes in combination with the equipment code to obtain an equipment detailed list, thereby completing the identification resolution of the real-time equipment data; finally, it uses a knowledge graph to hierarchically nest the equipment operation overview table, the equipment management table, and the equipment detailed list to obtain an identification code graph, thereby completing the identification registration of all equipment and transmitting it to the recursive resolution node.

[0066] In this embodiment, the identification resolution system docking module uniformly abstracts and manages the Internet of Things asset identification, and provides services such as identification code issuance, identification resolution, and identification registration for various scenarios and industry applications.

[0067] In the present invention, the identification resolution system includes a top-level node, a secondary node, an enterprise node, and a recursive resolution node; the operation mechanism of the identification resolution system is as follows:

[0068] According to the enterprise node prefix in the recursive resolution node in combination with the enterprise address, all the equipment identification codes in the enterprise node are associated and allocated to obtain a regional identification list, which is classified into the comprehensive secondary node in the secondary node;

[0069] According to the enterprise's affiliated industry in combination with the enterprise scale, the regional identification list is divided to obtain an industry identification sub-list, which is classified into the industry-type secondary node in the secondary node;

[0070] The top-level node connects and resolves the secondary node according to the docking application information in combination with the access application protocol to obtain identification statistical data, metadata, and master data;

[0071] The top-level node divides the metadata and master data of the comprehensive secondary node and the industry-type secondary node into several batches of synchronous data sets according to the identification statistical data in combination with the data synchronization channel;

[0072] The top-level node synchronously expands and feedback-updates the metadata in each batch of synchronous data sets according to the data attributes, and at the same time fills in the master data in a preset format according to the data type for full-volume and incremental synchronization;

[0073] The top-level node sends monitoring requests to all enterprise nodes and secondary nodes according to the preset parsing port, and at the same time receives the security status and operation logs of all enterprise nodes and secondary nodes.

[0074] In the present invention, the data analysis and processing module hierarchically expands the real-time device data of all devices according to the device identification code and in combination with the device management table, and obtains a device part association data set;

[0075] In this embodiment, hierarchical expansion means disassembling complex information according to logical levels or modules.

[0076] At the same time, according to the preset device part threshold set and in combination with the device deployment environment information, the real-time device data is detected, judged, stored and backed up:

[0077] If the acquisition interval of adjacent device real-time data is equal to the preset collection time difference, it is determined that the current device real-time data is valid data, and it is input into the anomaly detection model and combined with the device operation overview table for data quality detection, obtaining an abnormal device data set and a normal device data set;

[0078] If the acquisition interval of adjacent device real-time data is not equal to the preset collection time difference, it is determined that the current device real-time data is suspicious data, and at the same time, the operation state of the device is inspected:

[0079] If the operation state of the device is shutdown or standby state, the current suspicious data is marked and converted into valid data, and at the same time, it is added to the normal device data set;

[0080] If the operation state of the device is normal operation state, the current suspicious data is marked and converted into invalid data, and at the same time, the invalid data is cleaned using a data cleaning algorithm, and a suspicious data set is constructed in combination with the acquisition time;

[0081] If the operation state of the device is abnormal operation state, the current suspicious data is marked and converted into invalid data, and at the same time, it is added to the abnormal device data set.

[0082] In the present invention, the working process of the anomaly detection model is as follows:

[0083] A: The environment analysis layer of the anomaly detection model uses the reflection mapping algorithm to decompose and sample the device deployment environment information to obtain an environmental stress mapping diagram;

[0084] B: The feature extraction layer of the anomaly detection model uses an improved ant colony algorithm to extract features from the abnormal device data set and the normal device data set according to the preset information extraction mechanism, obtaining an abnormal feature matrix and a normal feature matrix;

[0085] C: The feature correlation layer of the anomaly detection model uses the frequent pattern growth algorithm and combines the environmental stress mapping diagram to perform information correlation on the anomaly feature matrix and the normal feature matrix respectively, obtaining the environment-device feature matrix;

[0086] D: The iterative training layer of the anomaly detection model uses the inertia compressed particle swarm algorithm and combines the preset device part threshold set to perform several iterative trainings on the environment-device feature matrix, obtaining the comprehensive monitoring weight matrix;

[0087] E: The prediction and update layer of the anomaly detection model uses the energy-non-saturated loss function and combines the device historical data and the device operation overview table to perform feedback optimization on the comprehensive monitoring weight matrix, and then predicts and outputs the anomaly device data and the normal device data.

[0088] In this embodiment, the energy-non-saturated loss function:

[0089] Let the discriminator be an autoencoder, and the energy function is defined as the reconstruction error E(x):

[0090] E(t) = ||Dec(Enc(t)) - t|| 2

[0091] where t is the comprehensive monitoring weight matrix, Enc(·) is the encoder, and Dec(·) is the decoder;

[0092] where the discriminator needs to satisfy: for the real device data t r , minimize the energy E(t r ); for the generated data G(z), maximize the energy E(G(z)); the discriminator gradient function The generator gradient function E z []; Introduce the non-saturated idea and define the discriminator loss function L D :

[0093]

[0094] The first term minimizes the real data energy, and the second term realizes the non-saturated maximization of the generated data energy through the logistic function;

[0095] Minimize the generated data energy, and use the non-saturated form to obtain the generator loss function L G :

[0096] L G = -E z [log(1 + e -E(G(z)) )

[0097] This function has a stronger gradient when E(G(z)) is larger, avoiding the gradient vanishing problem of the traditional EBGAN.

[0098] In the present invention, the asset positioning service module hierarchically splits the abnormal device dataset and the normal device dataset respectively according to the device identification code in accordance with the identification resolution coding rule, in combination with the device deployment environment information, so as to determine the abnormal parts of all devices. At the same time, the metadata and master data in the enterprise node are synchronously updated in combination with the device management table and the device details list, and the abnormal device dataset and the normal device dataset are filled in the device management table and the device details list to obtain the device status table, and then the deployment position positioning and abnormal part positioning of all devices are completed in combination with the positioning correction model.

[0099] In this embodiment, the hierarchical splitting is to hierarchically decompose a complex system / problem into smaller and more manageable sub-units according to the logical level to form a tree structure, ensuring that the information granularity of each level is consistent and there is no intersection.

[0100] In the present invention, the working process of the positioning correction model is as follows:

[0101] The communication analysis layer of the positioning correction model analyzes the communication protocols of all devices in combination with the device distance and the signal attenuation value according to the device identification code in combination with the security status, and obtains the protocol-environment-distance communication table;

[0102] In this embodiment, let the device set be i is the device serial number, and n is the total number of devices; definition: Ci is the communication protocol feature vector of device di; Ei = α·||xi - xAP|| + β·Ai, where Ei is the environmental attenuation factor (α is the distance attenuation coefficient, β is the signal attenuation coefficient, and Ai is the environmental absorption parameter); Si is the security status weight; then the protocol-environment-distance communication table Tcom is generated:

[0103] Among them, represents the tensor product operation;

[0104] The positioning analysis layer of the positioning correction model analyzes the deployment positions of all devices according to the preset positioning accuracy threshold and the signal anti-interference value, and obtains the device deployment accuracy table;

[0105] In this embodiment, let the positioning accuracy threshold be ∈th and the anti-interference coefficient be γ, then the device deployment accuracy evaluation Pi: Among them is the signal gradient at position xi, is the noise variance; the device deployment accuracy table Tpre is generated:

[0106] The data judgment layer of the positioning correction model expands and judges the comprehensive monitoring weight matrix according to the protocol-environment-distance communication table in combination with the device deployment accuracy table, and obtains the positioning data expansion matrix;

[0107] In this embodiment, a comprehensive monitoring weight matrix W ∈ R n×m , is extended through the joint expansion of two tables: W^ = W ⊙ (Tcom T ·Tpre) + λ·diag(P1,...,Pn); where ⊙ is the Hadamard product and λ is the regularization coefficient; the positioning data expansion matrix Mext is obtained: Mext = SVD(W^)·Σk; the first k principal components are retained (Σk is the singular value truncation matrix);

[0108] The positioning correction layer of the positioning correction model corrects the deployment position and abnormal parts of the device according to the positioning data expansion matrix and the abnormal device dataset, in combination with the identification coding atlas and the environmental impact factor, to obtain the positioning data of the optimal deployment position and abnormal parts;

[0109] In this embodiment, an optimization objective function is constructed, where X is the positioning data matrix of the original deployment position:

[0110]

[0111] where Yobs is the observation data matrix, L is the graph Laplacian matrix constructed based on the identification coding atlas, μ is the regularization parameter, and F is the Frobenius norm; the positioning data X of the optimal deployment position * :

[0112] X * = (Mext T Mext + μL) -1 Mext T Yobs;

[0113] The detection of abnormal parts is performed through residual analysis: Δ = ||Yobs - MextX * || > τ·σ env , (τ is the environmental factor threshold, and σ env is the standard deviation of the environmental impact factor);

[0114] In the present invention, the display warning module uses a web page to display the filled device management table and device details to the administrator in real time, and at the same time uses notifications to alarm the abnormal parts of the device in combination with the abnormal device data, and stores them in the asset positioning database in combination with the device status table.

[0115] Example:

[0116] The hardware environment required for the main program to run: server CPU with 16 cores, 32G of memory, and 1TGB of hard disk.

[0117] Software environment required for the main program to run: Linux CentOS 7.9 operating system, JAVA JDK8, Spring Boot, Spring Cloud, Flink, Kafka, Elasticsearch, MySQL, etc.

[0118] The main program Jar package runs in the way of nohup java - jar xxx.jar;

[0119] The sub - program calls the Flink window to distribute the device data in Kafka and the database Greenplum to the display warning module;

[0120] The device communication end uses different communication protocols to connect to the asset positioning end and several devices through the information transmission module;

[0121] Apply for the enterprise node prefix through the active identification carrier module according to the enterprise authentication information; the enterprise authentication information includes the company organization code, enterprise name, enterprise address, enterprise type and legal representative; at the same time, obtain the factory list of all enterprises according to the enterprise name and enterprise address combined with the enterprise type; cluster the devices in all factories of the enterprise to obtain the device list of all factories; analyze the devices to obtain all parts of each device and get the device part table;

[0122] At the same time, use the sensor network to collect the operation data of the devices to obtain the real - time device data; perform asymmetric encryption and decryption operations on the real - time device data, and then load it into the information transmission module for transmission to the asset positioning end; the devices include Internet of Things devices and positioning devices; the real - time device data includes the deployment location, security status, operation log and remote control data;

[0123] The identification and resolution docking module of the asset positioning terminal (Internet of Things asset positioning system) generates a factory code according to the identification and resolution system, using the enterprise node in combination with the factory list and the enterprise node prefix; at the same time, according to the device list in combination with the factory code, the device category code and the device code are generated in sequence; at the same time, according to the device part table in combination with the device real-time data, the code point codes in each part of the device are obtained; finally, the enterprise node prefix, company organization code, factory code, device category code, device code, part code and code point code are combined to obtain the device identification code, and then the identification code is issued for all devices; subsequently, the identification and resolution docking module aggregates the device real-time data of the same company according to the device identification code to obtain the device operation overview table, and at the same time, according to the factory code, the device operation overview table is combined with the device real-time data for further division and aggregation to obtain the device management table, and at the same time, according to the device attributes in combination with the device code, the device management table of the same device is extended to obtain the device details list, and then the identification and resolution of the device real-time data is completed; finally, the device operation overview table, the device management table and the device details list are hierarchically nested using the knowledge graph to obtain the identification code graph, and then the identification registration of all devices is completed and transmitted to the recursive resolution node;

[0124] The data analysis and processing module hierarchically expands the device real-time data of all devices according to the device identification code in combination with the device management table to obtain the device part association data set;

[0125] At the same time, according to the preset device part threshold set in combination with the device deployment environment information, the device real-time data is detected, judged and stored for backup:

[0126] If the acquisition interval of adjacent device real-time data is equal to the preset collection time difference, it is determined that the current device real-time data is valid data, and it is input into the anomaly detection model in combination with the device operation overview table for data quality detection to obtain the abnormal device data set and the normal device data set;

[0127] If the acquisition interval of adjacent device real-time data is not equal to the preset collection time difference, it is determined that the current device real-time data is suspicious data, and at the same time, the operating state of the device is inspected:

[0128] If the operating state of the device is shutdown or standby state, the current suspicious data is marked and converted into valid data, and at the same time, it is added to the normal device data set;

[0129] If the operating state of the device is normal operating state, the current suspicious data is marked and converted into invalid data, and at the same time, the invalid data is cleaned using the data cleaning algorithm, and a suspicious data set is constructed in combination with the acquisition time;

[0130] If the operating state of the device is abnormal operating state, the current suspicious data is marked and converted into invalid data, and at the same time, it is added to the abnormal device data set;

[0131] The Asset Location Service Module (IoT Asset Location Service Module) hierarchically splits the abnormal device dataset and the normal device dataset respectively according to the device identification code in accordance with the identification resolution coding rules, combined with the device deployment environment information, so as to determine the abnormal parts of all devices. At the same time, it synchronously updates the metadata and master data in the enterprise node by combining the device management table and the device details list, and fills the abnormal device dataset and the normal device dataset in the device management table and the device details list to obtain the device status table, and then completes the deployment location positioning and abnormal part positioning of all devices in combination with the positioning correction model;

[0132] The Display and Warning Module uses the web page to display the filled device management table and device details list to the administrator in real time. At the same time, it uses notifications to alarm the abnormal parts of the device in combination with the abnormal device data, and stores them in the asset location database (IoT Asset Location System Database GreenPlum) in combination with the device status table.

Claims

1. An Internet of Things asset positioning system based on an identification and resolution system, characterized in that: It includes a device communication end and an asset positioning end; among them, The device communication end is used to collect real-time device data through an active identification carrier module and generate a device identification code, and use different communication protocols through an information transmission module to connect and transmit data with the asset positioning end and several devices; The asset positioning end is used to parse the device identification code by combining with the identification resolution system through an identification resolution docking module, detect and back up the collected real-time device data through a data analysis and processing module using an anomaly detection model, and locate all devices through an asset positioning service module using a positioning correction model. At the same time, the device status and anomaly alarm notifications are displayed through a display and warning module.

2. The Internet of Things asset positioning system based on the identifier resolution system according to claim 1, wherein: The information transmission module uses different communication protocols to connect with the asset positioning end and several devices; the devices include Internet of Things devices and positioning devices.

3. The Internet of Things asset positioning system based on the identifier resolution system according to claim 1, wherein: The active identification carrier module applies for an enterprise node prefix according to the enterprise authentication information; The enterprise authentication information includes the company organization code, enterprise name, enterprise address, enterprise type, and legal representative; At the same time, the factory list of all enterprises is obtained according to the enterprise name, enterprise address, and enterprise type; And the devices in all factories of the enterprise are clustered to obtain the device list of all factories; The devices are parsed to obtain all parts of each device, and a device part table is obtained; at the same time, the operation data of the devices is collected to obtain real-time device data; and the real-time device data is subjected to asymmetric encryption and decryption operations, and then loaded into the information transmission module to be transmitted to the asset positioning end.

4. The Internet of Things asset positioning system based on the identification and resolution system according to claim 1, wherein: The identification resolution docking module generates a factory code by combining the enterprise node with the factory list and the enterprise node prefix according to the identification resolution system; At the same time, the device category code and device code are generated in sequence according to the device list and the factory code; at the same time, the code point codes in each part of the device are obtained according to the device part table and the real-time device data; finally, the enterprise node prefix, company organization code, factory code, device category code, device code, part code, and code point code are combined to obtain the device identification code, and then the identification code is issued for all devices; subsequently, the identification resolution docking module aggregates the real-time device data of the same company according to the device identification code to obtain a device operation overview table, and at the same time, the device operation overview table and the real-time device data are re-divided and aggregated according to the factory code to obtain a device management table. At the same time, the device management table of the same device is extended according to the device attributes and the device code to obtain a device detailed list, and then the identification resolution of the real-time device data is completed; finally, the device operation overview table, device management table, and device detailed list are hierarchically nested to obtain an identification code map, and then the identification registration of all devices is completed and transmitted to the recursive resolution node.

5. The Internet of Things asset positioning system based on the identifier resolution system according to claim 1, characterized in that: The identification resolution system includes a top-level node, a secondary node, an enterprise node, and a recursive resolution node; the operation mechanism of the identification resolution system is: According to the enterprise node prefix in the recursive resolution node and the enterprise address, all device identification codes in the enterprise node are associated and allocated to obtain a regional identification list, which is classified as a comprehensive secondary node in the secondary node; Divide the regional identification list according to the industry to which the enterprise belongs and the enterprise scale, obtain the industry identification sub-list, and classify it as the industry-type secondary node in the secondary node; The top-level node connects and parses the secondary node according to the docking application information and the access application protocol to obtain identification statistical data, metadata, and master data; The top-level node divides the metadata and master data of the comprehensive-type secondary node and the industry-type secondary node into several batches of synchronized data sets according to the identification statistical data and the data synchronization channel; The top-level node synchronously expands and feedback-updates the metadata in each batch of synchronized data sets according to the data attributes, and at the same time fills in the master data in a preset format according to the data type for full-volume and incremental synchronization; The top-level node sends monitoring requests to all enterprise nodes and secondary nodes according to the preset parsing port, and at the same time receives the security status and operation logs of all enterprise nodes and secondary nodes.

6. The Internet of Things asset positioning system based on the identification and resolution system according to claim 1, characterized in that: The data analysis and processing module hierarchically expands the device real-time data of all devices according to the device identification code and combines the device management table to obtain the device part association data set; At the same time, according to the preset device part threshold set and the device deployment environment information, detect, judge, store, and back up the device real-time data: If the acquisition interval of adjacent device real-time data is equal to the preset collection time difference, it is determined that the current device real-time data is valid data, and it is input into the anomaly detection model and combined with the device operation overview table for data quality detection to obtain the anomaly device data set and the normal device data set; If the acquisition interval of adjacent device real-time data is not equal to the preset collection time difference, it is determined that the current device real-time data is suspicious data, and at the same time, the operation state of the device is inspected: If the operation state of the device is shutdown or standby state, mark and convert the current suspicious data into valid data, and add it to the normal device data set at the same time; If the operation state of the device is normal operation state, mark and convert the current suspicious data into invalid data, clean the invalid data at the same time, and construct a suspicious data set in combination with the acquisition time; If the operation state of the device is abnormal operation state, mark and convert the current suspicious data into invalid data, and add it to the abnormal device data set at the same time.

7. The Internet of Things asset positioning system based on the identifier resolution system according to claim 6, characterized in that: The working process of the anomaly detection model is as follows: A: The environment analysis layer of the anomaly detection model uses the reflection mapping algorithm to decompose and sample the device deployment environment information to obtain the environmental stress mapping diagram; B: The feature extraction layer of the anomaly detection model uses the improved ant colony algorithm to extract features from the anomaly device data set and the normal device data set according to the preset information extraction mechanism to obtain the anomaly feature matrix and the normal feature matrix; C: The feature association layer of the anomaly detection model uses the frequent pattern growth algorithm to respectively associate the information of the anomaly feature matrix and the normal feature matrix with the environmental stress mapping diagram to obtain the environment-device feature matrix; D: The iterative training layer of the anomaly detection model uses the inertial compression particle swarm algorithm to perform several iterative trainings on the environment-device feature matrix in combination with the preset device part threshold set to obtain the comprehensive monitoring weight matrix; E: The prediction update layer of the anomaly detection model uses the energy-unsaturated loss function, combines the historical data of the device and the overall device operation table to feedback and optimize the comprehensive monitoring weight matrix, and then predicts and outputs the anomaly device data and normal device data.

8. The Internet of Things asset positioning system based on the identification resolution system according to claim 1, wherein: The asset location service module hierarchically splits the anomaly device dataset and the normal device dataset according to the device identification code in accordance with the identification resolution coding rules, combined with the device deployment environment information, and then determines the anomaly parts of all devices. At the same time, it synchronously updates the metadata and master data in the enterprise node by combining the device management table and the device details list, and fills the anomaly device dataset and the normal device dataset in the device management table and the device details list to obtain the device status table. Then, it combines the location correction model to complete the deployment location and anomaly part location of all devices.

9. The Internet of Things asset positioning system based on the identifier resolution system according to claim 1, characterized in that: The working process of the location correction model is as follows: The communication parsing layer of the location correction model parses the communication protocols of all devices in combination with the device distance and signal attenuation value according to the device identification code combined with the security status, and obtains the protocol-environment-distance communication table; The location analysis layer of the location correction model analyzes the deployment locations of all devices according to the preset location accuracy threshold and signal anti-interference value, and obtains the device deployment accuracy table; The data judgment layer of the location correction model makes an extended judgment on the comprehensive monitoring weight matrix according to the protocol-environment-distance communication table combined with the device deployment accuracy table, and obtains the location data expansion matrix; The location correction layer of the location correction model corrects the deployment locations and anomaly parts of the devices according to the location data expansion matrix and the anomaly device dataset, combined with the identification coding map and the environmental impact factor, and obtains the location data of the optimal deployment location and anomaly parts.

10. The Internet of Things asset positioning system based on the identifier resolution system according to claim 1, characterized in that: The display and warning module displays the filled device management table and device details list to the administrator in real time, and at the same time alarms the anomaly parts of the device combined with the anomaly device data, and stores them in the asset location database in combination with the device status table.

Images