A multi-source industrial equipment data acquisition and fusion system and method

Abstract

A multi-source industrial equipment data acquisition and fusion system and method, which relates to the technical field of data fusion; adaptively converts and performs edge preprocessing on the acquired information of industrial equipment, divides the acquired information into key information and non-key information, encrypts and transmits the key information and non-key information respectively, constructs a digital twin model of the industrial equipment and its acquisition point topology network, obtains the simulation information, time correction coefficient, and space correction coefficient of each acquisition point, obtains the correction information of each acquisition point according to the time correction coefficient and the space correction coefficient, constructs a spatio-temporal correction model, uses the spatio-temporal correction model to obtain the real-time correction information of each acquisition point, judges the operating state of each industrial equipment and generates an abnormal signal for feedback; can reduce the data processing volume of the cloud and can accurately reflect the operating state of industrial equipment.

Description

Technical Field

[0001] The present invention relates to the technical field of data fusion, and specifically to a multi-source industrial equipment data acquisition and fusion system and method. Background Art

[0002] Multi-source data acquisition and fusion for industrial equipment is an advanced solution developed in the modern industrial production environment to improve the efficiency and quality of the manufacturing process. With the popularization of the concept of Industry 4.0 and the development of intelligent manufacturing technology, data acquisition and analysis have become the key links to achieve intelligent production, providing strong support for the intelligent management of industrial production processes;

[0003] In the past, data acquisition for industrial equipment mostly targeted a single industrial equipment, rather than combining different industrial equipment. How to perform data acquisition and fusion for different industrial equipment is a problem that needs to be solved. Since data acquisition mostly uses sensors, the installation position of the sensors will directly affect their acquisition accuracy, and the distance from the center of the industrial equipment will directly cause deviations in the acquired information in terms of time and space, resulting in its inability to accurately reflect the operating state of the industrial equipment. In view of the deficiencies of the prior art, the present invention provides a multi-source industrial equipment data acquisition and fusion system and method. Summary of the Invention

[0004] The purpose of the present invention is to provide a multi-source industrial equipment data acquisition and fusion system and method.

[0005] The purpose of the present invention can be achieved through the following technical solutions: A multi-source industrial equipment data acquisition and fusion system includes the following modules:

[0006] A data acquisition module, used to input the basic information and acquisition information of each industrial equipment, and adaptively convert the acquisition information of each industrial equipment using a protocol adapter;

[0007] A data transmission module, used to perform edge preprocessing on the acquisition information of each industrial equipment, divide the acquisition information into key information and non-key information, and encrypt and transmit the key information and non-key information respectively using blockchain technology;

[0008] A data fusion module, used to construct a digital twin model of each industrial equipment and its corresponding acquisition point topology network, perform simulation on the digital twin model to obtain simulation information of each acquisition point, obtain a time correction coefficient of each acquisition point according to the acquisition point topology network, and obtain corresponding space correction coefficients according to the acquisition information and simulation information of each acquisition point;

[0009] A data evaluation module, which is used to obtain correction information according to the time correction coefficient and space correction coefficient of each collection point, construct a corresponding spatio-temporal correction model, use the spatio-temporal correction model to obtain the real-time correction information of each collection point respectively, judge the operating state of each industrial device according to the real-time correction information, and generate an abnormal signal for feedback.

[0010] Further, the process of inputting the basic information and collection information of each industrial device, and adaptively converting the collection information of each industrial device by using a protocol adapter includes:

[0011] Set an input unit, and input the basic information of each industrial device through the input unit. The basic information includes device parameters, communication parameters, physical parameters, and operation parameters;

[0012] Set several collection points on the industrial device, and set a temperature collection unit, a humidity collection unit, a pressure collection unit, a vibration collection unit, and an operation collection unit at the collection points to obtain the collection information of the industrial device, including temperature parameters, humidity parameters, pressure parameters, vibration parameters, and operation parameters;

[0013] Set a protocol adapter on the industrial device. Its hardware consists of an edge gateway and an FPGA, and its software consists of a protocol feature library and a lightweight CNN model, which is used to identify the protocol features of the industrial device and convert the collection information of the industrial device into a standardized JSON format.

[0014] Further, the process of edge preprocessing the collection information of each industrial device, dividing the collection information into key information and non-key information, and encrypting and transmitting the key information and non-key information respectively by using blockchain technology includes:

[0015] Set a preprocessing unit on the industrial device, which is used to preprocess the collection information of the industrial device, including clearing missing data, clearing abnormal data, and clearing duplicate data;

[0016] Regard the pressure parameters and vibration parameters in the collection information as key information, regard the temperature parameters, humidity parameters, and operation parameters among them as non-key information, and use blockchain technology to generate a blockchain light node ID for each industrial device;

[0017] Encrypt the key information and non-key information respectively, bind them with the blockchain light node ID of their corresponding industrial device, and attach their corresponding hash values to build a digital platform for the industrial device. For the key information, transmit it to the digital platform through a 5G+TSN network. For the non-key information and basic information, transmit them to the digital platform through LoRa.

[0018] Further, the process of constructing the digital twin models of each industrial device and their corresponding acquisition point topology networks and performing simulation on the digital twin models to obtain the simulation information of each acquisition point includes:

[0019] In the digital platform, use digital twin technology to construct the digital twin model of a single industrial device according to its basic information. In the digital twin model, convert the installation positions of each acquisition point and the center of the industrial device into an acquisition point topology network, where the nodes are acquisition points and the edges are physical connection relationships;

[0020] Use simulation software to perform simulation on the digital twin model of the industrial device according to its operating parameters, so as to output the simulation vibration parameters and simulation pressure parameters of each acquisition point. The simulation information includes simulation vibration parameters and simulation pressure parameters;

[0021] Construct the digital twin models of each industrial device and their acquisition point topology networks respectively, and use simulation software to perform simulation on them respectively, so as to output the corresponding simulation vibration parameters and simulation pressure parameters.

[0022] Further, the process of obtaining the time correction coefficient of each acquisition point according to the acquisition point topology network and obtaining the corresponding space correction coefficient according to the acquisition information and simulation information of each acquisition point includes:

[0023] For the acquisition point i on a single industrial device, the time correction coefficient of the monitored vibration parameter is ;

[0024] Among them, represents the physical propagation delay, is the connection distance between the acquisition point i and the center of the industrial device in the acquisition point topology network, is the sound speed in the industrial device material, is the clock deviation between the two;

[0025] Denote the vibration parameter monitored by the acquisition point i at the same moment and the simulation vibration parameter output by its digital twin model as S i (t) and S p,i (t), and obtain the space correction coefficient ;

[0026] ;

[0027] Among them, is the preset standard space correction coefficient, is the corresponding moment. The above are the time correction coefficient and space correction coefficient obtained for the vibration parameter. The same method is used to obtain the time correction coefficient and space correction coefficient corresponding to the pressure parameter.

[0028] Furthermore, the process of obtaining correction information based on the time correction coefficient and the space correction coefficient of each acquisition point and constructing a corresponding spatio-temporal correction model includes:

[0029] For the acquisition point i on a single industrial device, an update period is set. When an update period is reached, the time correction coefficient of the acquisition point i is obtained once and the space correction coefficient , and the corresponding corrected vibration parameter is obtained ;

[0030] ;

[0031] A first spatio-temporal correction set is generated according to the vibration parameters and operating parameters of different acquisition points and their corresponding corrected vibration parameters, and the first spatio-temporal correction set is divided into a training set and a test set;

[0032] A convolutional neural network is constructed. The vibration parameters and operating parameters in the training set are used as the input data of the convolutional neural network, and the corresponding corrected vibration parameters in the training set are used as the output data of the convolutional neural network. The convolutional neural network is trained to obtain an initial convolutional neural network;

[0033] The initial convolutional neural network is verified using the test set, and the initial convolutional neural network with an output less than or equal to a preset test error threshold is used as the first spatio-temporal correction model for vibration parameters;

[0034] The above is the first spatio-temporal correction model obtained for vibration parameters. The same method is adopted to obtain the second spatio-temporal correction model corresponding to the pressure parameters. The correction information includes corrected vibration parameters and corrected pressure parameters.

[0035] Furthermore, the process of using the spatio-temporal correction model to obtain the real-time correction information of each acquisition point, judging the operating state of each industrial device according to the real-time correction information, and generating an abnormal signal for feedback includes:

[0036] The vibration parameters and operating parameters of each acquisition point are input into the first spatio-temporal correction model in real time to obtain real-time corrected vibration parameters, and the pressure parameters and operating parameters of each acquisition point are input into the second spatio-temporal correction model in real time to obtain real-time corrected pressure parameters;

[0037] The real-time correction information includes real-time corrected vibration parameters and real-time corrected pressure parameters. A vibration parameter threshold and a pressure parameter threshold are set respectively, and the real-time corrected vibration parameters and real-time corrected pressure parameters of each acquisition point are compared with their corresponding vibration parameter threshold and pressure parameter threshold respectively;

[0038] When the real-time corrected vibration parameter is greater than the vibration parameter threshold, it is determined that the industrial equipment is currently in an abnormal vibration state, and a corresponding abnormal vibration signal is generated for feedback. When the real-time corrected pressure parameter is greater than the pressure parameter threshold, it is determined that the industrial equipment is currently in an abnormal pressure state, and a corresponding abnormal pressure signal is generated for feedback.

[0039] A multi-source industrial equipment data acquisition and fusion method includes the following steps:

[0040] Step S1: Enter the basic information and acquisition information of each industrial equipment, and use the protocol adapter to adaptively convert the acquisition information of each industrial equipment. Perform edge preprocessing on the acquisition information of each industrial equipment, divide the acquisition information into key information and non-key information, and use blockchain technology to encrypt and transmit the key information and non-key information respectively;

[0041] Step S2: Construct the digital twin model of each industrial equipment and its corresponding acquisition point topology network, perform simulation on the digital twin model to obtain the simulation information of each acquisition point, obtain the time correction coefficient of each acquisition point according to the acquisition point topology network, and obtain the corresponding space correction coefficient according to the acquisition information and simulation information of each acquisition point;

[0042] Step S3: Obtain the correction information according to the time correction coefficient and space correction coefficient of each acquisition point, construct the corresponding space-time correction model, use the space-time correction model to obtain the real-time correction information of each acquisition point respectively, judge the operating state of each industrial equipment according to the real-time correction information, and generate an abnormal signal for feedback.

[0043] Compared with the prior art, the beneficial effects of the present invention are:

[0044] By deploying a protocol adapter and a preprocessing unit on each industrial equipment respectively, the present invention can perform protocol conversion and preprocessing on the acquisition information of different industrial equipment at the edge end, which is beneficial to reducing the data processing volume in the cloud. By dividing the acquisition information into key information and non-key information and transmitting them through different paths, on the one hand, it can reduce the transmission delay of key information, and on the other hand, it can reduce the transmission power consumption of non-key information;

[0045] By constructing digital twin models of various industrial devices and their corresponding acquisition point topology networks, obtaining simulation information of each acquisition point under ideal conditions, and combining the actual acquisition information, the time correction coefficient and space correction coefficient of each acquisition point can be obtained. Furthermore, the acquisition information of each acquisition point can be corrected to obtain corrected information, and a corresponding spatio-temporal correction model can be constructed, which can optimize the deviation in time and the deviation in space simultaneously. The deep learning technology can be used to directly feedback the corresponding real-time correction information according to the acquisition information and operating parameters, accurately reflecting the operating state of industrial devices and timely prompting relevant personnel to repair industrial devices with anomalies. Brief Description of the Drawings

[0046] Figure 1 This is the schematic diagram of the present invention. Detailed Embodiment

[0047] As Figure 1 shown, a multi-source industrial device data acquisition and fusion system includes the following modules:

[0048] The data acquisition module is used to input the basic information and acquisition information of various industrial devices, and adaptively convert the acquisition information of various industrial devices by using a protocol adapter;

[0049] The data transmission module is used to perform edge preprocessing on the acquisition information of various industrial devices, divide the acquisition information into key information and non-key information, and encrypt and transmit the key information and non-key information respectively by using blockchain technology;

[0050] The data fusion module is used to construct digital twin models of various industrial devices and their corresponding acquisition point topology networks, perform simulation on the digital twin models to obtain simulation information of each acquisition point, obtain the time correction coefficient of each acquisition point according to the acquisition point topology network, and obtain the corresponding space correction coefficient according to the acquisition information and simulation information of each acquisition point;

[0051] The data evaluation module is used to obtain correction information according to the time correction coefficient and space correction coefficient of each acquisition point, construct a corresponding spatio-temporal correction model, obtain the real-time correction information of each acquisition point by using the spatio-temporal correction model, judge the operating state of each industrial device according to the real-time correction information, and generate an abnormal signal for feedback.

[0052] It should be further noted that in the specific implementation process, the process of inputting the basic information and acquisition information of various industrial devices and adaptively converting the acquisition information of various industrial devices by using a protocol adapter includes:

[0053] Set up an input unit to input the basic information of each industrial device through the input unit. The basic information refers to the fixed data read from each industrial device itself and is used to identify and manage industrial devices, including device parameters, communication parameters, physical parameters, and operation parameters;

[0054] The device parameters include device model, device serial number, and device manufacturer. The communication parameters include IP address, MAC address, communication protocol, and port information. The physical parameters include size, weight, color, and material. The operation parameters include operating temperature range, power requirements, and power consumption;

[0055] Set several collection points on each industrial device respectively, and obtain the collection information of each industrial device through the collection points, including temperature parameters, humidity parameters, pressure parameters, vibration parameters, and operation parameters. The collection information refers to the dynamic data of each industrial device and is used to monitor the operating status and environmental conditions of industrial devices;

[0056] Set a temperature collection unit, a humidity collection unit, a pressure collection unit, a vibration collection unit, and an operation collection unit on the collection points respectively, which are used to obtain the corresponding temperature parameters, humidity parameters, pressure parameters, vibration parameters, and operation parameters of each industrial device. The operation parameters include speed and rotation speed, acceleration, current and voltage, and power;

[0057] Set corresponding protocol adapters on each industrial device respectively. The hardware of the protocol adapter consists of an edge gateway and an FPGA, and the software consists of a protocol feature library and a lightweight CNN model. It can automatically identify the protocol features of each industrial device, such as the function code of Modbus and the Topic structure of ROS, and dynamically convert the collection information of each industrial device into a standardized JSON format.

[0058] It should be further noted that in the specific implementation process, the edge preprocessing of the collection information of each industrial device, dividing the collection information into key information and non-key information, and the process of encrypting and transmitting the key information and non-key information respectively using blockchain technology includes:

[0059] Set corresponding preprocessing units on each industrial device respectively, and preprocess the collection information of its corresponding industrial device through the preprocessing unit, including clearing missing data, clearing abnormal data, and clearing duplicate data;

[0060] Divide the collection information of each industrial device respectively, regard the pressure parameters and vibration parameters among them as key information, regard the temperature parameters, humidity parameters, and operation parameters among them as non-key information, and use blockchain technology to generate a unique blockchain light node ID for each industrial device;

[0061] Encrypt the key information and non-key information separately, bind them to the blockchain light node ID of the corresponding industrial device, and attach their corresponding hash values to build a digital platform for industrial devices, which is used to summarize and process the collected information of each industrial device;

[0062] For key information, transmit it to the digital platform through the 5G+TSN network, which can reduce transmission latency. For non-key information and basic information, transmit it to the digital platform through LoRa, which can reduce transmission power consumption.

[0063] It should be further noted that in the specific implementation process, the process of constructing the digital twin model of each industrial device and its corresponding acquisition point topology network, and simulating the digital twin model to obtain the simulation information of each acquisition point includes:

[0064] In the digital platform, use digital twin technology to build a corresponding digital twin model according to the basic information of a single industrial device, mark the installation positions of each acquisition point in the digital twin model, and convert the installation positions of each acquisition point and the center of the industrial device into a corresponding topology network, denoted as the acquisition point topology network, whose nodes are acquisition points and the edges are physical connection relationships;

[0065] Use simulation software to simulate the digital twin model of the industrial device according to its operating parameters, simulate the vibration propagation of the industrial device based on finite element analysis (FEA) to output the simulation vibration parameters of each acquisition point, and use computational fluid dynamics (CFD) to simulate the pressure distribution to output the simulation pressure parameters of each acquisition point;

[0066] Adopt the same method to build the digital twin model of each industrial device and its corresponding acquisition point topology network respectively, use simulation software to simulate them respectively, and output the corresponding simulation vibration parameters and simulation pressure parameters respectively. The simulation information includes the simulation vibration parameters and simulation pressure parameters of each acquisition point.

[0067] It should be further noted that in the specific implementation process, the process of obtaining the time correction coefficient of each acquisition point according to the acquisition point topology network and obtaining the corresponding space correction coefficient according to the acquisition information and simulation information of each acquisition point includes:

[0068] For the acquisition point i on a single industrial device, the time correction coefficient of the monitored vibration parameter is ;

[0069] Among them, represents the physical propagation delay, is the connection distance between the center of the acquisition point i and the industrial equipment in the acquisition point topology network, is the sound speed in the industrial equipment material, is the clock deviation between the two;

[0070] The vibration parameters monitored by the acquisition point i at the same moment and the simulated vibration parameters output by its digital twin model are respectively denoted as S i (t) and S p,i (t), then the spatial correction coefficient at this moment is:

[0071] ;

[0072] Among them, is the preset standard spatial correction coefficient, is the corresponding moment, and through dynamic update can compensate for the signal attenuation caused by the installation position difference.

[0073] The above are the time correction coefficient and spatial correction coefficient obtained for the vibration parameters. The same method is used to obtain the time correction coefficient and spatial correction coefficient corresponding to the pressure parameters.

[0074] It should be further noted that in the specific implementation process, the process of obtaining the correction information according to the time correction coefficient and spatial correction coefficient of each acquisition point and constructing the corresponding spatio-temporal correction model includes:

[0075] For the acquisition point i on a single industrial equipment, its vibration parameter S i (t) and the operating parameters of its industrial equipment are obtained in real time, and simulation is carried out according to the operating parameters in the digital twin model to output the corresponding simulated vibration parameter S p,i (t);

[0076] Set the update period. When an update period is reached, obtain the time correction coefficient of the acquisition point i once and the spatial correction coefficient , and output the calibrated corrected vibration parameter, denoted as ;

[0077] ;

[0078] Generate the first spatio-temporal correction set according to the vibration parameters, operating parameters and their corresponding corrected vibration parameters of different acquisition points, and divide the obtained first spatio-temporal correction set into a training set and a test set;

[0079] Construct a convolutional neural network, use the vibration parameters and operating parameters in the training set as the input data of the convolutional neural network, and use the corresponding corrected vibration parameters in the training set as the output data of the convolutional neural network, and train the convolutional neural network to obtain an initial convolutional neural network;

[0080] Use the test set to verify the model of the initial convolutional neural network, and output the initial convolutional neural network whose test error is less than or equal to the preset test error threshold as the first spatio-temporal correction model of the vibration parameters;

[0081] The above is the first spatio-temporal correction model obtained for the vibration parameters. The same method is used to obtain the second spatio-temporal correction model corresponding to the pressure parameters. The correction information includes the corrected vibration parameters and corrected pressure parameters at each acquisition point.

[0082] It should be further noted that in the specific implementation process, the process of using the spatio-temporal correction model to obtain the real-time correction information of each acquisition point, judging the operating state of each industrial device according to the real-time correction information, and generating an abnormal signal for feedback includes:

[0083] Input the vibration parameters and operating parameters of each acquisition point into the first spatio-temporal correction model in real time, and output the corresponding real-time corrected vibration parameters. Input the pressure parameters and operating parameters of each acquisition point into the second spatio-temporal correction model in real time, and output the corresponding real-time corrected pressure parameters;

[0084] The real-time correction information includes real-time corrected vibration parameters and real-time corrected pressure parameters. Set the vibration parameter threshold and pressure parameter threshold respectively, and compare the real-time corrected vibration parameters and real-time corrected pressure parameters of each acquisition point with their corresponding vibration parameter thresholds and pressure parameter thresholds respectively;

[0085] When the real-time corrected vibration parameter is greater than the vibration parameter threshold, it is judged that the industrial device is currently in a vibration abnormal state, and a corresponding vibration abnormal signal is generated for feedback. In other cases, it is judged that the industrial device is currently in a vibration normal state, and no operation is performed on it;

[0086] When the real-time corrected pressure parameter is greater than the pressure parameter threshold, it is judged that the industrial device is currently in a pressure abnormal state, and a corresponding pressure abnormal signal is generated for feedback. In other cases, it is judged that the industrial device is currently in a pressure normal state, and no operation is performed on it;

[0087] The operating state includes vibration abnormal state, vibration normal state, pressure abnormal state, pressure normal state. The abnormal signals include vibration abnormal signal and pressure abnormal signal, which are used to prompt relevant personnel to repair the industrial device with abnormal operating state in time.

[0088] An embodiment of the present invention further includes a multi-source industrial equipment data acquisition and fusion method, comprising the following steps:

[0089] Step S1: Enter the basic information and acquisition information of each industrial equipment, adaptively convert the acquisition information of each industrial equipment using a protocol adapter, perform edge preprocessing on the acquisition information of each industrial equipment, divide the acquisition information into key information and non-key information, and encrypt and transmit the key information and non-key information respectively using blockchain technology;

[0090] Step S2: Construct a digital twin model of each industrial equipment and its corresponding acquisition point topology network, perform simulation on the digital twin model to obtain simulation information of each acquisition point, obtain a time correction coefficient of each acquisition point according to the acquisition point topology network, and obtain corresponding space correction coefficients according to the acquisition information and simulation information of each acquisition point;

[0091] Step S3: Obtain correction information according to the time correction coefficient and space correction coefficient of each acquisition point, construct a corresponding spatio-temporal correction model, obtain real-time correction information of each acquisition point using the spatio-temporal correction model, judge the operating state of each industrial equipment according to the real-time correction information, and generate an abnormal signal for feedback.

[0092] The above embodiments are only used to illustrate the technical method of the present invention and not to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical method of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A multi-source industrial equipment data acquisition and fusion system, characterized in that: Includes the following modules: The data acquisition module is used to input the basic information and collected information of each industrial device, and use the protocol adapter to adaptively convert the collected information of each industrial device; The data transmission module is used to perform edge preprocessing on the collected information of each industrial device, divide the collected information into key information and non-key information, and use blockchain technology to encrypt and transmit the key information and non-key information respectively; The data fusion module is used to build the digital twin model of each industrial equipment and its corresponding collection point topological network, simulate the digital twin model to obtain the simulation information of each collection point, obtain the time correction coefficient of each collection point according to the collection point topological network, and obtain the corresponding space correction coefficient according to the collection information and simulation information of each collection point; The data evaluation module is used to obtain correction information based on the time correction coefficient and space correction coefficient of each collection point, and to build a corresponding time-space correction model. The time-space correction model is used to obtain the real-time correction information of each collection point, and the operating status of each industrial equipment is judged according to the real-time correction information, and an abnormal signal is generated for feedback; The process of obtaining the time correction coefficient and space correction coefficient of each acquisition point includes: For a single industrial equipment collection point i, the time correction coefficient for monitoring vibration parameters is: in, represents the physical propagation delay, is the connection distance between the collection point i and the center of the industrial equipment in the collection point topology network, is the speed of sound in industrial equipment materials, is the clock deviation between the two; The vibration parameters monitored at the same time point i and the simulated vibration parameters output by the digital twin model are recorded as and , get the spatial correction coefficient at the corresponding time ; ； in, is the preset standard space correction factor, To correspond to the moment, the above are the time correction coefficient and space correction coefficient obtained for the vibration parameters. The same method is adopted to obtain the time correction coefficient and space correction coefficient corresponding to the pressure parameters.

2. A multi-source industrial equipment data acquisition and fusion system according to claim 1, characterized in that: The process of entering basic information and collected information and adaptively converting the collected information includes: Setting an input unit, through which basic information of each industrial device is input, the basic information includes device parameters, communication parameters, physical parameters, and operating parameters; Several collection points are set on the industrial equipment, and temperature collection units, humidity collection units, pressure collection units, vibration collection units, and operation collection units are set on the collection points to obtain collection information of the industrial equipment, including temperature parameters, humidity parameters, pressure parameters, vibration parameters, and operation parameters; A protocol adapter is set up on the industrial equipment. Its hardware consists of an edge gateway and FPGA, and its software consists of a protocol feature library and a lightweight CNN model. It is used to identify the protocol features of the industrial equipment and convert the collected information of the industrial equipment into a standardized JSON format.

3. A multi-source industrial equipment data acquisition and fusion system according to claim 2, characterized in that: The process of edge preprocessing and collecting information, encrypting and transmitting key information and non-key information separately includes: A preprocessing unit is set on the industrial equipment to preprocess the collected information of the industrial equipment, including clearing missing data, clearing abnormal data, and clearing duplicate data; The pressure parameters and vibration parameters in the collected information are taken as key information, and the temperature parameters, humidity parameters, and operating parameters are taken as non-key information. Blockchain technology is used to generate blockchain light node IDs for various industrial equipment. After encrypting the critical information and non-critical information separately, they are bound to the blockchain light node ID of the corresponding industrial equipment, and their corresponding hash values ​​are attached to build a digital platform for industrial equipment. For critical information, it is transmitted to the digital platform through the 5G+TSN network, and for non-critical information and basic information, it is transmitted to the digital platform through LoRa.

4. A multi-source industrial equipment data acquisition and fusion system according to claim 3, characterized in that: The process of building a digital twin model and its collection point topology network and obtaining simulation information of each collection point includes: In the digital platform, digital twin technology is used to build a digital twin model of a single industrial equipment based on its basic information. In the digital twin model, the installation location of each collection point and the center of the industrial equipment are converted into a collection point topology network, where the nodes are collection points and the edges are physical connection relationships. Using simulation software to simulate the digital twin model of the industrial equipment according to its operating parameters to output simulated vibration parameters and simulated pressure parameters of each collection point, wherein the simulation information includes simulated vibration parameters and simulated pressure parameters; The digital twin models of each industrial equipment and their collection point topological networks are constructed respectively, and they are simulated respectively using simulation software to output the corresponding simulated vibration parameters and simulated pressure parameters.

5. A multi-source industrial equipment data acquisition and fusion system according to claim 4, characterized in that: The process of obtaining correction information and building a spatiotemporal correction model includes: For a collection point i on a single industrial device, set an update cycle. When an update cycle is reached, obtain the time correction coefficient of the collection point i once. and space correction factor , and obtain the corresponding corrected vibration parameters ; ； Generate a first spatiotemporal correction set according to vibration parameters and operating parameters of different acquisition points and their corresponding corrected vibration parameters, and divide the first spatiotemporal correction set into a training set and a test set; Constructing a convolutional neural network, taking the vibration parameters and operating parameters in the training set as input data of the convolutional neural network, taking the corresponding corrected vibration parameters in the training set as output data of the convolutional neural network, and training the convolutional neural network to obtain an initial convolutional neural network; The initial convolutional neural network is model verified using the test set, and an initial convolutional neural network with a value less than or equal to a preset test error threshold is output as the first spatiotemporal correction model of the vibration parameters; The above is a first spatiotemporal correction model obtained for vibration parameters. The same method is adopted to obtain a second spatiotemporal correction model corresponding to pressure parameters. The correction information includes corrected vibration parameters and corrected pressure parameters.

6. A multi-source industrial equipment data acquisition and fusion system according to claim 5, characterized in that: The process of obtaining real-time correction information from each collection point, determining the operating status of each industrial equipment, and generating abnormal signals for feedback includes: Inputting the vibration parameters and operating parameters of each acquisition point into the first time-space correction model in real time to obtain real-time corrected vibration parameters, and inputting the pressure parameters and operating parameters of each acquisition point into the second time-space correction model in real time to obtain real-time corrected pressure parameters; The real-time correction information includes real-time correction vibration parameters and real-time correction pressure parameters, respectively setting vibration parameter thresholds and pressure parameter thresholds, and comparing the real-time correction vibration parameters and real-time correction pressure parameters of each acquisition point with their corresponding vibration parameter thresholds and pressure parameter thresholds; When the real-time corrected vibration parameter is greater than the vibration parameter threshold, it is judged that the industrial equipment is currently in an abnormal vibration state, and a corresponding abnormal vibration signal is generated for feedback. When the real-time corrected pressure parameter is greater than the pressure parameter threshold, it is judged that the industrial equipment is currently in an abnormal pressure state, and a corresponding abnormal pressure signal is generated for feedback.

7. A multi-source industrial equipment data collection and fusion method, which is implemented based on the multi-source industrial equipment data collection and fusion system according to any one of claims 1 to 6, characterized in that: The method comprises: Step S1: input the basic information and collected information of each industrial device, and use the protocol adapter to adaptively convert the collected information of each industrial device, perform edge preprocessing on the collected information of each industrial device, divide the collected information into key information and non-key information, and use blockchain technology to encrypt and transmit the key information and non-key information respectively; Step S2: construct a digital twin model of each industrial equipment and its corresponding collection point topological network, simulate the digital twin model to obtain simulation information of each collection point, obtain the time correction coefficient of each collection point according to the collection point topological network, and obtain the corresponding space correction coefficient according to the collection information and simulation information of each collection point; Step S3: Correction information is obtained based on the time correction coefficient and space correction coefficient of each acquisition point, and a corresponding space-time correction model is constructed. The real-time correction information of each acquisition point is obtained using the space-time correction model. The operating status of each industrial equipment is determined based on the real-time correction information, and an abnormal signal is generated for feedback.

Images