Data processing method and system based on industrial internet

By integrating Bluetooth AOA positioning system, device status sensor and monitoring video acquisition system on the production line, combined with convolutional neural network, in-depth analysis and risk identification of multi-source data is achieved, and the complexity of data acquisition and processing in the existing technology is solved. , the problem of insufficient real-time risk identification has improved the safety and stability of the production line.

CN119089351APending Publication Date: 2024-12-06ANHUI TIANCHENG ZHICHAIN TECH CO LTD

Patent Information

Application Number
CN202411139067.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-20
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

The existing production monitoring system faces the problems of diversity and complexity of data acquisition and processing, insufficient real-time risk identification and response, and limited intelligent decision-making capabilities when processing multi-source data.

Method used

Using Bluetooth AOA positioning system, device status sensor, environmental sensor and monitoring video acquisition system, multi-source data is transmitted to the big data center in real time through the industrial Internet platform for cleaning and preprocessing, and using convolutional neural network to perform multi-source data In-depth analysis, identify potential risks such as equipment failures, material abnormalities, environmental abnormalities and fires, and generate corresponding alarms and emergency response measures through the intelligent decision-making module.

Benefits of technology

Improve the intelligence level of the industrial Internet system, enhance the safety and stability of the production line in complex environments, realize comprehensive monitoring and risk identification of all links of the production line, improve the response speed to emergency events, and reduce the cause Losses caused by misjudgment or delayed processing.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119089351A_ABST
    Figure CN119089351A_ABST
Patent Text Reader

Abstract

The invention discloses a data processing method and system based on an industrial internet. The method comprises the following steps: acquiring positioning data of personnel, equipment and materials in each link of a production line, operation state data and / or electrical parameters of the equipment, temperature, humidity and / or gas concentration parameters of a production place and monitoring video data of the production place; transmitting the acquired data to a big data center server through industrial internet equipment, and cleaning and preprocessing the data; identifying risks and anomalies; and analyzing the multi-source data through a pre-trained convolutional neural network, performing comprehensive risk identification, outputting an identification result, and executing a fusion intelligent decision. According to the data processing method and system based on the industrial internet, the intelligent level of an industrial internet system is improved, and the safety and stability of a production line in a complex environment are enhanced.
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to the field of industrial Internet technology, and in particular to a data processing method and system based on industrial Internet. [Background technology]

[0002] With the development of the industrial Internet, the intelligence and automation of production lines are constantly improving. How to effectively manage and monitor the status of each link in the production process has become a major challenge facing modern manufacturing. Industrial production lines usually include a variety of complex equipment, personnel flow and material management. The real-time collection, analysis and processing of various data are of great significance to ensure the smooth progress of production and safe production.

[0003] However, existing production monitoring systems often face the following major problems when processing these multi-source data: the diversity and complexity of data collection and processing; the lack of real-time risk identification and response; the limitation of intelligent decision-making capabilities; and the difficulty of fusing and analyzing multiple input data.

[0004] Therefore, how to design an intelligent data processing method based on the Industrial Internet so that it can effectively collect, process, and analyze multi-source data, and achieve comprehensive risk identification and intelligent decision-making through advanced algorithms has become an important research direction in the current technology field. [Summary of the invention]

[0005] In view of this, an embodiment of the present invention provides a data processing method and system based on the industrial Internet.

[0006] In a first aspect, an embodiment of the present invention provides a data processing method based on the industrial Internet, the method comprising:

[0007] S1. Collect the positioning data of personnel, equipment and materials in each link of the production line through the Bluetooth AOA positioning system, collect the operating status data and / or electrical parameters of the equipment through the equipment status sensor, collect the temperature, humidity and / or gas concentration parameters of the production site through the environmental sensor, and collect the monitoring video data of the production site through the camera;

[0008] S2, transmit the collected data to the big data center server through industrial Internet devices, and perform cleaning and preprocessing;

[0009] S3. Identify the failure risk of the equipment based on the equipment's location data, operating status data, and electrical parameters; identify the abnormal conditions of the materials based on the material's location data; identify the abnormal environmental conditions based on the environmental parameters provided by the environmental sensors; identify abnormal events based on the monitoring video data; and send a first-level warning notification to the maintenance personnel or issue a first-level alarm through the alarm device;

[0010] S4, analyze the multi-source data through the pre-trained convolutional neural network, conduct comprehensive risk identification, and output the identification results;

[0011] S5. Convert the recognition result into vector form and execute fusion intelligent decision, which includes sending a first-level warning notification to the maintenance personnel or issuing a first-level alarm through the alarm device, executing a second-level alarm, shutting down and / or reporting to the management terminal of the administrator.

[0012] According to the above aspects and any possible implementation, an implementation is further provided, wherein S1 specifically includes:

[0013] Bluetooth AOA positioning base stations are pre-deployed at the production site to collect positioning signals sent by Bluetooth tags installed on personnel, equipment and materials at various stages of the production line;

[0014] Installing a device status sensor on the device in advance, wherein the device status sensor includes a vibration sensor, a pressure sensor, a current sensor and / or a voltage sensor to monitor the vibration, internal pressure, current and / or voltage data of the device;

[0015] Pre-arrange environmental sensors at the production site, wherein the environmental sensors include temperature sensors, humidity sensors and / or gas sensors for collecting temperature, humidity and / or target gas concentration parameters of the production site;

[0016] Arrange cameras in advance at the production site to collect monitoring video data of the production site through the cameras;

[0017] Pre-configure mobile terminals for maintenance personnel;

[0018] The industrial Internet device is used to receive data transmitted by the Bluetooth AOA positioning base station, equipment status sensor, environmental sensor, camera and mobile terminal.

[0019] According to the above aspects and any possible implementation manner, an implementation manner is further provided, wherein the step S3 identifies the failure risk of the device according to the positioning data, the operating status data and the electrical parameters of the device, specifically including:

[0020] Clean and normalize the collected data to remove noise and outliers, and fuse the cleaned data to form a comprehensive operation data set of the equipment;

[0021] Perform feature extraction on the comprehensive operation data set of the equipment to extract key feature values, including key features of movement frequency, vibration amplitude, pressure change, current fluctuation and voltage fluctuation;

[0022] The failure risk of the equipment is calculated according to the failure risk scoring formula, and the failure risk scoring formula is as follows:

[0023] R=W1·F m +W2·F s +W3·P+W4·C+W5·V+W6·He -λT ,

[0024] W1+W2+W3+W4+W5+W6=1,

[0025]

[0026] Among them, R is the equipment failure risk score, F m is the moving frequency of the device, F s is the vibration characteristic value of the equipment, P is the pressure characteristic value, C is the current characteristic value, V is the voltage characteristic value, H is the equipment maintenance history score, e is a natural constant, λ is the time attenuation coefficient, T is the time from the last maintenance, W1, W2, W3, W4, W5 and W6 are the weight coefficients of the characteristic values ​​respectively, M i is the number of days the equipment is running after the ith maintenance, Q i represents the quality score of the ith maintenance, which is scored by the maintenance personnel, and n is the total number of maintenance times;

[0027] The acquired fault risk is compared with a preset risk threshold. When the fault risk score exceeds the risk threshold, fault warning information is generated and a first-level equipment warning notification is sent to the maintenance end of the maintenance personnel. The first-level equipment warning notification includes the equipment location, ID and fault type.

[0028] Obtain the location data of the maintenance personnel and the equipment at risk of failure. If the two fail to complete the positioning contact within the preset time, a first-level alarm will be issued through the alarm device;

[0029] Obtain the maintenance records and quality scores sent by the maintenance personnel's maintenance terminal, and update the equipment maintenance history database.

[0030] According to the above aspects and any possible implementation, an implementation is further provided, wherein the step S3 identifies the abnormal condition of the material according to the location data of the material, specifically including:

[0031] Clean and normalize the collected material location data to remove noise and outliers;

[0032] Conduct trajectory analysis on the material positioning data to extract the material's movement path and stop points;

[0033] According to the historical positioning data of the material, set the normal movement mode of the material, including the normal movement path, residence time and movement frequency;

[0034] Compare the current material movement path and stop points with the set normal movement pattern to identify abnormal material conditions;

[0035] If a path anomaly is identified, a path anomaly warning message is generated; if a stay time anomaly is identified, a stay time anomaly warning message is generated; if a movement frequency anomaly is identified, a movement frequency anomaly warning message is generated;

[0036] Sending a material level 1 warning notification to the maintenance personnel's maintenance terminal, wherein the material level 1 warning notification includes the material location, ID and abnormality type;

[0037] Obtain the location data of relevant personnel and abnormal materials, monitor the position relationship between relevant personnel and abnormal materials, and if the relevant personnel fail to complete the positioning contact within the preset time, a first-level alarm will be issued through the alarm device;

[0038] Obtain the maintenance records sent by the maintenance personnel's maintenance terminal and update the material exception processing database.

[0039] According to the above aspects and any possible implementation manner, an implementation manner is further provided, wherein the step S3 identifies an abnormal environmental condition according to environmental parameters provided by an environmental sensor, specifically including:

[0040] Obtain the temperature, humidity and / or target gas concentration parameters of the production site through environmental sensors, clean and normalize the collected environmental parameter data, and remove noise and outliers;

[0041] Set the normal range of each environmental parameter based on historical environmental data and standard operating procedures;

[0042] Monitor environmental parameters in real time, compare the real-time environmental parameters with the set normal range, and identify whether there are abnormal environmental conditions;

[0043] When an abnormal environmental condition is identified, abnormal warning information is generated, and an environmental level 1 warning notification is sent to the maintenance end of the maintenance personnel, wherein the environmental level 1 warning notification includes abnormal environmental parameters, location and abnormal type;

[0044] Obtain the location data of the maintenance personnel's maintenance terminal and abnormal environmental parameters;

[0045] Monitor the positional relationship between relevant personnel and abnormal environment sensors. If the maintenance personnel fail to complete the positioning contact within the preset range within the preset time, a first-level alarm will be issued through the alarm device;

[0046] Obtain the maintenance records sent by the maintenance personnel's maintenance terminal and update the environmental parameter exception handling database.

[0047] According to the above aspects and any possible implementation, an implementation is further provided, wherein identifying abnormal events according to monitoring video data in S3 specifically includes:

[0048] Obtain the installation parameters of the camera, select multiple fixed reference points within the camera's viewing angle, obtain the coordinates of the reference points through the Bluetooth tag, establish a two-dimensional monitoring coordinate system in the video image based on the physical space coordinates of the reference points, map the physical space coordinates of the reference points into the video image to form a monitoring coordinate system, and complete the coordinate mapping and positioning of the video and the physical space;

[0049] Acquire real-time monitoring video data from cameras at the production site, and preprocess the collected monitoring video data, wherein the preprocessing includes video denoising and image enhancement;

[0050] Analyze video data to detect suspected flames;

[0051] When a suspected flame is confirmed, a first-level fire warning notification is sent to the mobile terminal of the maintenance personnel, and the first-level fire warning notification includes a suspected flame monitoring image, occurrence time and location;

[0052] Obtain the maintenance personnel's maintenance end and the location data of the suspected flame;

[0053] Monitor the positional relationship between relevant personnel and suspected flames. If the maintenance personnel fail to complete the positioning contact within the preset range within the preset time, a first-level alarm will be issued through the alarm device, and the fire sprinkler heads within the preset range will be opened according to the positioning data of the suspected flame;

[0054] Obtain the maintenance records sent by the maintenance personnel's maintenance terminal and update the surveillance video exception handling database.

[0055] According to the above aspects and any possible implementation manner, an implementation manner is further provided, wherein the detecting whether there is a suspected flame specifically includes:

[0056] Extract the color information of each pixel in the video image and convert the RGB color space into the HSV color space; define the color range of the flame, set the H channel at 0%-50%, the S channel at 50%-100%, and the V channel at 50%-100%; perform color detection on each frame of the image, identify pixels that meet the flame color characteristics, and generate a binary flame candidate area image;

[0057] Analyze the differences between video frames, analyze the pixel motion within the flame candidate region through optical flow method, detect the random flickering and edge jump of the flame, use the background subtraction algorithm to identify the dynamic changes of the flame, narrow the flame candidate region of the flame candidate region image, and exclude non-flame color matching objects;

[0058] Use the SIFT similarity matching algorithm to compare the currently detected flame candidate area with the historical flame image library to determine whether it matches the previous flame event characteristics;

[0059] Different weights are assigned to color detection, dynamic pattern recognition and historical data verification results, and the flame suspected value is calculated based on the weights. The obtained flame suspected value is compared with the pre-set credibility threshold. When the flame suspected value exceeds the credibility threshold, it is judged that there is a suspected flame.

[0060] According to the above aspects and any possible implementation, there is further provided an implementation, wherein after confirming the existence of suspected flame, the method further comprises:

[0061] After confirming the existence of a suspected flame, the location of the suspected flame is obtained, and a range area covering a preset radius is generated as a danger zone;

[0062] Based on the Bluetooth AOA positioning system, determine the list of people located in the current dangerous area, and filter out people who have passed through the current dangerous area based on their historical movement data, and update the list of people to generate a list of dangerous people;

[0063] The list of dangerous persons and evacuation notice are sent to the mobile terminal of the maintenance personnel, and the fire alarm information of the names on the dangerous persons list is broadcast through the alarm device.

[0064] According to the above aspects and any possible implementation manner, an implementation manner is further provided, wherein S4 specifically includes:

[0065] Acquire the collected positioning data, operating status data and / or electrical parameters, temperature, humidity and / or gas concentration parameters, and monitoring video data to synchronize the timestamps of different data sources, perform preliminary processing on different types of data through feature extraction tools, and form a multi-source data set suitable for the input format of the convolutional neural network model based on a multi-input structure;

[0066] Load a pre-trained convolutional neural network model based on a multi-input structure;

[0067] The processed multi-source data set is used as input data and input into a convolutional neural network model based on a multi-input structure;

[0068] In the convolutional neural network model based on a multi-input structure, the input data is subjected to feature extraction through multi-layer convolution operations, and the extracted features are integrated and classified through the fully connected layer of the convolutional neural network model based on a multi-input structure;

[0069] Generate risk identification results based on the output of the convolutional neural network model based on a multi-input structure.

[0070] In a second aspect, an embodiment of the present invention provides an industrial Internet-based data processing system using the above method, including:

[0071] The production data collection module is used to collect the positioning data of personnel, equipment and materials in each link of the production line through the Bluetooth AOA positioning system, collect the operating status data and / or electrical parameters of the equipment through the equipment status sensor, collect the temperature, humidity and / or gas concentration parameters of the production site through the environmental sensor, and collect the monitoring video data of the production site through the camera;

[0072] The data processing module is used to transmit the collected data to the big data center server through the industrial Internet equipment, and to perform cleaning and preprocessing;

[0073] The risk emergency identification module is used to identify the failure risk of the equipment according to the equipment's location data, operating status data and electrical parameters, identify the abnormal conditions of the materials according to the material's location data, identify the abnormal environmental conditions according to the environmental parameters provided by the environmental sensors, identify abnormal events according to the monitoring video data, and send a first-level early warning notification to the maintenance personnel or issue a first-level alarm through the alarm device;

[0074] The comprehensive risk identification module is used to analyze multi-source data through a pre-trained convolutional neural network, conduct comprehensive risk identification, and output identification results;

[0075] The intelligent decision-making module is used to convert the recognition results into vector form and execute fusion intelligent decision-making, which includes sending a first-level warning notification to the maintenance personnel or issuing a first-level alarm through the alarm device, executing a second-level alarm, shutting down and / or reporting to the administrator's management terminal.

[0076] One of the above technical solutions has the following beneficial effects:

[0077] In the method of the embodiment of the present invention, a data processing method and system based on the industrial Internet are proposed. By introducing the Bluetooth AOA positioning system, equipment status sensors, environmental sensors and monitoring video acquisition systems, the present invention realizes comprehensive monitoring of personnel, equipment and materials in all links of the production line, and transmits these multi-source data to the big data center in real time through the industrial Internet platform for cleaning and preprocessing. The multi-source data is deeply analyzed and feature extracted using algorithms and models, which can effectively identify potential risks such as equipment failure, material anomalies, environmental anomalies and fire, and these identification results are comprehensively analyzed through the intelligent decision-making module to generate corresponding alarms and emergency response measures, which not only improves the intelligence level of the industrial Internet system, but also enhances the safety and stability of the production line in a complex environment, and has broad application prospects and practical value.

Brief Description of the Drawings

[0078] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0079] Figure 1 It is a flow chart of a data processing method based on the industrial Internet provided by an embodiment of the present invention;

[0080] Figure 2 It is a functional block diagram of a system provided by an embodiment of the present invention. [Specific implementation method]

[0081] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0082] Please refer to Figure 1 , which is a flow chart of a data processing method based on the industrial Internet provided by an embodiment of the present invention, such as Figure 1 As shown, the method comprises the following steps:

[0083] S1. Collect the positioning data of personnel, equipment and materials in each link of the production line through the Bluetooth AOA positioning system, collect the operating status data and / or electrical parameters of the equipment through the equipment status sensor, collect the temperature, humidity and / or gas concentration parameters of the production site through the environmental sensor, and collect the monitoring video data of the production site through the camera;

[0084] S2, transmit the collected data to the big data center server through industrial Internet devices, and perform cleaning and preprocessing;

[0085] S3. Identify the failure risk of the equipment based on the equipment's location data, operating status data, and electrical parameters; identify the abnormal conditions of the materials based on the material's location data; identify the abnormal environmental conditions based on the environmental parameters provided by the environmental sensors; identify abnormal events based on the monitoring video data; and send a first-level warning notification to the maintenance personnel or issue a first-level alarm through the alarm device;

[0086] S4, analyze the multi-source data through the pre-trained convolutional neural network, conduct comprehensive risk identification, and output the identification results;

[0087] S5. Convert the recognition result into vector form and execute fusion intelligent decision, which includes sending a first-level warning notification to the maintenance personnel or issuing a first-level alarm through the alarm device, executing a second-level alarm, shutting down and / or reporting to the management terminal of the administrator.

[0088] The embodiment of the present invention can efficiently collect personnel, equipment, material and environmental information in each link of the production line through various sensing devices such as Bluetooth AOA positioning system, equipment status sensor, environmental sensor and camera. The collected data is transmitted to the big data center through industrial Internet equipment. After cleaning and preprocessing, the noise and outliers in the data are eliminated, thereby ensuring the accuracy and reliability of subsequent data analysis. By comprehensively analyzing the positioning data, operating status data and electrical parameters of the equipment, the risk of equipment failure can be identified in real time. At the same time, by analyzing the trajectory of material positioning data and monitoring environmental parameters, material anomalies and environmental abnormalities can be quickly identified. In addition, combined with the analysis of monitoring video data, it can effectively detect fire safety hazards at the production site, and send early warning notifications to maintenance personnel or trigger alarm equipment in time, thereby improving the safety of the production line. The pre-trained convolutional neural network model can be used for multi-source The data is deeply analyzed, key features are extracted and comprehensive risk identification is performed. Through the intelligent analysis of the model, the present invention can accurately judge the potential various risks in the production line and output the identification results, providing a reliable basis for subsequent decision-making. Moreover, the designed convolutional neural network model has adaptive learning ability, and can continuously optimize and adjust the risk identification model according to the changes in the actual production environment and historical data; the identification results are converted into quantified risk vectors, and by integrating the intelligent decision-making module, corresponding emergency response measures are executed according to the risk assessment results, including secondary alarm, shutdown processing and reporting to the management end. This intelligent decision-making process not only improves the response speed of the production line to emergencies, but also effectively reduces the losses caused by misjudgment or delayed processing; in addition, the present invention has good scalability and can easily integrate other types of sensing equipment or analysis tools to adapt to production line application scenarios of different scales and complexities.

[0089] In a preferred embodiment, the S1 specifically includes:

[0090] Bluetooth AOA positioning base stations are pre-deployed at the production site to collect positioning signals sent by Bluetooth tags installed on personnel, equipment and materials at various stages of the production line;

[0091] Installing a device status sensor on the device in advance, wherein the device status sensor includes a vibration sensor, a pressure sensor, a current sensor and / or a voltage sensor to monitor the vibration, internal pressure, current and / or voltage data of the device;

[0092] Pre-arrange environmental sensors at the production site, wherein the environmental sensors include temperature sensors, humidity sensors and / or gas sensors for collecting temperature, humidity and / or target gas concentration parameters of the production site;

[0093] Arrange cameras in advance at the production site to collect monitoring video data of the production site through the cameras;

[0094] Pre-configure mobile terminals for maintenance personnel;

[0095] The industrial Internet device is used to receive data transmitted by the Bluetooth AOA positioning base station, equipment status sensor, environmental sensor, camera and mobile terminal.

[0096] The preferred embodiment of the present invention realizes comprehensive monitoring and data collection of all links of the production line by pre-arranging various sensor devices and positioning base stations in the production site, greatly improving the intelligent monitoring capability of the production line, realizing accurate positioning, comprehensive status monitoring and efficient data management, and ensuring the safety and operation efficiency of the production line.

[0097] In a preferred embodiment, the step S3 identifies the failure risk of the device according to the positioning data, operating status data and electrical parameters of the device, specifically including:

[0098] Clean and normalize the collected data to remove noise and outliers, and fuse the cleaned data to form a comprehensive operation data set of the equipment;

[0099] Perform feature extraction on the comprehensive operation data set of the equipment to extract key feature values, including key features of movement frequency, vibration amplitude, pressure change, current fluctuation and voltage fluctuation;

[0100] The failure risk of the equipment is calculated according to the failure risk scoring formula, and the failure risk scoring formula is as follows:

[0101] R=W1·F m +W2·F s +W3·P+W4·C+W5·V+W6·He -λT ,

[0102] W1+W2+W3+W4+W5+W6=1,

[0103]

[0104] Among them, R is the equipment failure risk score, F m is the moving frequency of the device, F sis the vibration characteristic value of the equipment, P is the pressure characteristic value, C is the current characteristic value, V is the voltage characteristic value, H is the equipment maintenance history score, e is a natural constant, λ is the time attenuation coefficient, T is the time from the last maintenance, W1, W2, W3, W4, W5 and W6 are the weight coefficients of the characteristic values ​​respectively, M i is the number of days the equipment is running after the ith maintenance, Q i represents the quality score of the ith maintenance, which is scored by the maintenance personnel, and n is the total number of maintenance times;

[0105] The acquired fault risk is compared with a preset risk threshold. When the fault risk score exceeds the risk threshold, fault warning information is generated and a first-level equipment warning notification is sent to the maintenance end of the maintenance personnel. The first-level equipment warning notification includes the equipment location, ID and fault type.

[0106] Obtain the location data of the maintenance personnel and the equipment at risk of failure. If the two fail to complete the positioning contact within the preset time, a first-level alarm will be issued through the alarm device;

[0107] Obtain the maintenance records and quality scores sent by the maintenance personnel's maintenance terminal, and update the equipment maintenance history database.

[0108] The embodiment of the present invention effectively removes noise and outliers by cleaning and normalizing the collected data, thereby ensuring the accuracy and consistency of the data; through the fault risk scoring formula, multiple key features can be quantified and comprehensively analyzed. The scoring formula fully considers the movement frequency, vibration characteristics, pressure, current and voltage changes of the equipment, the equipment maintenance history score and the time decay effect, so as to calculate the comprehensive fault risk score of the equipment. The scoring method is based on scientific weighted analysis and can accurately evaluate the fault risk level of the equipment; by comparing with the preset risk threshold, when the fault risk score exceeds the threshold, the fault warning information can be generated in time and sent to the maintenance personnel. The first-level warning notification not only includes the specific location, ID and fault type of the equipment, but also can be tracked in real time through the positioning data of the maintenance personnel and the faulty equipment. If the positioning contact is not completed within the preset time, the present invention will automatically trigger the first-level alarm mechanism to ensure that the equipment fault is quickly responded to and handled; by obtaining the maintenance records and quality scores of the maintenance personnel, the equipment maintenance history database is continuously updated, and by recording the equipment operation days and maintenance quality scores after each maintenance, the equipment maintenance history database can dynamically reflect the maintenance status and historical performance of the equipment, and provide important reference data for future fault risk assessment. The present invention realizes multi-dimensional status monitoring, accurate risk assessment, timely early warning and effective alarm mechanism in terms of equipment failure risk identification. Through scientific feature extraction and risk scoring methods, it greatly improves the intelligent level of equipment management, can significantly reduce the operational risks brought by equipment failure, and improve the overall operating efficiency and safety of the production line.

[0109] In a preferred embodiment, the step S3 of identifying abnormal conditions of materials according to the material location data specifically includes:

[0110] Clean and normalize the collected material location data to remove noise and outliers;

[0111] Conduct trajectory analysis on the material positioning data to extract the material's movement path and stop points;

[0112] According to the historical positioning data of the material, set the normal movement mode of the material, including the normal movement path, residence time and movement frequency;

[0113] Compare the current material movement path and stop points with the set normal movement pattern to identify abnormal material conditions;

[0114] If a path anomaly is identified, a path anomaly warning message is generated; if a stay time anomaly is identified, a stay time anomaly warning message is generated; if a movement frequency anomaly is identified, a movement frequency anomaly warning message is generated;

[0115] Sending a material level 1 warning notification to the maintenance personnel's maintenance terminal, wherein the material level 1 warning notification includes the material location, ID and abnormality type;

[0116] Obtain the location data of relevant personnel and abnormal materials, monitor the position relationship between relevant personnel and abnormal materials, and if the relevant personnel fail to complete the positioning contact within the preset time, a first-level alarm will be issued through the alarm device;

[0117] Obtain the maintenance records sent by the maintenance personnel's maintenance terminal and update the material exception processing database.

[0118] The present invention realizes efficient data processing, accurate trajectory analysis and anomaly identification based on historical patterns in identifying abnormal conditions of materials. The trajectory analysis can not only track the real-time position of the material, but also reflect the movement state and behavior characteristics of the material in the production line. The anomaly identification method based on historical data can better adapt to the normal behavior pattern of the material, thereby reducing the possibility of false alarms and significantly improving the intelligence and automation level of material management. Through timely early warning and alarm mechanisms, the present invention effectively reduces the production risks caused by material anomalies, ensures the stable operation of the production line, and has broad industrial application value.

[0119] In a preferred embodiment, the step S3 of identifying abnormal environmental conditions according to environmental parameters provided by environmental sensors specifically includes:

[0120] Obtain the temperature, humidity and / or target gas concentration parameters of the production site through environmental sensors, clean and normalize the collected environmental parameter data, and remove noise and outliers;

[0121] Set the normal range of each environmental parameter based on historical environmental data and standard operating procedures;

[0122] Monitor environmental parameters in real time, compare the real-time environmental parameters with the set normal range, and identify whether there are abnormal environmental conditions;

[0123] When an abnormal environmental condition is identified, abnormal warning information is generated, and an environmental level 1 warning notification is sent to the maintenance end of the maintenance personnel, wherein the environmental level 1 warning notification includes abnormal environmental parameters, location and abnormal type;

[0124] Obtain the location data of the maintenance personnel's maintenance terminal and abnormal environmental parameters;

[0125] Monitor the positional relationship between relevant personnel and abnormal environment sensors. If the maintenance personnel fail to complete the positioning contact within the preset range within the preset time, a first-level alarm will be issued through the alarm device;

[0126] Obtain the maintenance records sent by the maintenance personnel's maintenance terminal and update the environmental parameter exception handling database.

[0127] The present invention realizes efficient data processing, accurate real-time monitoring and timely abnormal response in identifying abnormal environmental conditions, significantly improving the environmental safety management capabilities of production sites. Through real-time early warning and intelligent alarm mechanisms, the present invention effectively reduces the impact of environmental anomalies on the production process and ensures the continuity and safety of production.

[0128] In a preferred embodiment, the step S3 of identifying abnormal events according to the monitoring video data specifically includes:

[0129] Obtain the installation parameters of the camera, select multiple fixed reference points within the camera's viewing angle, obtain the coordinates of the reference points through the Bluetooth tag, establish a two-dimensional monitoring coordinate system in the video image based on the physical space coordinates of the reference points, map the physical space coordinates of the reference points into the video image to form a monitoring coordinate system, and complete the coordinate mapping and positioning of the video and the physical space;

[0130] Acquire real-time monitoring video data from cameras at the production site, and preprocess the collected monitoring video data, wherein the preprocessing includes video denoising and image enhancement;

[0131] Analyze video data to detect suspected flames;

[0132] When a suspected flame is confirmed, a first-level fire warning notification is sent to the mobile terminal of the maintenance personnel, and the first-level fire warning notification includes a suspected flame monitoring image, occurrence time and location;

[0133] Obtain the maintenance personnel's maintenance end and the location data of the suspected flame;

[0134] Monitor the positional relationship between relevant personnel and suspected flames. If the maintenance personnel fail to complete the positioning contact within the preset range within the preset time, a first-level alarm will be issued through the alarm device, and the fire sprinkler heads within the preset range will be opened according to the positioning data of the suspected flame;

[0135] Obtain the maintenance records sent by the maintenance personnel's maintenance terminal and update the surveillance video exception handling database.

[0136] The present invention obtains the installation parameters of the camera, selects multiple fixed reference points within the camera's field of view, uses Bluetooth tags to obtain the physical space coordinates of the reference points, and establishes a two-dimensional monitoring coordinate system in the video image. This process maps the precise coordinates of the physical space to the video image, achieves a seamless combination of video and physical space, and ensures the accuracy and real-time nature of subsequent video data analysis; preprocesses the monitoring video data acquired in real time, including video denoising and image enhancement operations, which can significantly improve the quality of video data by removing noise in the video and enhancing the clarity of the image, thereby ensuring the video monitoring effect in a complex production environment and laying the foundation for the accurate detection of abnormal events; uses advanced image analysis technology to conduct in-depth analysis of the preprocessed video data to detect whether there is a suspected flame in the video. When the system confirms the presence of a suspected flame, it can immediately send a notification to the mobile terminal of the maintenance personnel. A first-level fire warning notification is sent, which includes the monitoring image of the suspected flame, the time of occurrence and the accurate location, to help maintenance personnel quickly understand and respond to fire risks; by obtaining the maintenance personnel's mobile terminal and the location data of the suspected flame, the position relationship between the relevant personnel and the suspected flame can be monitored in real time. When it is detected that the maintenance personnel fails to enter the positioning contact area within the preset range within the specified time, the system will automatically trigger a first-level alarm and automatically activate the fire sprinkler heads within the corresponding range according to the flame's location data. This function significantly improves the timeliness and accuracy of fire emergency response and reduces the potential losses caused by fire; by obtaining the maintenance records of the maintenance personnel and continuously updating the monitoring video anomaly processing database, the database records the processing process and results of all video anomaly events, which can provide important reference data for future anomaly detection and processing, help optimize the system's response strategy and improve the overall response efficiency.

[0137] In a preferred embodiment, the detecting whether there is a suspected flame specifically includes:

[0138] Extract the color information of each pixel in the video image and convert the RGB color space into the HSV color space; define the color range of the flame, set the H channel at 0%-50%, the S channel at 50%-100%, and the V channel at 50%-100%; perform color detection on each frame of the image, identify pixels that meet the flame color characteristics, and generate a binary flame candidate area image;

[0139] Analyze the differences between video frames, analyze the pixel motion within the flame candidate region through optical flow method, detect the random flickering and edge jump of the flame, use the background subtraction algorithm to identify the dynamic changes of the flame, narrow the flame candidate region of the flame candidate region image, and exclude non-flame color matching objects;

[0140] Use the SIFT similarity matching algorithm to compare the currently detected flame candidate area with the historical flame image library to determine whether it matches the previous flame event characteristics;

[0141] Different weights are assigned to color detection, dynamic pattern recognition and historical data verification results, and the flame suspected value is calculated based on the weights. The obtained flame suspected value is compared with the pre-set credibility threshold. When the flame suspected value exceeds the credibility threshold, it is judged that there is a suspected flame.

[0142] By converting each pixel in the video image from the RGB color space to the HSV color space and screening the H, S, and V channels according to the set flame color range, the present invention can quickly identify pixels that meet the flame color characteristics and generate a binary flame candidate area image. This color detection method not only improves the efficiency of flame detection, but also effectively reduces false alarms caused by lighting changes, thereby ensuring the accuracy of the flame candidate area; the optical flow method is used to analyze the pixel movement in the flame candidate area, detect the random flickering and edge jump of the flame, further narrow the flame candidate area, and combine the background subtraction algorithm to identify the dynamic changes of the flame, which can effectively exclude non-flame color matching objects, such as reflections and light spots. This multi-level dynamic feature analysis improves the accuracy of flame detection and reduces the possibility of false alarms; the SIFT similarity matching algorithm is used to compare the currently detected flame candidate area with the historical flame image library. By extracting key feature points in the image and judging the similarity between the current flame and the historical flame event features, the authenticity of the flame can be further confirmed. This method utilizes the accumulation of historical data to improve the reliability and accuracy of flame detection. By assigning different weights to color detection, dynamic pattern recognition and historical data verification results, the flame suspected value is calculated by comprehensively considering multiple detection results. When the flame suspected value exceeds the preset credibility threshold, the system automatically determines that there is a suspected flame. This weight-based comprehensive judgment method can balance the advantages and disadvantages of each detection method and improve the reliability of the overall detection. When a suspected flame is detected, a fire warning notification can be triggered immediately and responded through relevant mechanisms. The system can send fire warning information to maintenance personnel through the mobile terminal, including the monitoring image of the flame, the time of occurrence and the specific location, which effectively improves the timeliness and accuracy of fire emergency response and minimizes the potential harm of fire to the production environment.

[0143] In a preferred embodiment, after confirming the existence of suspected flame, the method further comprises:

[0144] After confirming the existence of a suspected flame, the location of the suspected flame is obtained, and a range area covering a preset radius is generated as a danger zone;

[0145] Based on the Bluetooth AOA positioning system, determine the list of people located in the current dangerous area, and filter out people who have passed through the current dangerous area based on their historical movement data, and update the list of people to generate a list of dangerous people;

[0146] The list of dangerous persons and evacuation notice are sent to the mobile terminal of the maintenance personnel, and the fire alarm information of the names on the dangerous persons list is broadcast through the alarm device.

[0147] After confirming the existence of a suspected flame, the present invention can quickly obtain the accurate location of the flame, and generate a danger zone covering the flame based on a preset radius. By accurately locating the location of the flame, the scope of the danger zone can be timely delineated to ensure that the relevant area can be effectively monitored and managed before the fire risk spreads, thereby protecting the safety of the production site to the greatest extent; using the Bluetooth AOA positioning system, it can accurately determine all personnel currently in the danger zone, and screen out personnel who have passed through the area based on the historical movement data of the personnel, and generate a list of dangerous personnel by updating the list of personnel. The system can grasp the situation of personnel in the danger zone and those who may be endangered in real time, and provide protection for fire hazard prevention; after the list of dangerous personnel is generated, it can immediately Send evacuation notices to the maintenance personnel's mobile terminals, and broadcast fire alarm information of the names of people on the dangerous personnel list through the alarm device, ensuring that when a possible fire occurs, relevant personnel can quickly receive the alarm information and evacuate the dangerous area in time, greatly improving the efficiency of fire emergency response, effectively reducing the risk of casualties and property losses, and avoiding confusion and production impact caused by false alarms; through real-time positioning of personnel in the dangerous area and historical data analysis, the personnel list can be dynamically updated to ensure that anyone who enters or has passed through the dangerous area can be identified and notified in time. Through the real-time broadcast function of the alarm device, the system can transmit the fire alarm information to the relevant personnel in the shortest time, further enhancing the response speed and effectiveness of emergency management.

[0148] In a preferred embodiment, the S4 specifically includes:

[0149] Acquire the collected positioning data, operating status data and / or electrical parameters, temperature, humidity and / or gas concentration parameters, and monitoring video data to synchronize the timestamps of different data sources, perform preliminary processing on different types of data through feature extraction tools, and form a multi-source data set suitable for the input format of the convolutional neural network model based on a multi-input structure;

[0150] Load a pre-trained convolutional neural network model based on a multi-input structure;

[0151] The processed multi-source data set is used as input data and input into a convolutional neural network model based on a multi-input structure;

[0152] In the convolutional neural network model based on a multi-input structure, the input data is subjected to feature extraction through multi-layer convolution operations, and the extracted features are integrated and classified through the fully connected layer of the convolutional neural network model based on a multi-input structure;

[0153] Generate risk identification results based on the output of the convolutional neural network model based on a multi-input structure.

[0154] The embodiment of the present invention uses a feature extraction tool to perform preliminary processing on different types of data to form a multi-source data set suitable for convolutional neural network model analysis, thereby ensuring the accuracy and consistency of the data; the convolutional neural network model based on a multi-input structure is capable of processing multiple data inputs at the same time, and performing deep feature extraction on the input data through multi-layer convolution operations. The model can extract key features from time series data, image data, and environmental data, achieve a multi-dimensional understanding of complex production environments, and improve the accuracy and depth of risk identification; it can accurately judge the operating status of the production line and generate corresponding risk identification results. Through this intelligent analysis process, the present invention greatly improves the safety and operational efficiency of the production line, and provides managers with timely and accurate risk warnings and decision-making basis for countermeasures; the convolutional neural network model of the present invention has good scalability and adaptability, can continuously optimize the analysis model according to changes in the actual production environment, and adapt to different types of data inputs.

[0155] The convolutional neural network model based on the multi-input structure of the present invention is designed as follows:

[0156] 1. Data input type: Positioning data: real-time location data of personnel, equipment and materials collected by Bluetooth AOA positioning base stations deployed in the production site; Equipment status data: parameters such as equipment vibration, internal pressure, current and voltage collected by vibration sensors, pressure sensors, current sensors and voltage sensors; Environmental parameter data: including temperature, humidity and gas concentration parameters, collected by environmental sensors deployed in the production site; Image data: monitoring video frame data from cameras in the production site.

[0157] 2. Data synchronization and preliminary processing:

[0158] Synchronize the timestamps of different data sources to ensure that all data are analyzed on the same timeline; use feature extraction tools to perform preliminary data processing, including normalization, noise reduction, and outlier removal; form a multi-source data set suitable for input into the CNN model, including converting time series data and image data into the standard input format required by CNN.

[0159] 3. Model structure:

[0160] Input layer: Input 1 (positioning data): positioning data collected by Bluetooth AOA positioning base station, shape (T, F1), T is the number of time steps, F1 is the positioning feature dimension; Input 2 (device status data): data from vibration, pressure, current and voltage sensors, shape (T, F2); Input 3 (environmental parameter data): temperature, humidity and gas concentration data, shape (T, F3); Input 4 (image data): monitoring video frame from the camera, shape (H, W, C), H is the image height, W is the image width, C is the number of channels;

[0161] Convolutional layer and pooling layer:

[0162] Positioning data channel: Convolution layer 1: 1D convolution, convolution kernel size is 3, number is 64, step size is 1, padding method is the same, activation function is ReLU; Pooling layer 1: 1D maximum pooling, pooling window size is 2, step size is 2; Output shape: the output shape after pooling is (T / 2, 64);

[0163] Device status data channel: Convolution layer 2: 1D convolution, convolution kernel size is 3, number is 64, step size is 1, padding method is the same, activation function is ReLU; Pooling layer 2: 1D maximum pooling, pooling window size is 2, step size is 2; Output shape: the output shape after pooling is (T / 2, 64);

[0164] Environmental parameter data channel: Convolution layer 3: 1D convolution, convolution kernel size is 3, number is 32, step size is 1, padding method is the same, activation function is ReLU; Pooling layer 3: 1D maximum pooling, pooling window size is 2, step size is 2; Output shape: The output shape after pooling is (T / 2, 32);

[0165] Image data channel: Convolution layer 4: 2D convolution, convolution kernel size is 3x3, number is 32, step size is 1x1, padding method is the same, activation function is ReLU; Pooling layer 4: 2D maximum pooling, pooling window size is 2x2, step size is 2x2; Output shape: The output shape after pooling is (H / 2, W / 2, 32);

[0166] Feature fusion and fully connected layer:

[0167] Flatten the pooled output of each channel into a one-dimensional vector: positioning data channel: the flattened dimension is (T / 2)*64; device status data channel: the flattened dimension is (T / 2)*64; environmental parameter data channel: the flattened dimension is (T / 2)*32; image data channel: the flattened dimension is (H / 2)*(W / 2)*32;

[0168] Feature fusion layer: connect all flattened vectors into a large vector: the fused dimension is: (T / 2)*(64+64+32)+(H / 2)*(W / 2)*32;

[0169] Fully connected layer 1: fully connected layer, output dimension is 256, activation function is ReLU;

[0170] Fully connected layer 2: fully connected layer, output dimension is 128, activation function is ReLU;

[0171] Final output layer: The output dimension is 4, the activation function is Sigmoid, and it is used for binary classification tasks.

[0172] 4. Model training and optimization:

[0173] Loss function: use cross entropy loss, optimizer: use Adam optimizer, and the initial learning rate is set to 0.001;

[0174] Training and validation: The dataset is divided into a training set and a validation set, usually 80% training data and 20% validation data. An early stopping mechanism is used to prevent overfitting, and model parameters are optimized through cross-validation.

[0175] It should be noted that the convolutional neural network model based on a multi-input structure of the present invention is not limited to the above-mentioned design, and other existing neural network models can also be adopted as needed.

[0176] The comprehensive risk feature vector is nonlinearly transformed using the Sigmoid activation function, and the output recognition result is a risk probability vector, each value corresponds to the recognition result of a risk category, and different weights are assigned to each risk category according to the probability value of each risk category, and the sum value is calculated, and the sum value is compared with different level thresholds. According to the comparison result, a fusion intelligent decision is executed, and the intelligent decision includes sending a first-level early warning notification to the maintenance personnel or issuing a first-level alarm through the alarm device, executing a second-level alarm, shutting down, and / or reporting to the management end of the manager;

[0177] Specifically, when the comprehensive risk score does not exceed the low risk threshold, no action is required; when the comprehensive risk score exceeds the low risk threshold but does not exceed the medium risk threshold, a first-level warning notification is sent to the maintenance personnel or a first-level alarm is issued through the alarm device; when the comprehensive risk score exceeds the medium risk threshold but does not exceed the highest risk threshold, a second-level alarm is triggered and reported to the manager's management end; when the risk score exceeds the highest threshold, an emergency shutdown is performed and automatically reported to the manager's management end.

[0178] The embodiments of the present invention further provide device embodiments for implementing the steps and methods in the above method embodiments.

[0179] The embodiment of the present invention provides a data processing system based on the industrial Internet using the above method. Figure 2 FIG. 1 is a functional block diagram of a system provided by an embodiment of the present invention, wherein the system includes:

[0180] The production data collection module is used to collect the positioning data of personnel, equipment and materials in each link of the production line through the Bluetooth AOA positioning system, collect the operating status data and / or electrical parameters of the equipment through the equipment status sensor, collect the temperature, humidity and / or gas concentration parameters of the production site through the environmental sensor, and collect the monitoring video data of the production site through the camera;

[0181] The data processing module is used to transmit the collected data to the big data center server through the industrial Internet equipment, and to perform cleaning and preprocessing;

[0182] The risk emergency identification module is used to identify the failure risk of the equipment according to the equipment's location data, operating status data and electrical parameters, identify the abnormal conditions of the materials according to the material's location data, identify the abnormal environmental conditions according to the environmental parameters provided by the environmental sensors, identify abnormal events according to the monitoring video data, and send a first-level early warning notification to the maintenance personnel or issue a first-level alarm through the alarm device;

[0183] The comprehensive risk identification module is used to analyze multi-source data through a pre-trained convolutional neural network, conduct comprehensive risk identification, and output identification results;

[0184] The intelligent decision-making module is used to convert the recognition results into vector form and execute fusion intelligent decision-making, which includes sending a first-level warning notification to the maintenance personnel or issuing a first-level alarm through the alarm device, executing a second-level alarm, shutting down and / or reporting to the administrator's management terminal.

[0185] An embodiment of the present invention also proposes a computer-readable storage medium, which stores one or more programs, and the one or more programs include instructions, which, when executed by a server including multiple applications, can enable the server to execute the node working method provided in any embodiment of the present invention.

[0186] The systems, devices, modules or units described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0187] For the convenience of description, the above device is described as various units or modules according to their functions. Of course, when implementing the present invention, the functions of each unit or module can be implemented in the same or multiple software and / or hardware.

[0188] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0189] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0190] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0191] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0192] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0193] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0194] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0195] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0196] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0197] The present invention may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present invention may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in local and remote computer storage media, including storage devices.

[0198] Each embodiment of the present invention is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0199] The above description is only an embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent substitution, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of the claims of the present invention.

Claims

1. A data processing method based on industrial Internet, characterized in that: The method comprises: S1. Collect the positioning data of personnel, equipment and materials in each link of the production line through the Bluetooth AOA positioning system, collect the operating status data and / or electrical parameters of the equipment through the equipment status sensor, collect the temperature, humidity and / or gas concentration parameters of the production site through the environmental sensor, and collect the monitoring video data of the production site through the camera; S2, transmit the collected data to the big data center server through industrial Internet devices, and perform cleaning and preprocessing; S3. Identify the failure risk of the equipment based on the equipment's location data, operating status data, and electrical parameters; identify the abnormal conditions of the materials based on the material's location data; identify the abnormal environmental conditions based on the environmental parameters provided by the environmental sensors; identify abnormal events based on the monitoring video data; and send a first-level warning notification to the maintenance personnel or issue a first-level alarm through the alarm device; S4, analyze the multi-source data through the pre-trained convolutional neural network, conduct comprehensive risk identification, and output the identification results; S5. Convert the recognition result into vector form and execute fusion intelligent decision, which includes sending a first-level warning notification to the maintenance personnel or issuing a first-level alarm through the alarm device, executing a second-level alarm, shutting down and / or reporting to the management terminal of the administrator.

2. The method according to claim 1, characterized in that The S1 specifically includes: Bluetooth AOA positioning base stations are pre-deployed at the production site to collect positioning signals sent by Bluetooth tags installed on personnel, equipment and materials at various stages of the production line; Installing a device status sensor on the device in advance, wherein the device status sensor includes a vibration sensor, a pressure sensor, a current sensor and / or a voltage sensor to monitor the vibration, internal pressure, current and / or voltage data of the device; Pre-arrange environmental sensors at the production site, wherein the environmental sensors include temperature sensors, humidity sensors and / or gas sensors for collecting temperature, humidity and / or target gas concentration parameters of the production site; Arrange cameras in advance at the production site to collect monitoring video data of the production site through the cameras; Pre-configure mobile terminals for maintenance personnel; The industrial Internet device is used to receive data transmitted by the Bluetooth AOA positioning base station, equipment status sensor, environmental sensor, camera and mobile terminal.

3. The method according to claim 2, characterized in that In S3, the failure risk of the equipment is identified according to the positioning data, operating status data and electrical parameters of the equipment, specifically including: Clean and normalize the collected data to remove noise and outliers, and fuse the cleaned data to form a comprehensive operation data set of the equipment; Perform feature extraction on the comprehensive operation data set of the equipment to extract key feature values, including key features of movement frequency, vibration amplitude, pressure change, current fluctuation and voltage fluctuation; The failure risk of the equipment is calculated according to the failure risk scoring formula, and the failure risk scoring formula is as follows: R=W1·F m +W2·F s +W3·P+W4·C+W5·V+W6·He -λT , W1+W2+W3+W4+W5+W6=1, Among them, R is the equipment failure risk score, F m is the moving frequency of the device, F s is the vibration characteristic value of the equipment, P is the pressure characteristic value, C is the current characteristic value, V is the voltage characteristic value, H is the equipment maintenance history score, e is a natural constant, λ is the time attenuation coefficient, T is the time from the last maintenance, W1, W2, W3, W4, W5 and W6 are the weight coefficients of the characteristic values ​​respectively, M i is the number of days the equipment is running after the ith maintenance, Q i represents the quality score of the ith maintenance, which is scored by the maintenance personnel, and n is the total number of maintenance times; The acquired fault risk is compared with a preset risk threshold. When the fault risk score exceeds the risk threshold, fault warning information is generated and a first-level equipment warning notification is sent to the maintenance end of the maintenance personnel. The first-level equipment warning notification includes the equipment location, ID and fault type. Obtain the location data of the maintenance personnel and the equipment at risk of failure. If the two fail to complete the positioning contact within the preset time, a first-level alarm will be issued through the alarm device; Obtain the maintenance records and quality scores sent by the maintenance personnel's maintenance terminal, and update the equipment maintenance history database.

4. The method according to claim 3, characterized in that In S3, the abnormal condition of the material is identified according to the location data of the material, which specifically includes: Clean and normalize the collected material location data to remove noise and outliers; Conduct trajectory analysis on the material positioning data to extract the material's movement path and stop points; According to the historical positioning data of the material, set the normal movement mode of the material, including the normal movement path, residence time and movement frequency; Compare the current material movement path and stop points with the set normal movement pattern to identify abnormal material conditions; If a path anomaly is identified, a path anomaly warning message is generated; if a stay time anomaly is identified, a stay time anomaly warning message is generated; if a movement frequency anomaly is identified, a movement frequency anomaly warning message is generated; Sending a material level 1 warning notification to the maintenance personnel's maintenance terminal, wherein the material level 1 warning notification includes the material location, ID and abnormality type; Obtain the location data of relevant personnel and abnormal materials, monitor the position relationship between relevant personnel and abnormal materials, and if the relevant personnel fail to complete the positioning contact within the preset time, a first-level alarm will be issued through the alarm device; Obtain the maintenance records sent by the maintenance personnel's maintenance terminal and update the material exception processing database.

5. The method according to claim 4, characterized in that In S3, identifying abnormal environmental conditions according to environmental parameters provided by environmental sensors specifically includes: Obtain the temperature, humidity and / or target gas concentration parameters of the production site through environmental sensors, clean and normalize the collected environmental parameter data, and remove noise and outliers; Set the normal range of each environmental parameter based on historical environmental data and standard operating procedures; Monitor environmental parameters in real time, compare the real-time environmental parameters with the set normal range, and identify whether there are abnormal environmental conditions; When an abnormal environmental condition is identified, abnormal warning information is generated, and an environmental level 1 warning notification is sent to the maintenance end of the maintenance personnel, wherein the environmental level 1 warning notification includes abnormal environmental parameters, location and abnormal type; Obtain the location data of the maintenance personnel's maintenance terminal and abnormal environmental parameters; Monitor the positional relationship between relevant personnel and abnormal environment sensors. If the maintenance personnel fail to complete the positioning contact within the preset range within the preset time, a first-level alarm will be issued through the alarm device; Obtain the maintenance records sent by the maintenance personnel's maintenance terminal and update the environmental parameter exception handling database.

6. The method according to claim 5, characterized in that The step S3 of identifying abnormal events based on the monitoring video data specifically includes: Obtain the installation parameters of the camera, select multiple fixed reference points within the camera's viewing angle, obtain the coordinates of the reference points through the Bluetooth tag, establish a two-dimensional monitoring coordinate system in the video image based on the physical space coordinates of the reference points, map the physical space coordinates of the reference points into the video image to form a monitoring coordinate system, and complete the coordinate mapping and positioning of the video and the physical space; Acquire real-time monitoring video data from cameras at the production site, and preprocess the collected monitoring video data, wherein the preprocessing includes video denoising and image enhancement; Analyze video data to detect suspected flames; When a suspected flame is confirmed, a first-level fire warning notification is sent to the mobile terminal of the maintenance personnel, and the first-level fire warning notification includes a suspected flame monitoring image, occurrence time and location; Obtain the maintenance personnel's maintenance end and the location data of the suspected flame; Monitor the positional relationship between relevant personnel and suspected flames. If the maintenance personnel fail to complete the positioning contact within the preset range within the preset time, a first-level alarm will be issued through the alarm device, and the fire sprinkler heads within the preset range will be opened according to the positioning data of the suspected flame; Obtain the maintenance records sent by the maintenance personnel's maintenance terminal and update the surveillance video exception handling database.

7. The method according to claim 6, characterized in that The detecting whether there is a suspected flame specifically includes: Extract the color information of each pixel in the video image and convert the RGB color space into the HSV color space; define the color range of the flame, set the H channel at 0%-50%, the S channel at 50%-100%, and the V channel at 50%-100%; perform color detection on each frame of the image, identify pixels that meet the flame color characteristics, and generate a binary flame candidate area image; Analyze the differences between video frames, analyze the pixel motion within the flame candidate region through optical flow method, detect the random flickering and edge jump of the flame, use the background subtraction algorithm to identify the dynamic changes of the flame, narrow the flame candidate region of the flame candidate region image, and exclude non-flame color matching objects; Use the SIFT similarity matching algorithm to compare the currently detected flame candidate area with the historical flame image library to determine whether it matches the previous flame event characteristics; Different weights are assigned to color detection, dynamic pattern recognition and historical data verification results, and the flame suspected value is calculated based on the weights. The obtained flame suspected value is compared with the pre-set credibility threshold. When the flame suspected value exceeds the credibility threshold, it is judged that there is a suspected flame.

8. The method according to claim 6, characterized in that The process of confirming the existence of suspected flames also includes: After confirming the existence of a suspected flame, the location of the suspected flame is obtained, and a range area covering a preset radius is generated as a danger zone; Based on the Bluetooth AOA positioning system, determine the list of people located in the current dangerous area, and filter out people who have passed through the current dangerous area based on their historical movement data, and update the list of people to generate a list of dangerous people; The list of dangerous persons and evacuation notice are sent to the mobile terminal of the maintenance personnel, and the fire alarm information of the names on the dangerous persons list is broadcast through the alarm device.

9. The method according to claim 1, characterized in that: The S4 specifically includes: Acquire the collected positioning data, operating status data and / or electrical parameters, temperature, humidity and / or gas concentration parameters, and monitoring video data to synchronize the timestamps of different data sources, perform preliminary processing on different types of data through feature extraction tools, and form a multi-source data set suitable for the input format of the convolutional neural network model based on a multi-input structure; Load a pre-trained convolutional neural network model based on a multi-input structure; The processed multi-source data set is used as input data and input into a convolutional neural network model based on a multi-input structure; In the convolutional neural network model based on a multi-input structure, the input data is subjected to feature extraction through multi-layer convolution operations, and the extracted features are integrated and classified through the fully connected layer of the convolutional neural network model based on a multi-input structure; Generate risk identification results based on the output of the convolutional neural network model based on a multi-input structure.

10. An industrial Internet-based data processing system using the method according to any one of claims 1 to 9, characterized in that: include: The production data collection module is used to collect the positioning data of personnel, equipment and materials in each link of the production line through the Bluetooth AOA positioning system, collect the operating status data and / or electrical parameters of the equipment through the equipment status sensor, collect the temperature, humidity and / or gas concentration parameters of the production site through the environmental sensor, and collect the monitoring video data of the production site through the camera; The data processing module is used to transmit the collected data to the big data center server through the industrial Internet equipment, and to perform cleaning and preprocessing; The risk emergency identification module is used to identify the failure risk of the equipment according to the equipment's location data, operating status data and electrical parameters, identify the abnormal conditions of the materials according to the material's location data, identify the abnormal environmental conditions according to the environmental parameters provided by the environmental sensors, identify abnormal events according to the monitoring video data, and send a first-level early warning notification to the maintenance personnel or issue a first-level alarm through the alarm device; The comprehensive risk identification module is used to analyze multi-source data through a pre-trained convolutional neural network, conduct comprehensive risk identification, and output identification results; The intelligent decision-making module is used to convert the recognition results into vector form and execute fusion intelligent decision-making, which includes sending a first-level warning notification to the maintenance personnel or issuing a first-level alarm through the alarm device, executing a second-level alarm, shutting down and / or reporting to the administrator's management terminal.

Citation Information

Patent Citations

  • Self-learning omnibearing mobile monitoring system

    CN103795984A

  • Smart construction site management system based on global variable frequency positioning and automatic recognition technologies

    CN107146168A

  • Safety production early warning prevention and control method

    CN117351643A

  • Building construction management method based on multi-data fusion

    CN117788223A

Cited By

  • Intelligent judicial security management and control system

    CN119296014A

  • Message transmitting and receiving method and system based on industrial internet and medium

    CN119383245A

  • Industrial production monitoring method and system based on large model, terminal and storage medium

    CN119644969A

  • Computer board card system

    CN119829351A

  • Self-adaptive multi-modal image recognition method and system based on artificial intelligence

    CN119848457A