Smart factory management method, system and device and readable storage medium

Abstract

The invention provides a smart factory management method, system and device and a readable storage medium, and relates to the technical field of smart factories, and the method comprises the steps: obtaining at least one artificial intelligence algorithm model trained by automatic training software in an AI module; the video of the industrial park is decoded in real time through the AI module, and a corresponding artificial intelligence algorithm model is scheduled to perform intelligent analysis and monitoring so as to realize abnormal alarm notification; performing real-time parameter monitoring on equipment in the industrial park through an SCADA module; and full-life-cycle management of the equipment is realized based on the basic information of the equipment through the EAM module. According to the method, the system, the device and the readable storage medium, the problem that an existing factory management method often focuses on post-event tracing and punishment and cannot meet the requirements of an enterprise for prevention and alarm in advance can be solved.

Description

Technical Field

[0001] The present invention relates to the field of smart factory technology, and in particular to a smart factory management method, system, device and readable storage medium. Background Art

[0002] In the modern industrial environment, especially in the context of intelligent manufacturing and smart factories, enterprises have increasingly stringent requirements for safe production and equipment operation. Traditional management models often focus on post-event tracing and punishment, mainly focusing on post-accident handling and accountability. Although this method can improve safety awareness and sense of responsibility to a certain extent, it cannot effectively deal with potential safety hazards and equipment failures, and cannot meet the needs of enterprises for pre-event prevention and warning. Summary of the invention

[0003] The technical problem to be solved by the present invention is to provide a smart factory management method, system, device and readable storage medium in response to the above-mentioned deficiencies in the prior art, so as to solve the problem that the existing factory management methods often focus on post-event tracing and punishment and cannot meet the needs of enterprises for pre-event prevention and warning.

[0004] In a first aspect, the present invention provides a smart factory management method, which is applied to a smart factory management system. The smart factory management system is constructed by integrating three software architectures: artificial intelligence AI, supervisory control and data acquisition SCADA, and enterprise asset management EAM. The method comprises:

[0005] Obtain at least one artificial intelligence algorithm model trained by the automated training software in the AI ​​module;

[0006] The AI ​​module is used to decode the video of the industrial park in real time, and the corresponding artificial intelligence algorithm model is dispatched for intelligent analysis and monitoring to realize abnormal alarm notification;

[0007] Real-time parameter monitoring of the equipment in the industrial park is performed through a SCADA module;

[0008] The EAM module enables full life cycle management of equipment based on its basic information.

[0009] Furthermore, before obtaining at least one artificial intelligence algorithm model trained by the automated training software in the AI ​​module, the method further includes:

[0010] The automated training software is built based on the artificial intelligence development process to provide a one-stop service for the entire process including data collection, data management, data annotation, model training, model verification, and model deployment;

[0011] At least one artificial intelligence algorithm model is trained by the automated training software.

[0012] Furthermore, the training of at least one artificial intelligence algorithm model by the automated training software specifically includes:

[0013] Receive a training data set, set labels, a verification data set, and a specified algorithm type imported by a user through the automated training software;

[0014] The training data set is labeled by the set labels, and the artificial intelligence algorithm model is trained based on the labeled training data set and the specified algorithm type;

[0015] The effectiveness of the artificial intelligence algorithm model is verified through the verification data set, and the verified artificial intelligence algorithm model is deployed.

[0016] Furthermore, the AI ​​module is used to decode the video of the industrial park in real time, and the corresponding artificial intelligence algorithm model is dispatched to perform intelligent analysis and monitoring to realize abnormal alarm notification, which specifically includes:

[0017] The AI ​​module reads the video stored in the video surveillance convergence software in the industrial park and performs real-time decoding;

[0018] According to the different scenario analysis requirements of users, the pre-configured artificial intelligence algorithm model is called to perform intelligent analysis and monitoring on the video decoding data from the corresponding real-time streaming protocol RTSP stream address;

[0019] Send the analysis results after the intelligent analysis monitoring to the designated information software according to user requirements and the set data interface and protocol specifications;

[0020] If the analysis result indicates an abnormality, an abnormal alarm notification is sent to supervisors at the corresponding level.

[0021] Furthermore, the artificial intelligence algorithm model includes a fireworks safety compliance monitoring model, and the intelligent analysis and monitoring steps of the fireworks safety compliance monitoring model include:

[0022] Detect the color image in the video decoding data based on the fireworks image features to determine whether there is a fire area;

[0023] In response to the presence of a fire area, estimating a relative area of ​​the fire area;

[0024] Determine whether the area is expanding by combining the previous and next frame results in the video decoding data;

[0025] A warning of a corresponding level is implemented based on the determination result of whether the area is expanding.

[0026] Furthermore, the artificial intelligence algorithm model includes a personnel and vehicle trajectory compliance monitoring model, and the intelligent analysis and monitoring steps of the personnel and vehicle trajectory compliance monitoring model include:

[0027] Based on the image segmentation technology, the pedestrian safety passage, double yellow line, intersection and grid line area in each image in the video decoding data are obtained;

[0028] Acquire the positions of motor vehicles, non-motor vehicles and on-site personnel in each image based on target detection technology;

[0029] Determine whether non-motor vehicles and on-site personnel are walking in the safe passage according to the positions of the pedestrian safe passage area and non-motor vehicles and on-site personnel, mark on-site personnel and non-motor vehicles that do not walk in the safe passage and report the violation information;

[0030] Obtain the motion trajectory of each motor vehicle in each image based on tracking technology;

[0031] Estimate the speed of motor vehicles based on their motion trajectories and the frame rate of the video stream, mark vehicles that exceed a preset speed, and report violation information;

[0032] Based on the double yellow lines, intersections and grid line areas, determine whether each motor vehicle crosses the double yellow lines, drives in the opposite direction or parks illegally. Vehicles that cross the double yellow lines, drive in the opposite direction or park illegally are marked and the violation information is reported.

[0033] Furthermore, the real-time parameter monitoring of the equipment in the industrial park by the SCADA module specifically includes:

[0034] The industrial field data of the industrial park is collected by the data collection system in the SCADA module, and transmitted to the monitoring center through the communication system, so that the monitoring center can monitor and manage the site according to the industrial field data.

[0035] Furthermore, the EAM module implements the full life cycle management of the equipment based on the basic information of the equipment, specifically including:

[0036] Through the EAM module, the equipment operation status in the production process is monitored based on the basic information of the equipment, the equipment utilization rate is calculated, the maintenance information affecting production is generated and transmitted, the equipment maintenance management is guided, and the equipment use compliance management is implemented.

[0037] In a second aspect, the present invention provides a smart factory management system, which is constructed by integrating three software architectures: artificial intelligence AI, supervisory control and data acquisition SCADA, and enterprise asset management EAM. The system includes an AI module, a SCADA module, and an EAM module that are interconnected.

[0038] The AI ​​module is used to obtain at least one artificial intelligence algorithm model trained by the automated training software;

[0039] The AI ​​module is also used to decode the video of the industrial park in real time and dispatch the corresponding artificial intelligence algorithm model for intelligent analysis and monitoring to realize abnormal alarm notification;

[0040] The SCADA module is used to perform real-time parameter monitoring on the equipment in the industrial park;

[0041] The EAM module is used to implement full life cycle management of equipment based on basic information of the equipment.

[0042] In a third aspect, the present invention provides a smart factory management device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to implement the smart factory management method described in the first aspect above.

[0043] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the smart factory management method described in the first aspect is implemented.

[0044] The smart factory management method, system, device and readable storage medium provided by the present invention are based on a smart factory management system constructed by integrating the three software architectures of AI, SCADA and EAM. First, at least one artificial intelligence algorithm model trained by the automated training software in the AI ​​module is obtained; then the video of the industrial park is decoded in real time through the AI ​​module, and the corresponding artificial intelligence algorithm model is dispatched for intelligent analysis and monitoring to realize abnormal alarm notification; then the equipment of the industrial park is monitored in real time through the SCADA module; and the full life cycle management of the equipment is realized based on the basic information of the equipment through the EAM module, so as to realize the integrated management of the equipment, ensure the high efficiency of the equipment, meet the needs of the enterprise for pre-prevention and warning, and avoid the occurrence of safety and equipment quality accidents to the greatest extent. It solves the problem that the existing factory management methods often focus on post-event tracing and punishment, and cannot meet the needs of the enterprise for pre-event prevention and warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 This is a flow chart of a smart factory management method according to Embodiment 1 of the present invention;

[0046] Figure 2 A schematic diagram of the software architecture of the smart factory management method according to an embodiment of the present invention;

[0047] Figure 3 This is a schematic diagram of the AI ​​software architecture of an embodiment of the present invention;

[0048] Figure 4 A schematic diagram of a human-machine interface of a data acquisition system according to an embodiment of the present invention;

[0049] Figure 5 This is a schematic diagram of the EAM software architecture of an embodiment of the present invention;

[0050] Figure 6 A flowchart of an artificial intelligence development process according to an embodiment of the present invention;

[0051] Figure 7 A schematic diagram of the service flow of the automated training software according to an embodiment of the present invention;

[0052] Figure 8 A schematic diagram of fireworks safety compliance monitoring according to an embodiment of the present invention;

[0053] Fig. 9 A schematic diagram of compliance monitoring of personnel and vehicle trajectories according to an embodiment of the present invention;

[0054] Fig.10 This is a schematic diagram of the structure of a smart factory management system according to Embodiment 2 of the present invention;

[0055] Fig.11 This is a structural diagram of a smart factory management device according to Example 3 of the present invention. DETAILED DESCRIPTION

[0056] In order to enable those skilled in the art to better understand the technical solution of the present invention, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0057] It can be understood that, in the absence of conflict, the various embodiments of the present invention and the various features in the embodiments can be combined with each other.

[0058] It can be understood that, for the convenience of description, the drawings of the present invention only show the parts related to the present invention, while the parts irrelevant to the present invention are not shown in the drawings.

[0059] It can be understood that each unit and module involved in the embodiments of the present invention may correspond to only one physical structure, or may be composed of multiple physical structures, or multiple units and modules may be integrated into one physical structure.

[0060] It can be understood that, without conflict, the functions and steps marked in the flowcharts and block diagrams of the present invention may occur in an order different from that marked in the drawings.

[0061] It is understood that the flowcharts and block diagrams of the present invention illustrate the possible architectures, functions, and operations of the systems, devices, equipment, and methods according to the various embodiments of the present invention. Each box in the flowchart or block diagram may represent a unit, module, program segment, or code, which contains executable instructions for implementing the specified functions. Moreover, each box or combination of boxes in the block diagram and flowchart may be implemented by a hardware-based system that implements the specified functions, or may be implemented by a combination of hardware and computer instructions.

[0062] It can be understood that the units and modules involved in the embodiments of the present invention can be implemented by software or hardware. For example, the units and modules can be located in a processor.

[0063] Embodiment 1:

[0064] This embodiment provides a smart factory management method, which is applied to a smart factory management system. The smart factory management system is constructed by integrating three software architectures: AI (Artificial Intelligence), SCADA (Supervisory Control and Data Acquisition), and EAM (Enterprise Asset Management). Figure 1 As shown, the method includes:

[0065] Step S101: Obtain at least one artificial intelligence algorithm model trained by the automated training software in the AI ​​module.

[0066] Step S102: The AI ​​module is used to decode the video of the industrial park in real time, and the corresponding artificial intelligence algorithm model is dispatched to perform intelligent analysis and monitoring to realize abnormal alarm notification.

[0067] In this embodiment, the smart factory management system provides an AI module for intelligent analysis and monitoring, realizes the deployment location of each video aggregation, and realizes functions such as video decoding, algorithm scheduling, statistical presentation, event alarm, and system docking.

[0068] Optionally, before obtaining at least one artificial intelligence algorithm model trained by the automated training software in the AI ​​module, the method further includes:

[0069] The automated training software is built based on the artificial intelligence development process to provide a one-stop service for the entire process including data collection, data management, data annotation, model training, model verification, and model deployment;

[0070] At least one artificial intelligence algorithm model is trained by the automated training software.

[0071] In this embodiment, the artificial intelligence development process needs to go through three stages: data preparation, model development, and model use. Automated training software is developed according to these three stages. Simple algorithms that use more data for training often have much better results than complex algorithms that are subject to the Dimension Curse. Excellent application results come not only from sophisticated and complex algorithms, but also from large-scale analysis and mining of massive training data. Automated training software provides a one-stop service for the entire process from data collection, data management, data labeling, model training, model verification, to model deployment, which can realize flexible iteration and upgrading of algorithm models.

[0072] Optionally, the training of at least one artificial intelligence algorithm model by the automated training software specifically includes:

[0073] Receive a training data set, set labels, a verification data set, and a specified algorithm type imported by a user through the automated training software;

[0074] The training data set is labeled by the set labels, and the artificial intelligence algorithm model is trained based on the labeled training data set and the specified algorithm type;

[0075] The effectiveness of the artificial intelligence algorithm model is verified through the verification data set, and the verified artificial intelligence algorithm model is deployed.

[0076] In this embodiment, the training steps of the automated training software include:

[0077] 1) Data collection: It can support uploading the training data to the automated training software through local SDK (Software Development Kit) and API (Application Programming Interface).

[0078] 2) Data management: supports data import and export, data viewing, and training data set management.

[0079] 3) Data labeling: You can set the labels supported by the data set and label the data according to the labels.

[0080] 4) Model training: You can specify the training data set and algorithm type, and monitor and manage the training tasks.

[0081] 5) Model verification: After the model training is completed, the effectiveness of the model can be verified by setting a verification data set.

[0082] 6) Model deployment: After verification, the model can be released, and the on-site program can obtain update notifications of the new version and perform local dynamic deployment and update of the model.

[0083] 7) Full-process iteration: Through multiple full-process iterations of model training, continuous improvement of the algorithm model can be achieved.

[0084] It should be noted that the labeling step of the automated training software can first use a pre-trained model or automatic labeling tool to preliminarily label the data set, and then after the preliminary automatic labeling, manually review and correct errors, and use these corrected data to fine-tune the pre-trained model. It can also adopt an active learning strategy to let the model select the most uncertain data points for manual review. In the automatic labeling process, quality control measures are implemented to gradually improve the labeling quality of the data set through continuous iterative optimization, combined with manual review and automatic labeling.

[0085] Optionally, the AI ​​module is used to decode the video of the industrial park in real time, and a corresponding artificial intelligence algorithm model is dispatched to perform intelligent analysis and monitoring to realize abnormal alarm notification, specifically including:

[0086] The AI ​​module reads the video stored in the video surveillance convergence software in the industrial park and performs real-time decoding;

[0087] According to the different scene analysis requirements of users, the video decoding data from the corresponding RTSP (Real Time Streaming Protocol) stream address is called by the pre-configured artificial intelligence algorithm model for intelligent analysis and monitoring;

[0088] Send the analysis results after the intelligent analysis monitoring to the designated information software according to user requirements and the set data interface and protocol specifications;

[0089] If the analysis result indicates an abnormality, an abnormal alarm notification is sent to supervisors at the corresponding level.

[0090] In this embodiment, the AI ​​module includes modules such as real-time video decoding, algorithm model scheduling, intelligent analysis monitoring, customer system docking, abnormal alarm notification, analysis result storage and comprehensive management background. Through the collaboration between the modules, the intelligent upgrade and transformation of the customer's existing video surveillance convergence software is realized. The functions of each module are as follows:

[0091] a) Real-time video decoding: Connect with the customer's monitoring video aggregation software, read the stored video recordings and decode them, and prepare for video analysis;

[0092] b) Algorithm model scheduling: According to the different scenario analysis requirements of customers, the corresponding algorithm model and control strategy are configured for the video data from the corresponding RTSP stream address;

[0093] c) Intelligent analysis and monitoring: intelligently analyze each channel of video data according to the analysis functions and corresponding cycles set to meet customer needs;

[0094] d) Customer system connection: Send the structured analysis results to the customer-specified information software according to the customer's needs and the set data interface and protocol specifications;

[0095] e) Abnormal alarm notification: Violation events or alarm results are notified to supervisors at the corresponding level in real time through SMS, WeChat public account, DingTalk, etc.;

[0096] f) Analysis result storage: save the analysis results for users to view, query and obtain evidence;

[0097] g) Comprehensive management background: realize functions such as user permissions, system settings, system monitoring, and result management.

[0098] Optionally, the artificial intelligence algorithm model includes a fireworks safety compliance monitoring model, and the intelligent analysis and monitoring steps of the fireworks safety compliance monitoring model include:

[0099] Detect the color image in the video decoding data based on the fireworks image features to determine whether there is a fire area;

[0100] In response to the presence of a fire area, estimating a relative area of ​​the fire area;

[0101] Determine whether the area is expanding by combining the previous and next frame results in the video decoding data;

[0102] A corresponding level of warning is implemented based on the determination result of whether the area is expanding.

[0103] In this embodiment, the fire image features such as the color, shape, and movement trajectory of the smoke and the color and flickering frequency of the flame are used to determine whether there is a fire area and the severity of the fire, and to issue a warning of the corresponding level accordingly. Once the fire is found to have a tendency to continue to expand, the warning level is gradually increased.

[0104] The warning levels include:

[0105] a) Low-level warning: It may only remind you of potential risks, such as detecting cigarette butts or small-scale smoke in the monitoring area, but it does not yet pose a fire threat.

[0106] b) Medium-level warning: When obvious smoke or small flames are detected, the system will issue a medium-level warning, indicating that further observation and assessment of the fire situation is required.

[0107] c) High-level warning: When the fire and smoke detection algorithm detects obvious flames or a lot of smoke, a high-level warning is triggered, which usually means that immediate action is required, such as activating the fire extinguishing system or notifying the fire department.

[0108] d) Emergency level warning: This is the highest level of warning, usually triggered when the system detects a large-scale fire or a fire that spreads rapidly, requiring immediate evacuation of personnel and fire extinguishing.

[0109] The specific warning levels and response measures may vary according to different application scenarios and safety requirements. For example, in some systems, the warning levels may be combined with specific color coding to facilitate quick identification and response.

[0110] Optionally, the artificial intelligence algorithm model includes a personnel and vehicle trajectory compliance monitoring model, and the intelligent analysis and monitoring steps of the personnel and vehicle trajectory compliance monitoring model include:

[0111] Based on the image segmentation technology, the pedestrian safety passage, double yellow line, intersection and grid line area in each image in the video decoding data are obtained;

[0112] Acquire the positions of motor vehicles, non-motor vehicles and on-site personnel in each image based on target detection technology;

[0113] Determine whether non-motor vehicles and on-site personnel are walking in the safe passage according to the positions of the pedestrian safe passage area and non-motor vehicles and on-site personnel, mark on-site personnel and non-motor vehicles that do not walk in the safe passage and report the violation information;

[0114] Obtain the motion trajectory of each motor vehicle in each image based on tracking technology;

[0115] Estimate the speed of motor vehicles based on their motion trajectories and the frame rate of the video stream, mark vehicles that exceed a preset speed, and report violation information;

[0116] Based on the double yellow lines, intersections and grid line areas, determine whether each motor vehicle crosses the double yellow lines, drives in the opposite direction or parks illegally. Vehicles that cross the double yellow lines, drive in the opposite direction or park illegally are marked and the violation information is reported.

[0117] In this embodiment, the personnel and vehicle trajectory compliance monitoring model is used to perform real-time analysis of on-site personnel and vehicles in the monitoring screen to intelligently identify at least the following non-compliant behaviors: 1) on-site personnel and non-motor vehicles walking on non-sidewalks; 2) motor vehicles driving in the wrong direction; 3) motor vehicles speeding; 4) motor vehicles illegally parking.

[0118] Step S103: Real-time parameter monitoring of the equipment in the industrial park is performed through a SCADA module.

[0119] Specifically, the industrial field data of the industrial park is collected by the data collection system in the SCADA module, and transmitted to the monitoring center through the communication system, so that the monitoring center monitors and manages the site according to the industrial field data.

[0120] In this embodiment, the SCADA module consists of three parts: a monitoring center, a communication system, and a data acquisition system. The data acquisition system includes a PLC (Programmable Logic Controller), a terminal, field instruments, a field human-machine interface, etc., wherein the field instruments include a temperature transmitter, a pressure transmitter, a flow meter, etc. In the case of no one on duty, the data acquisition system can automatically collect industrial field data with a long communication distance, and transmit various flow parameters to the terminal, and finally transmit the parameters to the monitoring center, which analyzes and processes the data. The monitoring center uses the communication system to obtain information such as the temperature, pressure, flow, and equipment operating status of the field equipment to monitor and manage the site.

[0121] Step S104: Implement full life cycle management of the equipment based on the basic information of the equipment through the EAM module.

[0122] Specifically, through the EAM module, the equipment operation status in the production process is monitored based on the basic information of the equipment, the equipment utilization rate is calculated, the maintenance information affecting production is generated and transmitted, the equipment maintenance management is guided, and the equipment use compliance management is realized.

[0123] In this embodiment, the EAM module monitors the operating status of the equipment in the production process based on the basic information of the equipment (such as the asset model, equipment ledger, operating status, maintenance history, etc.), calculates the equipment utilization OEE (Overall Equipment Effectiveness), produces and transmits maintenance information related to production, and guides equipment maintenance management; implements equipment use compliance management to ensure that the equipment is operated correctly in the correct state. Ensure the compliance of the equipment during use and cleaning. Accurately manage the technical properties of equipment, scales, rooms, and containers, and guide calibration and cleaning behaviors to be carried out in a timely and effective manner.

[0124] In the present invention, the management of park safety in smart factories uses visual AI as an efficient tool, and production management is mainly the management of equipment, among which SCADA focuses on real-time parameter monitoring of equipment and EAM focuses on the full life cycle management of equipment; AI visual analysis requires extremely fast data processing capabilities and low-latency transmission, SCADA / EAM requires millisecond-level transmission speeds and large connection points, and 5G capabilities and edge computing capabilities provide factories with extremely low-latency transmission and intelligent analysis guarantees. Therefore, the advantages of 5G+MEC (Mobile Edge Computing) in low-latency and wide connections can be fully utilized to achieve large concurrent connections of factory equipment, real-time visual intelligent analysis, and real-time transmission of analysis results, helping to create a benchmark for smart factories.

[0125] In a specific embodiment, the smart factory management method may include:

[0126] 1. Overall technical thinking

[0127] Research on 5G+MEC private network architecture: build 5G base stations and MEC edge cloud, cover the production workshop with 5G network, and realize the interconnection of manufacturing equipment through 5G network. AI+SCADA+EAM analysis software: 1) Provide AI module to realize the deployment location of each video collection, realize video decoding, algorithm scheduling, statistical presentation, event alarm, system docking and other functions; provide EAM module to realize the equipment ledger, inspection, maintenance and repair functions; provide SCADA module to realize real-time data collection and abnormal signal analysis alarm, post-analysis and other functions of equipment; 2) Provide standard algorithm service to realize the function of monitoring whether personnel have crossed the boundary, whether they are wearing safety helmets in compliance with regulations and the time when personnel are on duty and off duty, and provide equipment integration service to realize the full life cycle management function of equipment.

[0128] It should be noted that the present invention integrates and develops a set of smart factory software based on the three architectures of AI, SCADA and EAM to achieve integrated management of equipment.

[0129] 2. Software Architecture

[0130] like Figure 2 As shown in the figure, through the development of 5G+MEC+AI+SCADA+EAM smart factory software, the management of park security and equipment management is supplemented to build an integrated digital smart factory. Among them, the architecture analysis of AI, SCADA, and EAM is as follows:

[0131] 2.1 AI Software Architecture

[0132] The project construction can use Unicom AI industrial compliance monitoring software products with independent property rights and industry-leading performance, such as Figure 3As shown in the figure, it mainly includes modules such as real-time video decoding, algorithm model scheduling, intelligent analysis monitoring, customer system docking, abnormal alarm notification, analysis result storage and comprehensive management background. Through the collaboration between the modules, the intelligent upgrade and transformation of the customer's existing video surveillance convergence software can be achieved. The functions of each module are as follows:

[0133] a) Real-time video decoding: Connect with the customer's monitoring video aggregation software, read the stored video recordings and decode them, and prepare for video analysis;

[0134] b) Algorithm model scheduling: According to the different scenario analysis requirements of customers, the corresponding algorithm model and control strategy are configured for the video data from the corresponding RTSP stream address;

[0135] c) Intelligent analysis and monitoring: intelligently analyze each channel of video data according to the analysis functions and corresponding cycles set to meet customer needs;

[0136] d) Customer system connection: Send the structured analysis results to the customer-specified information software according to the customer's needs and the set data interface and protocol specifications;

[0137] e) Abnormal alarm notification: Violation events or alarm results are notified to supervisors at the corresponding level in real time through SMS, WeChat public account, DingTalk, etc.;

[0138] f) Analysis result storage: save the analysis results for users to view, query and obtain evidence;

[0139] g) Comprehensive management background: realize functions such as user permissions, system settings, system monitoring, and result management.

[0140] 2.2 SCADA software architecture

[0141] SCADA software mainly consists of three parts: monitoring center, communication system and data acquisition system. The upper computer and the lower computer are connected by the communication system. The upper computer is mainly responsible for monitoring, and the lower computer is mainly responsible for control.

[0142] Monitoring center: With the help of the communication system, the temperature, pressure, flow rate, equipment operation status and other information of the on-site equipment are obtained to monitor and manage the station. The monitoring system is mainly composed of the control department of the control center, the local station control department and the remote control station monitoring department.

[0143] The monitoring center mainly uses a dual Ethernet structure and uses TCP (Transmission Control Protocol) / IP (Internet Protocol) protocol to provide security for the local area network. The control department of the dispatch center includes external equipment such as engineer stations, operator stations, servers, front-end machine subsystems, large-screen projectors, and UPS subsystems. In order to ensure the security and reliability of data services, the production line digital construction product data acquisition management platform system is installed on the latest Windows server version operating system, and uses two backup data servers with exactly the same functions and configurations. When one data server fails, the other server will automatically switch to the main server operation mode. Local station control is mainly achieved by connecting a computer equipped with local control software to the terminal and sending relevant instructions to the terminal. The remote monitoring station can monitor the actual operation of the on-site equipment and cannot issue commands directly.

[0144] Communication system: responsible for long-distance transmission of information. Communication methods can be divided into three types: wired transmission, wireless transmission and hybrid transmission.

[0145] Data acquisition system: mainly composed of PLC, terminal, field instrument, field human-machine interface, etc. Field instrument includes temperature transmitter, pressure transmitter, flow meter, etc., which can automatically collect industrial field data with long communication distance without human supervision, and transmit various flow parameters to the terminal, and finally transmit the parameters to the monitoring center, which will analyze and process the data. The human-machine interface diagram of the data acquisition system is shown in the figure below. Figure 4 As shown, it is used to display the parameter settings, operating data, alarm information, etc. of each device.

[0146] 2.3 EAM Software Architecture

[0147] EAM software architecture diagram Figure 5 As shown, EAM monitors the operating status of equipment in the production process based on basic information of equipment (such as equipment asset model, equipment ledger, operating status, maintenance history, etc.), calculates equipment utilization OEE, produces and transmits maintenance information related to production, and guides equipment maintenance management; implements equipment use compliance management to ensure that the equipment is operated correctly in the correct state. Ensure the compliance of equipment during use and cleaning. Accurately manage the technical properties of equipment, scales, rooms, and containers, and guide calibration and cleaning behaviors to be carried out in a timely and effective manner.

[0148] EAM is aimed at on-site equipment management in factories / workshops to ensure compliance with equipment usage, such as: the equipment is operated correctly in good condition and real data is recorded; ensuring that the equipment meets cleaning requirements and operating parameters are within the compliance range, without any adverse impact on product quality. Accurately manage the technical attributes of equipment, scales, clean areas, and containers, so as to guide calibration and cleaning activities to be carried out in a timely and effective manner.

[0149] 3. Functional performance requirements

[0150] AI part:

[0151] 3.1 AI algorithm training software

[0152] like Figure 6 As shown in Figure 1, the AI ​​development process needs to go through three stages: data preparation, model development, and model use. Automated training software is developed according to these three stages. Simple algorithms trained with more data often have much better results than complex algorithms subject to the Dimension Curse. Excellent application results come not only from sophisticated and complex algorithms, but also from large-scale analysis and mining of massive training data.

[0153] like Figure 7 As shown in the figure, the automated training software provides one-stop service for the entire process from data collection, data management, data labeling, model training, model verification, to model deployment, which can realize flexible iteration and upgrading of algorithm models.

[0154] 1) Data collection: It can support uploading the training data to the automated training software through local SDK (Software Development Kit) and API (Application Programming Interface).

[0155] 2) Data management: supports data import and export, data viewing, and training data set management.

[0156] 3) Data labeling: You can set the labels supported by the data set and label the data according to the labels.

[0157] 4) Model training: You can specify the training data set and algorithm type, and monitor and manage the training tasks.

[0158] 5) Model verification: After the model training is completed, the effectiveness of the model can be verified by setting a verification data set.

[0159] 6) Model deployment: After verification, the model can be released, and the on-site program can obtain update notifications of the new version and perform local dynamic deployment and update of the model.

[0160] 7) Full-process iteration: Through multiple full-process iterations of model training, continuous improvement of the algorithm model can be achieved.

[0161] It should be noted that the labeling steps of the automated training software may include:

[0162] a) Use a pre-trained model: Use a pre-trained model that has been trained on similar data as a starting point to perform preliminary annotation on the dataset. For example, use the YOLO series model for automatic annotation of object detection.

[0163] b) Automatic labeling tool: Use the automatic labeling tool Label-Studio, which supports image and video processing and uses custom models for automatic labeling.

[0164] c) Fine-tuning the pre-trained model: After the initial automatic labeling, manual review and error correction are performed, and the corrected data is used to fine-tune the pre-trained model to improve the accuracy of the labeling.

[0165] d) Active learning: Using active learning strategies, the model selects the most uncertain data points for manual review, thus efficiently utilizing manual review to improve model performance.

[0166] e) Quality control: During the automatic labeling process, quality control measures are implemented, such as consistency checks between labelers, to ensure the accuracy and consistency of data labeling.

[0167] f) Iterative optimization: Through continuous iterative optimization, combined with manual review and automatic labeling, the labeling quality of the dataset is gradually improved.

[0168] g) Optimize the effect of intelligent labeling: If you are not satisfied with the accuracy of intelligent labeling, you can optimize the effect of intelligent labeling through manual verification and correction, and then use this part of the data to retrain the model to further improve the labeling accuracy.

[0169] 3.2. Fireworks safety compliance monitoring

[0170] Fire definition: Visible fire appears in the picture.

[0171] This function analyzes the real-time video stream obtained by the surveillance camera. Once a visible open fire area is detected in the picture, the abnormal fire event is reported and the administrator is notified through WeChat for Enterprise. For example, the diagram of fireworks safety compliance monitoring is as follows: Figure 8 As shown; once the fire is found to have a tendency to continue to expand, the warning level will be gradually raised.

[0172] The implementation process of fireworks safety compliance monitoring is as follows:

[0173] 1) Access surveillance video stream

[0174] 2) Algorithm layer performs video stream decoding

[0175] 3) Get color image

[0176] 4) Fire area detection based on deep learning

[0177] 5) Estimate the relative size of the fire area

[0178] 6) Combine the previous and next frame results to determine whether the area is expanding

[0179] 7) Once a fire area is discovered, report the dangerous incident immediately and notify the administrator through WeChat for Enterprise.

[0180] 8) If the fire is found to have a tendency to continue to expand, the warning level will be gradually increased.

[0181] It should be noted that the AI ​​fire and smoke detection algorithm will judge the severity of the fire based on the image features captured in the video surveillance, such as the color, shape, movement trajectory of the smoke and the color and flickering frequency of the flame, and issue an early warning of the corresponding level accordingly.

[0182] The warning levels include:

[0183] a) Low-level warning: It may only remind you of potential risks, such as detecting cigarette butts or small-scale smoke in the monitoring area, but it does not yet pose a fire threat.

[0184] b) Medium-level warning: When obvious smoke or small flames are detected, the system will issue a medium-level warning, indicating that further observation and assessment of the fire situation is required.

[0185] c) High-level warning: When the fire and smoke detection algorithm detects obvious flames or a lot of smoke, a high-level warning is triggered, which usually means that immediate action is required, such as activating the fire extinguishing system or notifying the fire department.

[0186] d) Emergency level warning: This is the highest level of warning, usually triggered when the system detects a large-scale fire or a fire that spreads rapidly, requiring immediate evacuation of personnel and fire extinguishing.

[0187] The specific warning levels and response measures may vary according to different application scenarios and safety requirements. For example, in some systems, the warning levels may be combined with specific color coding to facilitate quick identification and response.

[0188] Implementation effect:

[0189] The algorithm supports custom areas for identification (taking into account the optical fiber, temperature and other conditions of the area, and deploying a sufficient number of surveillance cameras to ensure full coverage and no blind spots). When the fire area is not less than 60x60 pixels, the recognition accuracy rate exceeds 80%.

[0190] Risk points:

[0191] 1) In long-distance areas and when the open flame is not obvious, the recognition rate is relatively low.

[0192] 2) Color is also one of the main characteristics of open flames. The open flames photographed in night infrared mode may not be obvious compared to the surrounding environment, and the recognition rate will be greatly affected. It can be replaced with a black light level full-color camera, or a thermal camera can be added for open flame monitoring.

[0193] 3) During the development and deployment process, the implementation of the above algorithm will be appropriately adjusted according to the actual situation on site and the computing power resource conditions.

[0194] 3.3、Trajectory compliance monitoring of personnel and vehicles

[0195] This function conducts real-time analysis of on-site personnel and vehicles in the monitoring screen, and intelligently identifies the following non-compliant behaviors: 1) On-site personnel walking on non-sidewalks; 2) Motor vehicles driving in the wrong direction; 3) Motor vehicles speeding.

[0196] like Fig. 9 As shown, the technical implementation steps of this part are:

[0197] 1) Obtain pedestrian safety passages, double yellow lines, intersections and grid line areas in the picture based on image segmentation technology;

[0198] 2) Obtain the positions of motor vehicles, non-motor vehicles and on-site personnel in the picture based on target detection technology;

[0199] 3) Determine whether non-motor vehicles and on-site personnel are walking in the safe passage according to the pedestrian safe passage area and the positions of non-motor vehicles and on-site personnel, mark on-site personnel and non-motor vehicles that do not walk in the safe passage and report violation information;

[0200] 4) Obtain the motion trajectory of each motor vehicle in the picture based on tracking technology;

[0201] 5) Estimate the speed of motor vehicles based on their motion trajectory and video stream frame rate (the camera will be calibrated after installation to ensure the accuracy of speed estimation), mark vehicles that exceed 30km / h on ordinary roads and 5km / h at intersections, and report violation information; and identify motor vehicles crossing double yellow lines, driving in the wrong direction, speeding, and illegal parking.

[0202] SCADA+EAM part:

[0203] 3.4、SCADA function introduction

[0204] a) Basic functions

[0205] Data collection: equipment operating status, status time, real-time operating data, fault alarm data, and technical parameters of some equipment;

[0206] It should be noted that SCADA focuses on real-time parameter monitoring of equipment, while EAM focuses on the full life cycle management of equipment.

[0207] Edge computing: utilization rate, output, shift output, running time, stop time, failure time, etc.;

[0208] Long-distance transmission: 4G / ADSL (Asymmetric Digital Subscriber Line)

[0209] Remote monitoring: data presentation, real-time curve, historical curve, historical data export, alarm reminder, alarm record, data item statistics, report output;

[0210] b) Production management

[0211] Production status: running, manual, semi-automatic, fully automatic, adjustment, shutdown, failure, etc., duration of various status, etc.;

[0212] Production statistics: shift production, yesterday's production, today's production, monthly production, total production statistics, report export, etc.;

[0213] Efficiency management: production progress, estimated completion time, time utilization rate, performance utilization rate, yield rate, statistical analysis, etc.

[0214] c) Equipment Management

[0215] Equipment records: managed by area and floor

[0216] Alarm management: system failure, equipment failure alarm, WeChat push / large screen display / andon system, etc.

[0217] d) Data management

[0218] Report management: report, storage, query, export and other management

[0219] Historical data storage, data statistical analysis

[0220] Data interaction: open interface, production data and equipment data interact with MES (Manufacturing Execution System) / ERP (Enterprise Resource Planning) / WMS (Warehouse Management System) and other systems;

[0221] 3.5. Introduction to EAM functions

[0222]

[0223]

[0224] Taking the EAM system as an example, if you want to add a new function, such as AI-based predictive maintenance, the steps may include:

[0225] 1) Data collection: Use IoT (Internet of Things) devices to collect asset performance data.

[0226] 2) Data integration: Integrate the collected data with the existing asset information in the EAM system.

[0227] 3) Model development: Develop machine learning models to analyze data and predict potential asset failures.

[0228] 4) System Integration: Integrate the AI ​​model into the EAM system so that it can automatically trigger maintenance work orders.

[0229] 5) User training: Train the maintenance team on how to use the new predictive maintenance capabilities.

[0230] 6) Test and deploy: Test new features in a live environment before fully deploying them.

[0231] 7) Monitoring and optimization: Continuously monitor the effectiveness of AI predictive maintenance and optimize based on feedback.

[0232] It should be noted that the smart factory management method provided by the present invention has the following beneficial effects:

[0233] a) In line with the needs of enterprises for smart factories, we will create safety compliance and equipment management software based on AI+SCADA+EAM technology from the two aspects of park safety and production control. We will use MEC to improve edge computing capabilities, use AI technology to replace manual control of park safety production operations to achieve the goal of "0" accidents, and use SCADA+EAM technology to achieve full life cycle management of equipment to ensure the high efficiency of equipment;

[0234] b) Develop AI+SCADA+EAM software. AI includes modules such as real-time video decoding, algorithm model scheduling, intelligent analysis and monitoring, customer system docking, abnormal alarm notification, analysis result storage and comprehensive management background. SCADA+EAM includes modules such as equipment data collection, data connection, data analysis, data abnormality alarm, equipment ledger, equipment inspection, equipment maintenance, equipment repair, equipment data report, etc.

[0235] By adding one set of software and one industry product, costs are expected to be reduced by 30%, efficiency increased by 20%, staff reduced by 25%, and the intelligence rate increased by 50%.

[0236] c) Complete the construction of enterprise data collection and monitoring and control application information system according to the established plan.

[0237] d) Improve the accuracy of data collection and equipment operation data collection to more than 99%.

[0238] e) Through AI intelligent analysis, unsafe behaviors of workers can be reduced or even avoided, effectively solving problems that affect production due to personnel safety.

[0239] The smart factory management method provided by the embodiment of the present invention is based on a smart factory management system constructed by integrating the three software architectures of AI, SCADA and EAM. First, at least one artificial intelligence algorithm model trained by the automated training software in the AI ​​module is obtained; then the video of the industrial park is decoded in real time through the AI ​​module, and the corresponding artificial intelligence algorithm model is dispatched for intelligent analysis and monitoring to realize abnormal alarm notification; the equipment of the industrial park is monitored in real time through the SCADA module; and the full life cycle management of the equipment is realized based on the basic information of the equipment through the EAM module, so as to realize the integrated management of the equipment, ensure the high efficiency of the equipment, meet the needs of the enterprise for pre-prevention and warning, and minimize the occurrence of safety and equipment quality accidents. It solves the problem that the existing factory management methods often focus on post-event tracing and punishment, and cannot meet the needs of the enterprise for pre-event prevention and warning.

[0240] Embodiment 2:

[0241] like Fig.10 As shown, this embodiment provides a smart factory management system, which is constructed by integrating three software architectures: artificial intelligence AI, supervisory control and data acquisition SCADA, and enterprise asset management EAM. The system includes an AI module 11, a SCADA module 12, and an EAM module 13 that are interconnected;

[0242] The AI ​​module 11 is used to obtain at least one artificial intelligence algorithm model trained by the automated training software;

[0243] The AI ​​module 11 is also used to decode the video of the industrial park in real time and dispatch the corresponding artificial intelligence algorithm model to perform intelligent analysis and monitoring to realize abnormal alarm notification;

[0244] The SCADA module 12 is used to monitor the parameters of the equipment in the industrial park in real time;

[0245] The EAM module 13 is used to implement full life cycle management of the equipment based on the basic information of the equipment.

[0246] Optionally, the AI ​​module 11 is also used to build the automated training software including a one-stop service for the entire process from data collection, data management, data annotation, model training, model verification to model deployment based on the artificial intelligence development process;

[0247] The AI ​​module 11 is also used to train at least one artificial intelligence algorithm model through the automated training software.

[0248] Optionally, the training of at least one artificial intelligence algorithm model by the automated training software specifically includes:

[0249] Receive a training data set, set labels, a verification data set, and a specified algorithm type imported by a user through the automated training software;

[0250] The training data set is labeled by the set labels, and the artificial intelligence algorithm model is trained based on the labeled training data set and the specified algorithm type;

[0251] The effectiveness of the artificial intelligence algorithm model is verified through the verification data set, and the verified artificial intelligence algorithm model is deployed.

[0252] Optionally, the real-time decoding of the video of the industrial park and the scheduling of the corresponding artificial intelligence algorithm model for intelligent analysis and monitoring to achieve abnormal alarm notification specifically include:

[0253] The AI ​​module reads the video stored in the video surveillance convergence software in the industrial park and performs real-time decoding;

[0254] According to the different scenario analysis requirements of users, the pre-configured artificial intelligence algorithm model is called to perform intelligent analysis and monitoring on the video decoding data from the corresponding real-time streaming protocol RTSP stream address;

[0255] Send the analysis results after the intelligent analysis monitoring to the designated information software according to user requirements and the set data interface and protocol specifications;

[0256] If the analysis result indicates an abnormality, an abnormal alarm notification is sent to supervisors at the corresponding level.

[0257] Optionally, the artificial intelligence algorithm model includes a fireworks safety compliance monitoring model, and the intelligent analysis and monitoring steps of the fireworks safety compliance monitoring model include:

[0258] Detect the color image in the video decoding data based on the fireworks image features to determine whether there is a fire area;

[0259] In response to the presence of a fire area, estimating a relative area of ​​the fire area;

[0260] Determine whether the area is expanding by combining the previous and next frame results in the video decoding data;

[0261] A warning of a corresponding level is implemented based on the determination result of whether the area is expanding.

[0262] Optionally, the artificial intelligence algorithm model includes a personnel and vehicle trajectory compliance monitoring model, and the intelligent analysis and monitoring steps of the personnel and vehicle trajectory compliance monitoring model include:

[0263] Based on the image segmentation technology, the pedestrian safety passage, double yellow line, intersection and grid line area in each image in the video decoding data are obtained;

[0264] Acquire the positions of motor vehicles, non-motor vehicles and on-site personnel in each image based on target detection technology;

[0265] Determine whether non-motor vehicles and on-site personnel are walking in the safe passage according to the positions of the pedestrian safe passage area and non-motor vehicles and on-site personnel, mark on-site personnel and non-motor vehicles that do not walk in the safe passage and report the violation information;

[0266] Obtain the motion trajectory of each motor vehicle in each image based on tracking technology;

[0267] Estimate the speed of motor vehicles based on their motion trajectories and the frame rate of the video stream, mark vehicles that exceed a preset speed, and report violation information;

[0268] Based on the double yellow lines, intersections and grid line areas, determine whether each motor vehicle crosses the double yellow lines, drives in the opposite direction or parks illegally. Vehicles that cross the double yellow lines, drive in the opposite direction or park illegally are marked and the violation information is reported.

[0269] Optionally, the SCADA module 12 is specifically used for:

[0270] The industrial field data of the industrial park is collected by the data collection system in the SCADA module, and transmitted to the monitoring center through the communication system, so that the monitoring center can monitor and manage the site according to the industrial field data.

[0271] Optionally, the EAM module 13 is specifically used for:

[0272] Based on the basic information of the equipment, the equipment operation status in the production process is monitored, the equipment utilization rate is calculated, maintenance information affecting production is generated and transmitted, equipment maintenance management is guided, and equipment use compliance management is implemented.

[0273] Embodiment 3:

[0274] refer to Fig.11 This embodiment provides a smart factory management device, including a memory 21 and a processor 22. The memory 21 stores a computer program, and the processor 22 is configured to run the computer program to execute the smart factory management method in Example 1.

[0275] The memory 21 is connected to the processor 22. The memory 21 may be a flash memory, a read-only memory or other memory. The processor 22 may be a central processing unit or a single-chip microcomputer.

[0276] Embodiment 4:

[0277] This embodiment provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the smart factory management method in the above-mentioned embodiment 1 is implemented.

[0278] The computer-readable storage medium includes volatile or non-volatile, removable or non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, computer program modules or other data). Computer-readable storage media include, but are not limited to, RAM (Random Access Memory), ROM (Read-Only Memory), EEPROM (Electrically Erasable Programmable read only memory), flash memory or other memory technology, CD-ROM (Compact Disc Read-Only Memory), digital versatile disk (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer.

[0279] In summary, the smart factory management method, system, device and readable storage medium provided by the embodiments of the present invention are based on a smart factory management system constructed by integrating the three software architectures of AI, SCADA and EAM. First, at least one artificial intelligence algorithm model trained by the automated training software in the AI ​​module is obtained; then, the video of the industrial park is decoded in real time through the AI ​​module, and the corresponding artificial intelligence algorithm model is dispatched for intelligent analysis and monitoring to realize abnormal alarm notification; then, the equipment in the industrial park is monitored in real time through the SCADA module; and the full life cycle management of the equipment is realized based on the basic information of the equipment through the EAM module, so as to realize the integrated management of the equipment, ensure the high efficiency of the equipment, meet the needs of the enterprise for pre-prevention and warning, and minimize the occurrence of safety and equipment quality accidents. It solves the problem that the existing factory management methods often focus on post-event tracing and punishment, and cannot meet the needs of the enterprise for pre-event prevention and warning.

[0280] It is to be understood that the above embodiments are merely exemplary embodiments used to illustrate the principles of the present invention, but the present invention is not limited thereto. For those of ordinary skill in the art, various modifications and improvements can be made without departing from the spirit and essence of the present invention, and these modifications and improvements are also considered to be within the scope of protection of the present invention.

Claims

1. A smart factory management method, characterized in that: Applied to a smart factory management system, the smart factory management system is constructed by integrating three software architectures: artificial intelligence AI, supervisory control and data acquisition SCADA, and enterprise asset management EAM. The method includes: Obtain at least one artificial intelligence algorithm model trained by the automated training software in the AI ​​module; The AI ​​module is used to decode the video of the industrial park in real time, and the corresponding artificial intelligence algorithm model is dispatched for intelligent analysis and monitoring to realize abnormal alarm notification; Real-time parameter monitoring of the equipment in the industrial park is performed through a SCADA module; The EAM module enables full life cycle management of equipment based on its basic information.

2. The method according to claim 1, characterized in that Before obtaining at least one artificial intelligence algorithm model trained by the automated training software in the AI ​​module, the method further includes: The automated training software is built based on the artificial intelligence development process to provide a one-stop service for the entire process including data collection, data management, data annotation, model training, model verification, and model deployment; At least one artificial intelligence algorithm model is trained by the automated training software.

3. The method according to claim 2, characterized in that The step of training at least one artificial intelligence algorithm model by the automated training software specifically includes: Receive a training data set, set labels, a verification data set, and a specified algorithm type imported by a user through the automated training software; The training data set is labeled by the set labels, and the artificial intelligence algorithm model is trained based on the labeled training data set and the specified algorithm type; The effectiveness of the artificial intelligence algorithm model is verified through the verification data set, and the verified artificial intelligence algorithm model is deployed.

4. The method according to claim 1, characterized in that: The AI ​​module is used to decode the video of the industrial park in real time, and the corresponding artificial intelligence algorithm model is dispatched to perform intelligent analysis and monitoring to realize abnormal alarm notification, specifically including: The AI ​​module reads the video stored in the video surveillance convergence software in the industrial park and performs real-time decoding; According to the different scenario analysis requirements of users, the pre-configured artificial intelligence algorithm model is called to perform intelligent analysis and monitoring on the video decoding data from the corresponding real-time streaming protocol RTSP stream address; Send the analysis results after the intelligent analysis monitoring to the designated information software according to user requirements and the set data interface and protocol specifications; If the analysis result indicates an abnormality, an abnormal alarm notification is sent to supervisors at the corresponding level.

5. The method according to claim 4, characterized in that The artificial intelligence algorithm model includes a fireworks safety compliance monitoring model, and the intelligent analysis and monitoring steps of the fireworks safety compliance monitoring model include: Detect the color image in the video decoding data based on the fireworks image features to determine whether there is a fire area; In response to the presence of a fire area, estimating a relative area of ​​the fire area; Determine whether the area is expanding by combining the previous and next frame results in the video decoding data; A warning of a corresponding level is implemented based on the determination result of whether the area is expanding.

6. The method according to claim 4, characterized in that The artificial intelligence algorithm model includes a personnel and vehicle trajectory compliance monitoring model, and the intelligent analysis and monitoring steps of the personnel and vehicle trajectory compliance monitoring model include: Based on the image segmentation technology, the pedestrian safety passage, double yellow line, intersection and grid line area in each image in the video decoding data are obtained; Acquire the positions of motor vehicles, non-motor vehicles and on-site personnel in each image based on target detection technology; Determine whether non-motor vehicles and on-site personnel are walking in the safe passage according to the positions of the pedestrian safe passage area and non-motor vehicles and on-site personnel, mark on-site personnel and non-motor vehicles that do not walk in the safe passage and report the violation information; Obtain the motion trajectory of each motor vehicle in each image based on tracking technology; Estimate the speed of motor vehicles based on their motion trajectories and the frame rate of the video stream, mark vehicles that exceed a preset speed and report violation information; Based on the double yellow lines, intersections and grid line areas, determine whether each motor vehicle crosses the double yellow lines, drives in the opposite direction or parks illegally. Vehicles that cross the double yellow lines, drive in the opposite direction or park illegally are marked and the violation information is reported.

7. The method according to claim 1, characterized in that The real-time parameter monitoring of the equipment in the industrial park by the SCADA module specifically includes: The industrial field data of the industrial park is collected by the data collection system in the SCADA module, and transmitted to the monitoring center through the communication system, so that the monitoring center can monitor and manage the site according to the industrial field data.

8. The method according to claim 1, characterized in that The EAM module implements the full life cycle management of the equipment based on the basic information of the equipment, specifically including: Through the EAM module, the equipment operation status in the production process is monitored based on the basic information of the equipment, the equipment utilization rate is calculated, the maintenance information affecting production is generated and transmitted, the equipment maintenance management is guided, and the equipment use compliance management is implemented.

9. A smart factory management system, characterized in that: The smart factory management system is constructed by integrating three software architectures: artificial intelligence AI, supervisory control and data acquisition SCADA, and enterprise asset management EAM. The system includes an AI module, a SCADA module, and an EAM module that are interconnected; The AI ​​module is used to obtain at least one artificial intelligence algorithm model trained by the automated training software; The AI ​​module is also used to decode the video of the industrial park in real time and dispatch the corresponding artificial intelligence algorithm model for intelligent analysis and monitoring to realize abnormal alarm notification; The SCADA module is used to monitor the parameters of the equipment in the industrial park in real time; The EAM module is used to implement full life cycle management of equipment based on basic information of the equipment.

10. A smart factory management device, characterized in that: It comprises a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to implement the smart factory management method as described in any one of claims 1-8.

11. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the smart factory management method according to any one of claims 1 to 8 is implemented.

Images