Special equipment management method and system based on hybrid model

Abstract

The invention provides a special equipment management method and system based on a hybrid model, and belongs to the technical field of special equipment monitoring, and the method comprises the steps that a plurality of management units are arranged on an equipment kettle and a pipeline, each management unit comprises a communication module and a storage module, and the management units update parameters on the equipment kettle and the pipeline; parameters are stored through the storage module and uploaded to the cloud through networking of the communication module. The parameter updating of the management unit further comprises the following steps that S1, the management unit obtains the working state of the equipment kettle, connects the working state of the equipment kettle and the working state of the pipeline, and converts the corresponding working state into working parameters; s2, the management unit obtains working time sequence parameters of the equipment kettle and the pipeline through the communication module, and early warning is conducted on the working states of the equipment kettle and the pipeline through the working time sequence parameters; and S3, generating a time sequence view according to the early warning information.

Description

Technical Field

[0001] This specification relates to the technical field of special equipment management, and in particular to a special equipment management method and system based on a hybrid model. Background Art

[0002] With the rapid development of China's chemical industry, the number of chemical companies has increased dramatically, encompassing a wide range of sectors, from large state-owned enterprises to small and medium-sized private enterprises. However, due to differences in scale, capital investment, and technological capabilities, the level of special equipment management among chemical companies varies widely. Large enterprises typically have more comprehensive special equipment management systems and supporting resources, while small and medium-sized enterprises often face problems such as insufficient management resources and outdated technology. This disparity in management levels has led to uneven performance in the operating efficiency, safety, and maintenance costs of special equipment, which in turn has hampered the sustainable development of the entire industry.

[0003] An existing Chinese patent, publication number CN112180855A, is titled "A Factory Equipment Operation Management System," comprising a central control host, multiple controllers, and a mobile terminal. The multiple controllers collect operational status data for different unit equipment within the factory and transmit it to the central control host and mobile terminal. Upon receiving the operational status data, the central control host and mobile terminal can display the equipment's operational status and send control instructions to the controllers. In this application, workshop equipment is connected to the controllers, which are connected to the central control host and mobile terminal via wired or wireless networking, enabling data transmission and command sending, and enabling remote monitoring and operation of the equipment.

[0004] The above technology aims to monitor and intervene by associating controllers on each device, but the above monitoring system can only be applied to ordinary factories and equipment.

[0005] Therefore, it is necessary to provide a special equipment management method and system based on a hybrid model, which can realize monitoring applications on special equipment. Summary of the Invention

[0006] The embodiments of this specification provide a special equipment management method and system based on a hybrid model, which can set corresponding data reading terminals according to the properties of the equipment and pipelines in the workshop, obtain the working parameters of the equipment and pipelines, and upload them to the cloud. On the other hand, the cloud imports the company's information to obtain the standard working parameters of the equipment and pipelines, and then monitor the working status of the equipment and pipelines.

[0007] In some embodiments, a hybrid model-based special equipment management method includes multiple management units installed on the equipment kettle and pipeline, each management unit including a communication module and a storage module. The management unit updates parameters on the equipment kettle and pipeline, stores the parameters in the storage module, and uploads them to the cloud via the communication module. Managing unit update parameters also includes the following steps: S1: The management unit obtains the working status of the equipment kettle, the working status of the connecting equipment kettle and the pipeline, and converts the corresponding working status into working parameters; S2: The management unit obtains the working sequence parameters of the equipment kettle and pipeline through the communication module, and issues an early warning of the working status of the equipment kettle and pipeline based on the working sequence parameters; S3: Generate a time series view based on the warning information.

[0008] Furthermore, the management unit also includes a displacement detection module, which obtains the location information of the management unit. When the location information changes, the communication module is triggered to send the changed location information to the cloud through the communication module.

[0009] Furthermore, the working principle of the communication module also includes relaying and forwarding information on a workshop basis, including setting up a communication management relay in the workshop, the communication management relay obtaining the IP address of the management unit under the same local area network in the workshop, and obtaining the storage parameters sent by the corresponding communication module after binding the IP address. Within the set relay time, the storage parameters under multiple management units obtained are uploaded to the cloud.

[0010] Furthermore, the equipment kettle includes a reaction kettle, a distillation kettle, a low-temperature liquid storage tank, and a heater, and the pipeline includes a feed pipeline, a kettle connecting pipeline, a discharge pipeline, an exhaust pipeline, and an air intake pipeline.

[0011] Furthermore, the management unit includes a data reading end and a data acquisition end. The data reading end includes a temperature acquisition module, a vibration acquisition module, a power supply end and a Bluetooth sending module. The data acquisition end includes a Bluetooth receiving end, a storage module and a communication module. The data reading end is set on each equipment kettle or a single pipeline.

[0012] Furthermore, the storage module on the data acquisition end is also provided with an RFID communication module.

[0013] It also includes a special equipment management system based on a hybrid model, including a management unit set on each equipment kettle, a data reading terminal set on each pipeline system, and a cloud processor containing a hybrid model. The data reading end includes a temperature acquisition module, a vibration acquisition module, a power supply end and a Bluetooth sending module. The management unit also includes a Bluetooth receiving end; The management unit includes a data acquisition terminal, a storage module, a communication module, and an RFID communication module; The hybrid model in the cloud processor is based on a coupling model of image features and text information. It obtains the working timing parameters of each device and the personnel information of each department by calling the company's DCS system, CS system and ERP system. The equipment working parameters obtained by the data reading end are sent to the coupling model, and the coupling model outputs the early warning timing view of the workshop.

[0014] Furthermore, the working principle of the coupled model includes the following steps: S31: Extract features from the image through a convolutional neural network, represent text information through dictionary index encoding or glyph feature maps, and obtain a dataset; S32: Perform model training and optimization on the data set to obtain an early warning time series diagram; S33: Correct the training coefficient and the optimization coefficient according to the early warning timing diagram.

[0015] The beneficial effects of the present invention are: 1. Compared with the existing feature-based device management model based on handwritten registration, this application can comprehensively process enterprise personnel information and device operating parameters, obtain actual device operating parameters through a sensor network, and achieve accurate monitoring of the equipment; 2. Obtain the equipment's operating parameters, including the specific progress of the equipment during processing, through the sensor network. By monitoring the pipeline system, the equipment's operating status can be corrected. 3. By introducing the intelligent coupling model system, chemical companies can achieve intelligent and refined management of special equipment, improve equipment operation efficiency and safety, and reduce maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] This specification will be further described in the form of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting, and in these embodiments, like numbers represent like structures, wherein: Figure 1 is a schematic diagram of the working principle shown in some embodiments of this specification; Figure 2 It is a schematic diagram of data correction according to some embodiments of this specification; Figure 3 This is a nameplate installation diagram according to some embodiments of this specification.

[0017] Explanation of the accompanying reference numerals: 101, communication module; 102, storage module; 103, displacement detection module; 104, buffer layer; 105, mounting base; 106, through hole; 107, bolt; 108, nameplate body. DETAILED DESCRIPTION

[0018] To more clearly illustrate the technical solutions of the embodiments of this specification, the following briefly describes the drawings required for describing the embodiments. Obviously, the drawings described below are merely examples or embodiments of this specification. Those skilled in the art can apply this specification to other similar scenarios based on these drawings without inventive effort. Unless otherwise apparent from the context or otherwise noted, the same reference numerals in the figures represent the same structure or operation.

[0019] It should be understood that the terms "system," "device," "unit," and / or "module" used herein are a method for distinguishing different components, elements, parts, portions, or assemblies at different levels. However, if other terms can achieve the same purpose, the terms may be replaced by other expressions.

[0020] As used in this specification and claims, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" do not refer to the singular but also include the plural. Generally speaking, the terms "comprises" and "include" only indicate the inclusion of the steps and elements specifically identified, and these steps and elements do not constitute an exclusive list. A method or apparatus may also include other steps or elements.

[0021] Flowcharts are used throughout this specification to illustrate the operations performed by systems according to embodiments of this specification. It should be understood that preceding or following operations do not necessarily need to be performed in exact order. Instead, the steps may be processed in reverse order or simultaneously. Furthermore, other operations may be added to these processes, or one or more operations may be removed from these processes.

[0022] Example: In this embodiment, it is further extended that the special equipment intelligent identification and management system consists of three parts: intelligent identification nameplate, communication management relay and software environment.

[0023] The intelligent identification nameplate is one of the system's core components, primarily consisting of a storage module, a communication module, a displacement detection module, and a power supply system. These modules are highly integrated and encapsulated in an explosion-proof enclosure to ensure stable operation in the harsh environments of chemical plants. The intelligent identification nameplate is installed on the base plate of the special equipment support. This base plate is typically not replaced or removed during equipment installation and operation, and allows management and inspection personnel to quickly locate the component, ensuring the long-term stability of the nameplate.

[0024] The storage module stores all relevant information about the special equipment, including but not limited to the model, installation date, maintenance records, and inspection information. This information can be entered manually or stored as an image. Information in the storage module can be accessed remotely through the communication management relay; or by scanning or sensing using a smartphone equipped with the management system software. This method allows for rapid data access during special equipment management and inspection, providing a comprehensive understanding of all relevant information about the equipment and improving management efficiency.

[0025] When the physical position of the intelligent identification nameplate or special equipment changes, the displacement detection module will issue an alarm and send the alarm information to the communication management relay. The communication management relay then issues a warning to the backend, prompting management to address the issue promptly. This setting is intended to prevent insufficient coordination between the production and management departments, resulting in special equipment locations being changed without management's knowledge, or equipment duplication. This situation is quite common in chemical companies. Most of the reasons for this are that the production department arbitrarily adjusts the location of special equipment to meet deadlines or adjust processes, while others are due to disorganized management records and the use of duplicate equipment. In chemical companies, after the safety production management department has conducted a safety assessment, the location of special equipment is strictly prohibited from being arbitrarily changed. Even if the special equipment has reached the end of its service life, replacement must be carried out in accordance with the special equipment model parameters specified in the safety assessment. To prevent this behavior, a displacement detection module is added to the smart identification nameplate. If the position of the special equipment or the smart identification nameplate is changed, the module triggers the communication mechanism, which sends an alert to the communication management relay through the communication module. This module has two options. The first is the Beidou satellite navigation system, a global satellite positioning and communication system developed by China, which is the successor to the US Global Positioning System (GPS). The third mature satellite navigation system after Russia's GLONASS. This system can display the location of the equipment and can locate and track special equipment. The disadvantage is that it is expensive and has little practical significance for a large number of special equipment. The second method is based on the detection of displacement sensors. During the system initialization phase, the initial position of the displacement sensor is written into the system and is highly correlated with the position in the three-dimensional electronic map in the communication management relay. As long as the sensor senses an abnormality, it immediately triggers an alarm communication and sends an alarm signal to the communication management relay. The second method has more advantages in terms of economy and accuracy.

[0026] The intelligent identification nameplate has a built-in short-distance communication module, which is only responsible for data interaction with the communication management relay. In order to ensure its low power consumption, the module is designed with a trigger mechanism. When a special equipment or the intelligent identification nameplate on the special equipment is unauthorizedly displaced or is approaching inspection or the design service life, the intelligent identification nameplate automatically communicates with the communication management relay, and the communication relay immediately informs the background through remote communication to gain adjustment time for the enterprise manager. As far as the communication method is concerned, due to the presence of flammable and explosive environments or high temperature, high humidity and strong electromagnetic field environments on site, wired communication is undoubtedly the method with the least risk, but in the process safety and process of chemical enterprises, In the modernization of control, numerous industrial systems, such as PLCs, DCSs, and CSs, are deployed. Most of these systems rely on wired communication, necessitating the pre-installation of numerous cables that are spread throughout the factory floor, consuming significant resources. However, wireless communication solutions exist for environments with flammable and explosive conditions, or those with high temperatures, high humidity, and strong electromagnetic fields. Currently, new wireless communication architectures, such as radio frequency identification (RFID) technology, modular sensor networks combined with RFID modular network connectors, can meet the environmental requirements of chemical companies. However, these issues require comprehensive consideration based on the actual application scenarios and requirements. Therefore, based on the principle of high efficiency and low investment, wireless communication modules are the preferred choice.

[0027] The communication management relay consists of a storage module and a communication module, and is responsible for managing all smart identification nameplates in the workshop. A communication management relay is often installed in each workshop. Simply put, the communication management relay is a data management center based on the workshop. It centrally stores and backs up the information of each smart identification nameplate, exchanges data with each smart identification nameplate internally, and exchanges data with the background externally. The equipment must meet the workshop's explosion-proof and corrosion-proof requirements. Since its location is not too restricted, it can be configured more flexibly. Active power supply can be used for power supply. Internal communication is mainly wireless, and external communication can be configured with wired or wireless communication based on on-site resource conditions.

[0028] The communication management relay's storage module stores workshop layout information. Since most chemical plant equipment is three-dimensional, the workshop layout stored in the storage module is effectively a three-dimensional electronic map. The location of each intelligent identification nameplate is clearly marked on the map, allowing managers to quickly locate equipment. Furthermore, since the communication management relay has limited physical dimensions, it can be larger, and its data storage capacity is much greater than that of intelligent identification nameplates. In addition to a three-dimensional map with the location of each piece of equipment marked, it also stores basic and newly added information for each piece of special equipment. Intelligent identification nameplates only store basic equipment information. To access newly added information, access the communication management relay through the backend, or scan or sense the intelligent identification nameplate with a smartphone and connect to the communication management relay to download it.

[0029] The communication module has two communication modes. One is short-distance communication with each intelligent identification nameplate. As mentioned above, wireless communication is the main method. This communication module is mainly used for communication with the intelligent identification nameplate, and is used for functions such as displacement warning, near-expiry warning, and downloading new information to smartphones. The other is long-distance communication with the backend. Depending on the different on-site resource conditions, it can be set to wired and 5G communication. It is mainly responsible for data exchange with the backend, transmitting the warning and near-expiry warning signals of the intelligent identification nameplate to the backend. In addition, each new information, such as inspection information and special equipment failure risk information, is transmitted to the communication relay storage module through the backend. This dual communication mode ensures real-time data transmission and system stability.

[0030] The system's software environment integrates management, scheduling, and early warning. The software is connected to the company's internal ERP system and can precisely manage individual physical objects. Through this software, companies can achieve real-time monitoring, data analysis, and predictive maintenance of equipment, thereby improving the intelligence level of special equipment management and reducing special equipment failure rates and maintenance costs. At the same time, the software also includes docking with the data platforms of external regulatory departments, such as the management software of the production safety management department and the big data platform of the inspection department. After the company's special equipment undergoes supervisory inspections or regular inspections, the conclusions in the reports and the existing hidden dangers and risk factors can interact with the software and be stored as new information in the background or communication management relay. This data is then used to build a big data model. Through learning and training of the model, more accurate and efficient intelligent management models can be explored.

[0031] Use artificial intelligence technology to design and implement algorithms and models to solve practical problems or optimize system performance. This includes extracting features from data through methods such as machine learning and deep learning, building models that can adapt to complex environments, and establishing data algorithms and models. First, data collection and processing must be carried out. It is necessary to clarify the needs and scenarios for chemical special equipment, understand the problem goals, and determine the data types and sources. In terms of data types, for individual special equipment, one type is management data and the other is process data. Management data includes daily management such as equipment inspections and special equipment inspection data. This type of data mainly reflects changes in the appearance, materials, wall thickness, etc. of the special equipment itself. The source of this type of data is mainly through manual collection through the management personnel’s smartphone APP and the docking of the software system with the inspection agency’s big data platform for data updates. For this type of information Information is not limited to numerical parameters, but may also be graphic information, including some test or inspection reports. Another type of data is the impact of slight differences in the process conditions of special equipment on the equipment. Generally speaking, the operating parameters of special equipment in a workshop or section are the same, but there are also differences. For example, the pressure and temperature of the equipment cannot be absolutely consistent within a process system, and the differences in operator behavior will affect the changes in the process parameters of the equipment. This type of data is of great practical significance for managers to analyze production processes, equipment service life and process modifications. This type of data can be automatically collected and imported into the special equipment intelligent management system through data interfaces in the enterprise's internal DCS system, CS system and ERP system to realize the automatic collection of special equipment related data.

[0032] Selecting algorithms and models: To combine intelligent management systems for special equipment with artificial intelligence (AI) to achieve intelligent management systems, it is crucial to choose appropriate data analysis methods. There are many types of AI models, but based on the volume and data types of chemical companies, while chemical equipment is numerous, the data types are relatively few. Furthermore, the vast majority of collected data can be classified and processed. Therefore, a supervised learning model architecture is more suitable for intelligent management systems for chemical equipment. However, considering that these data types may include graphic information such as 3D maps and reports, it is possible to consider building a hybrid model that integrates image features with text information. A hybrid model may include a combination of an image feature extraction module and a text processing module. A convolutional neural network (CNN) is used to extract features from images, and then the text information is represented through dictionary index encoding or glyph feature maps. The data set is trained and optimized through the model architecture, and evaluation and verification are carried out on this basis. Ultimately, two results are achieved in actual application scenarios. First, through model training, predictive analysis can be performed on the operating trends of all special equipment of the enterprise, including dynamic equipment management, production processes, process flow analysis, and differential analysis of the impact of production process differences and human factors on special equipment. Second, through the intelligent analysis results of AI, a strong basis is provided for the company's managers.

[0033] To build an intelligent management system for special equipment, it is necessary to link multiple internal enterprise systems such as DCS, CS and ERP systems, as well as some external data platforms, such as the inspection department's big data platform, and integrate them with AI into a complete architecture. The AI model is trained through a large amount of data so that the model can understand and execute the key functions of the special equipment intelligent management software, and is tested and verified in actual operation to ensure that it meets the expected goals and promptly discovers and resolves potential problems. The configuration of the AI model and internal software is continuously adjusted according to the test results and business needs to improve the intelligence level and overall performance of the system. Through the above deployment, the special equipment intelligent management system can be intelligent with the help of AI, thereby improving the management efficiency and operational efficiency of special equipment, optimizing product production, reducing the risk of special equipment use, and laying a solid foundation for the sustainable development of the enterprise.

[0034] This application explores the design and implementation of an AI-based intelligent identification and management system for special equipment, starting from the current status of special equipment management in domestic chemical companies. By analyzing the system's functional requirements and implementation paths, feasible directions for future research are proposed. With the continuous development of AI technology, its application prospects in chemical companies, including but not limited to the management of special equipment, are broad. By introducing an intelligent identification and management system, chemical companies can achieve intelligent and refined management of special equipment, improve equipment operating efficiency and safety, and reduce maintenance costs. In the future, as the technology matures, AI-based intelligent equipment management technology will play an even more important role in the management of special equipment in chemical companies, driving the entire chemical industry's transformation towards intelligent and digital technology. Finally, this article also hopes that intelligent management of special equipment will not only be applied to the management of special equipment in chemical companies, but also have the opportunity to be applied to electromechanical special equipment such as elevators and lifting machinery. This is because electromechanical special equipment and chemical special equipment have similar management characteristics. Data collection for electromechanical special equipment is also a key content of the data management system. The nameplates of electromechanical special equipment should also be upgraded to intelligent technology. By empowering intelligent identification nameplates with more data, IoT technology integrates and transmits this data to a control center for subsequent data analysis and calculations. By integrating the IoT system into a data management platform, while enabling real-time monitoring of special equipment, data processing shifts from traditional manual processing to intelligent processing, thus avoiding the instability and high error rate of manual operations, while saving costs and improving work efficiency.

[0035] For chemical companies, data related to special equipment are mainly in the form of numerical values, and there may also be some information such as three-dimensional maps, positioning, and pictures and texts. These pictures and texts are mainly some location information, original data and other information. When initializing the smart nameplate, in order to reduce the workload or manual errors in the manual entry process, it is more convenient to use scanning and taking pictures. After building a platform with AI, AI needs to extract and identify the effective information in the pictures and texts, so we choose to use a hybrid model to identify and extract the effective information in the pictures.

[0036] Please refer to Figure 3 , Figure 3 The intelligent nameplate and its installation method are a preferred embodiment and structure of this embodiment, and do not limit the technical scope to be protected by this application. A lightweight steel plate has a buffer layer 104 composed of heat-insulating and shock-resistant materials inside, and two groups of through holes 106 (screw holes) are symmetrically distributed on the base. The group of screw holes closer to the center line is fixed to the support plate of the equipment by bolts, and the group of screw holes farther from the center line fixes the nameplate body 108 to the mounting base 105 by bolts.

[0037] The main body of the smart nameplate is composed of a package, which contains a storage module, a position detection module, a communication module and a power supply unit. The working principle of the position detection module is to obtain abnormal vibration to reflect whether the equipment has abnormal movement. The position detection module is provided with a mercury bead in the center of two ring-shaped wires. The ring-shaped wires are arranged parallel to the support pad (or support pad). When the equipment body is offset or the smart nameplate is offset due to any unauthorized process, the mercury bead moves to the edge. When the mercury bead rolls to the ring-shaped wire, the circuit under the wire is turned on, generating an electrical signal, and transmitting the mobile signal to the communication module through the communication unit, and further sending it to the cloud through the management relay.

[0038] There are multiple management units on the equipment kettle and pipeline. The management unit contains a communication module and a storage module. The management unit updates the parameters on the equipment kettle and pipeline, and stores the parameters through the storage module. Please refer to Figure 1 , upload to the cloud through the communication module; Managing unit update parameters also includes the following steps: S1: The management unit obtains the working status of the equipment kettle, the working status of the connecting equipment kettle and the pipeline, and converts the corresponding working status into working parameters; S2: The management unit obtains the working sequence parameters of the equipment kettle and pipeline through the communication module, and issues an early warning of the working status of the equipment kettle and pipeline based on the working sequence parameters; S3: Generate a time series view based on the warning information.

[0039] In another embodiment, the communication module also includes two communication systems, short-distance and long-distance. The short-distance communication and data exchange are performed with the intelligent identification nameplate, and the long-distance data exchange is performed with the background or cloud.

[0040] Please refer to Figure 2 , the working principle of the coupled model includes the following steps: S31: Extract features from the image through a convolutional neural network, represent text information through dictionary index encoding or glyph feature maps, and obtain a dataset; S32: Perform model training and optimization on the data set to obtain an early warning time series diagram; S33: Correct the training coefficient and the optimization coefficient according to the early warning timing diagram.

[0041] The model training process involves using optimizers (such as stochastic gradient descent and Adam) to update the model weights. Optimizer parameters (such as learning rate and momentum) affect training speed and convergence. The model coefficients refer to the model's weight parameters, which are updated at each iteration based on the gradient of the loss function. The optimizer parameters or model weights are dynamically adjusted based on the current model's prediction results (such as warning accuracy and equipment operating status false alarm rate). Machine learning is performed using AI, and data analysis is performed within a set period of time, with reports sent to management.

[0042] To sum up, the present application can realize wireless monitoring of special chemical equipment and pipeline systems, calculate the working parameters of the object through sensors installed on the monitored object, and obtain early warning information by combining the working time and working status of the equipment with online data. The two are compared and the early warning information is obtained. For equipment that does not meet the standard working status parameters, an early warning is issued, and the early warning information is sent to the employees who are about to conduct inspections.

Claims

1. A special equipment management method based on a hybrid model, characterized in that: Multiple management units are set on the equipment kettle and pipeline. The management units include communication modules and storage modules. The management units update the parameters on the equipment kettle and pipeline, store the parameters through the storage module, and upload them to the cloud through the communication module. Managing unit update parameters also includes the following steps: S1: The management unit obtains the working status of the equipment kettle, the working status of the connecting equipment kettle and the pipeline, and converts the corresponding working status into working parameters; S2: The management unit obtains the working sequence parameters of the equipment kettle and pipeline through the communication module, and issues an early warning of the working status of the equipment kettle and pipeline based on the working sequence parameters; S3: Generate a time series view based on the warning information.

2. A special equipment management method based on a hybrid model according to claim 1, characterized in that: The management unit also includes a displacement detection module, which obtains the location information of the management unit. When the location information changes, the communication module is triggered to send the changed location information to the cloud through the communication module.

3. A special equipment management method based on a hybrid model as claimed in claim 2, characterized in that: The working principle of the communication module also includes relaying and forwarding information on a workshop basis, including setting up a communication management relay in the workshop. The communication management relay obtains the IP address of the management unit under the same local area network in the workshop. After binding the IP address, the communication management relay obtains the storage parameters sent by the corresponding communication module, and uploads the obtained storage parameters under multiple management units to the cloud within the set relay time.

4. A special equipment management method based on a hybrid model as claimed in claim 3, characterized in that: The equipment kettle includes a reaction kettle, a distillation kettle, a low-temperature liquid storage tank, and a heater. The pipelines include a feed pipeline, a kettle connecting pipeline, a discharge pipeline, an exhaust pipeline, and an air intake pipeline.

5. A special equipment management method based on a hybrid model as claimed in claim 4, characterized in that: The management unit includes a data reading end and a data acquisition end. The data reading end includes a temperature acquisition module, a vibration acquisition module, a power supply end and a Bluetooth sending module. The data acquisition end includes a Bluetooth receiving end, a storage module and a communication module. The data reading end is set on each equipment kettle or a single pipeline.

6. A special equipment management method based on a hybrid model as described in claim 5, characterized in that: The storage module on the data acquisition end is also provided with an RFID communication module.

7. A hybrid model-based special equipment management system, comprising the hybrid model-based special equipment management method according to claim 6, characterized in that: It includes a management unit installed on each equipment kettle, a data reading terminal installed on each pipeline system, and a cloud processor containing a hybrid model. The data reading end includes a temperature acquisition module, a vibration acquisition module, a power supply end and a Bluetooth sending module. The management unit also includes a Bluetooth receiving end; The management unit includes a data acquisition terminal, a storage module, a communication module, and an RFID communication module; The hybrid model in the cloud processor is based on a coupling model of image features and text information. It obtains the working timing parameters of each device and the personnel information of each department by calling the company's DCS system, CS system and ERP system. The equipment working parameters obtained by the data reading end are sent to the coupling model, and the coupling model outputs the early warning timing view of the workshop.

8. A hybrid model-based special equipment management system as described in claim 7, characterized in that: The coupled model works by following these steps: S31: Extract features from the image through a convolutional neural network, represent text information through dictionary index encoding or glyph feature maps, and obtain a dataset; S32: Perform model training and optimization on the data set to obtain an early warning time series diagram; S33: Correct the training coefficient and the optimization coefficient according to the early warning timing diagram.

Images