A driving multi-source data wireless transparent transmission system and method based on a 5G-CPE

Abstract

The application discloses a kind of based on 5G-CPE's driving multi-source data wireless transmission system, namely method, it is characterized in that, including: end-network-edge-cloud four-level hierarchical architecture, in turn is terminal acquisition layer, convergence access layer, edge processing layer, cloud intelligent analysis layer;The terminal acquisition layer includes driving main control PLC, safety monitoring PLC, video acquisition equipment, multidimensional sensing terminal, for output driving control data, safety monitoring data, video stream, equipment working condition sensing original data;The application adopts physical isolation industry 5G special network+three-level directional access strategy dual security protection, isolates public network invasion risk, guarantees driving production control data communication security, end-network-edge-cloud integrated fusion AI diagnosis architecture, driving multidimensional data is jointly analyzed, realizes equipment fault early prediction, changes from after-the-fact maintenance to predictive intelligent operation and maintenance.

Description

Technical Field

[0001] This invention relates to the fields of industrial internet, 5G private network communication, and intelligent operation and maintenance technology of cranes, and particularly to a wireless transparent transmission system and method for multi-source data of cranes based on 5G-CPE. Background Technology

[0002] In recent years, a new round of technological revolution and industrial transformation has been developing in depth. The construction of new infrastructure such as 5G, industrial internet, big data centers, and artificial intelligence has been accelerated, and they are increasingly becoming key drivers supporting the digital, networked, and intelligent transformation and upgrading of the real economy.

[0003] The steel industry mainly includes ironmaking, steelmaking, steel casting, steel rolling, warehousing and logistics. The steel industry has a long production process and complex technology, and currently faces major challenges such as low equipment maintenance efficiency, lack of transparency in the production process, fragmented downstream demand, and significant pressure to adopt green production methods. There is great potential for developing intelligent manufacturing and lean management models.

[0004] Existing overhead cranes are equipped with control systems and safety monitoring systems, but lack remote monitoring capabilities. They are typically managed through periodic inspections and replacements, and reactive maintenance. The most challenging aspect of controlling an overhead crane is its real-time operational status. As the crane operates independently above the factory building and is constantly moving during operation, any malfunction requires personnel to climb onto the crane for on-site inspection. This is time-consuming and labor-intensive for maintenance personnel. Furthermore, the constant movement and changing operating conditions of the crane make it difficult to maintain accurate information about equipment malfunctions and preserve the initial scene of the incident.

[0005] Existing transmission channels struggle to simultaneously meet the demands of low latency control commands, high video throughput, and high-frequency sensor acquisition. They fail to differentiate data processing based on crane load, operating speed, and travel orientation, resulting in a high false alarm rate for fixed warning thresholds. Furthermore, the lack of isolation between industrial control, monitoring, and office networks poses risks of crane control data leakage and unauthorized interference. The dispersed storage of equipment data, coupled with the absence of a unified fusion and analysis platform, prevents the early prediction of crane malfunctions using vibration, soundprint, and temperature AI models, limiting alarms to reactive measures. Summary of the Invention

[0006] The purpose of this invention is to address the shortcomings of traditional heavy-duty overhead cranes in steel plants, such as unstable transmission, data silos between multiple devices, false alarms due to fixed threshold warnings, weak network security, and lack of integrated AI predictive maintenance. This invention provides a wireless transparent transmission system and method for multi-source data from overhead cranes based on 5G-CPE. Utilizing a physically isolated 5G private network within the plant area, and using the vehicle-mounted 5G-CPE as a unified aggregation gateway, it achieves unified acquisition and encapsulation of multi-protocol data from PLC control, safety monitoring, multi-channel video, and gearbox temperature, vibration, and sound sensors. It automatically binds and transmits layered data based on crane load, speed, and gear condition tags. Coupled with a cloud-based LSTM+DBSCAN clustering AI early warning model, it enables remote visualization of the overhead crane, predictive maintenance of the gearbox, and quantitative calculation of remaining equipment lifespan. This provides a wireless transparent transmission system and method for multi-source data from overhead cranes based on 5G-CPE.

[0007] The technical solution of the present invention is: a wireless transparent transmission system for multi-source data in vehicles based on 5G-CPE, characterized in that it includes: a four-level layered architecture of terminal-network-edge-cloud, namely, terminal acquisition layer, aggregation access layer, edge processing layer, and cloud intelligent analysis layer. The terminal acquisition layer includes a crane main control PLC, a safety monitoring PLC, video acquisition equipment, and a multi-dimensional sensing terminal, which are used to output crane control data, safety monitoring data, video streams, and raw data from equipment operating condition sensors. The aggregation access layer is a vehicle-mounted 5G-CPE gateway, and the 5G-CPE has a built-in multi-protocol adaptation and acquisition module, a working condition tag generation module, a unified data encapsulation module, and a secure access control module. The 5G-CPE serves as the sole data aggregation node for the vehicle's mobile terminal and communicates wirelessly with the industrial 5G private network, which is physically isolated from the public network. The edge processing layer includes a ground edge server and an engineer inspection workstation, used to receive and unpack data packets uploaded by 5G-CPE, cache real-time driving data, and send control commands to the vehicle. The cloud-based intelligent analysis layer is a multi-source data fusion AI analysis platform with a built-in time-series database and AI fault diagnosis model. It is used to receive tagged driving data synchronized from the edge layer and complete multi-source data fusion storage, operating condition matching analysis and equipment fault early warning. The 5G-CPE reads the multi-protocol raw data output by heterogeneous devices at the terminal acquisition layer through the multi-protocol adaptation acquisition module. After standardization and conversion, the data is bound to the real-time operating condition tag of the vehicle by the operating condition tag generation module, and then encapsulated into a unified data packet by the unified data encapsulation module. This data is then transmitted hierarchically to the edge processing layer through the industrial 5G private network and finally synchronized to the cloud intelligent analysis platform to complete AI intelligent diagnosis. The security access control module is configured with a three-level targeted access policy to restrict illegal cross-level data interaction.

[0008] Preferably, the multi-protocol adaptation acquisition module is compatible with MODBUS TCP protocol, OPC UA protocol, and sensor device private communication protocol, and can parse four types of heterogeneous data in parallel: vehicle control, safety monitoring, video, and vibration / temperature / acoustic sensing, and output structured data frames in a unified format.

[0009] Preferably, the working condition label generation module collects working condition parameters such as vehicle load, travel speed, trolley / carriage travel direction, and lifting height in real time, encodes the working condition parameters to generate unique working condition labels, and embeds them into the header of each structured data frame to realize the binding of data with the real-time operating status of the vehicle.

[0010] Preferably, the unified data encapsulation module classifies the transmission priority of data frames with operating condition labels into different levels: driving safety control signals have the highest priority, equipment sensor operating condition data have the medium priority, and high-definition video streams have the lowest priority; the encapsulated data packets are marked with service priority identifiers, and the 5G private network dynamically allocates transmission bandwidth and latency resources according to the identifiers.

[0011] Preferably, the security access control module is configured with a three-level directional access policy, including: Level 1, only allowing the vehicle-mounted 5G-CPE to unidirectionally upload data to the edge processing layer; Level 2, only allowing the edge processing layer to issue legitimate inspection, speed adjustment, and shutdown control commands to the vehicle terminal; Level 3, the cloud-based intelligent analysis platform can only read historically stored data synchronized by the edge layer and prohibits reverse writing to the vehicle control system; all cross-level data interactions automatically retain access logs.

[0012] Preferably, the cloud-based intelligent analysis platform has a built-in AI fault diagnosis model including a vibration spectrum analysis model, an equipment acoustic signature recognition model, and a temperature trend prediction model. The AI ​​fault diagnosis model matches the corresponding fault alarm threshold under the working condition according to the working condition label carried by the data frame, and identifies equipment faults such as motor overheating, bearing wear, and track abnormal noise based on multi-source fusion data, and automatically generates operation and maintenance early warning work orders.

[0013] Preferably, the industrial 5G private network is equipped with a factory area 5G base station and a vehicle-mounted 5G-CPE that supports seamless and rapid 5G cell switching during large-scale vehicle movement, maintaining a stable wireless communication link throughout the process and avoiding data transmission interruptions and lags during vehicle movement.

[0014] This invention also provides a method for wireless transparent transmission of multi-source data in vehicles based on 5G-CPE, which, when applied to the above system, includes the following steps: Step S1: The terminal acquisition layer collects raw data from the vehicle's main control system, safety monitoring, video, vibration / temperature / acoustic sensor in parallel, and sends it to the vehicle-mounted 5G-CPE; Step S2: The multi-protocol adaptation and acquisition module inside the 5G-CPE uniformly parses and transforms the raw data of multiple protocols, and outputs standardized structured data frames; Step S3: Synchronously collect real-time operating parameters of vehicle load, speed, and orientation; the operating condition label generation module generates operating condition labels and embeds them into structured data frames. Step S4: The unified data encapsulation module encapsulates tagged data frames into data packets according to service priority, and transmits them wirelessly to the ground edge processing layer through the physically isolated industrial 5G private network in a hierarchical manner. Step S5: The edge server unpacks and splits the data packets, caches real-time driving data locally, and issues vehicle control commands to the 5G-CPE according to the on-site operation requirements; Step S6: The edge server periodically uploads all multi-source data with operating condition tags to the cloud-based intelligent analysis platform. Step S7: The cloud platform stores the entire lifecycle of vehicle driving data. The AI ​​fault diagnosis model combines the corresponding operating condition thresholds to analyze multi-source data, identify potential equipment faults, and push early warning information. Step S8: The security access control module executes the three-level directed access policy throughout the process, records all data interaction logs, and achieves secure and traceable transmission.

[0015] Preferably, in step S4, during industrial 5G private network transmission, a low-latency dedicated channel is allocated to the highest priority vehicle safety control signal, and a large-bandwidth time-division transmission channel is allocated to the low-priority high-definition video stream, so as to simultaneously complete the parallel backhaul of multiple types of data.

[0016] Preferably, in step S7, the cloud-based AI fault diagnosis model distinguishes different operating scenarios such as no-load, light-load, heavy-load, high-speed travel, and low-speed inspection based on the operating condition labels carried by the data. Each scenario is independently configured with fault alarm thresholds, and the analysis standards are dynamically matched to reduce the probability of false alarms and missed alarms.

[0017] The present invention has the following beneficial effects: (1) Innovatively propose a dedicated 5G-CPE vehicle-mounted aggregation network architecture for mobile driving scenarios, which solves the problems of easy damage to wired communication and unstable WiFi switching under large-scale vehicle movement. A single gateway is compatible with all heterogeneous devices in the vehicle, eliminating data silos.

[0018] (2) Design a multi-protocol unified encapsulation transparent transmission mechanism to be compatible with mainstream industrial control protocols and proprietary sensor protocols in parallel, so as to realize low-latency wireless backhaul of multiple types of data such as control, video and sensing.

[0019] (3) Create an adaptive tagging data processing mechanism for driving conditions, and dynamically adjust the fault warning threshold according to real-time load and speed to greatly reduce the false alarm and missed alarm problems caused by fixed thresholds.

[0020] (4) Adopting a dual security protection of physical isolation industrial 5G private network + three-level targeted access strategy to isolate the risk of public network intrusion and ensure the security of train production control data communication. The integrated AI diagnostic architecture of terminal-network-edge-cloud will jointly analyze multi-dimensional train data to realize early prediction of equipment failure and transform post-maintenance into predictive intelligent operation and maintenance. Attached Figure Description

[0021] Figure 1 This is the network topology diagram of the present invention. Figure 2 This is a control diagram of the driving system of the present invention. Figure 3 This is a schematic diagram of the 5G private network of the present invention. Figure 4 This is a schematic diagram of crane data tagging according to the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] The terminology used in the embodiments of this invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The singular forms “a,” “the,” and “the” used in the embodiments of this invention and the appended claims are also intended to include the plural forms, unless the context clearly indicates otherwise. “Multiple” generally includes at least two, but does not exclude the inclusion of at least one.

[0024] It should be understood that the term "and / or" used in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.

[0025] Depending on the context, the words “if” or “suppose” as used here can be interpreted as “when” or “in response to determination” or “in response to detection.” Similarly, depending on the context, the phrases “if determination” or “if detection (of the stated condition or event)” can be interpreted as “when determination” or “in response to determination” or “when detection (of the stated condition or event)” or “in response to detection (of the stated condition or event).”

[0026] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a product or system comprising a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a product or system. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the product or system that includes said element.

[0027] The present invention discloses a 5G-CPE-based vehicle multi-source wireless transparent transmission system, characterized in that it includes a four-layer architecture of terminal-network-edge-cloud, wherein the four layers are, in order, terminal acquisition layer, aggregation access layer, edge processing layer, and cloud intelligent analysis layer.

[0028] In this embodiment, the terminal acquisition layer is deployed on a heavy-duty overhead crane in a steel plant and includes a crane control unit, a safety monitoring unit, and a third-party sensing unit. The crane control unit includes an ABB main control PLC and a CMS control system, which outputs lifting weight, trolley / gantry speed, running gear, and direction OPC signals. The safety monitoring unit includes a Siemens safety PLC, multiple industrial cameras, and infrared temperature measurement equipment, which output safety interlock signals and multiple video streams. The third-party sensing unit includes a reducer temperature and vibration integrated sensor and a contact-type sound collector, which output triaxial vibration, lubricating oil temperature, and raw equipment acoustic data via MODBUS TCP; the aggregation access layer is an on-board 5G-CPE gateway, which has a built-in multi-protocol parsing module, a working condition tag generation module, a hierarchical data encapsulation module, a private network security control module, and a local cache unit. All terminal acquisition layer devices are uniformly connected to the 5G-CPE via industrial switches. An independent industrial 5G private network is deployed in the factory area. The industrial 5G private network is physically isolated from the enterprise office public network. The 5G-CPE can wirelessly communicate with the 5G base station in the factory area throughout the entire vehicle's travel.

[0029] The edge processing layer is located in the ground inspection office and includes a 5G terminal, an iba PDA high-speed acquisition system, and a local visual operation and maintenance front end. It is used to receive data packets uploaded by 5G-CPE, complete protocol unpacking, real-time display of vehicle status, playback of multiple video streams, issue inspection and speed control commands to the vehicle, and cache driving time sequence data for a short period of time to generate daily work cycle and cumulative load statistics reports.

[0030] The cloud-based intelligent analysis layer is an integrated intelligent operation and maintenance platform for vehicles. It has a built-in working condition classification time series database, a reducer AI early warning sub-model, a voiceprint clustering sub-model, and an equipment lifespan calculation module. It receives full data with working condition tags synchronized from the edge processing layer, and completes multi-source data fusion storage, intelligent early warning of equipment failure, and quantitative calculation of the remaining working cycle life of the vehicle.

[0031] The 5G-CPE synchronously parses heterogeneous raw data from OPC UA, MODBUS TCP, and video RTSP through a multi-protocol parsing module, converting it into standardized time-series data frames. The operating condition label generation module reads the PLC's weight, speed, and gear signals in real time, generates corresponding operating condition codes, and embeds them into the header of each data frame. The hierarchical data encapsulation module encapsulates data packets according to service priority and wirelessly transmits them to the edge processing layer via a physically isolated 5G private network. In weak signal scenarios, data is cached by a local caching unit, and transmission resumes after signal recovery. The private network security control module configures a three-layer directional access strategy to manage cross-level data interaction, and the cloud-based intelligent analysis layer uses data with operating condition labels to drive AI models to complete predictive operation and maintenance analysis.

[0032] In this embodiment, the working condition label generation module automatically divides 12 types of standard operating conditions, including heavy-load full-speed ascent, heavy-load full-speed descent, light-load full-speed ascent, light-load full-speed descent, unloaded full-speed ascent, unloaded full-speed descent, trolley moving right at full speed, trolley moving left at full speed, main trolley moving forward at full speed, main trolley moving backward at full speed, auxiliary trolley moving forward at full speed, and auxiliary trolley moving backward at full speed. Based on the lifting weight, it distinguishes between unloaded, light-load, and heavy-load ranges, and generates a unique working condition code label by combining the mechanism speed, operating gear, and operating direction.

[0033] In this embodiment, the hierarchical data encapsulation module divides the transmission priority into three levels: the first level is driving safety control signals, which are allocated a dedicated 5G low-latency transmission channel; the second level is high-frequency sensor data such as speed reducer vibration, temperature, and acoustic signature; and the third level is multiple high-definition video streams, which are allocated a high-bandwidth time-division transmission channel, and multiple types of data are synchronously and parallelly wirelessly transmitted back.

[0034] In this embodiment, the private network security control module is configured with a three-layer directional access strategy: the first layer only allows the vehicle-mounted 5G-CPE to upload data unidirectionally to the edge processing layer; the second layer only allows the edge processing layer to issue legal control commands for inspection, speed adjustment, and shutdown to the vehicle terminal; the third layer, the cloud-based intelligent analysis platform, only reads historical stored data synchronized at the edge and prohibits reverse writing to the vehicle PLC control system, and all cross-layer data interactions are automatically logged.

[0035] In this embodiment, the cloud-based intelligent analysis layer has a built-in voiceprint clustering sub-model. The voiceprint clustering sub-model extracts sound features using MFCC Mel-Cepstral Coefficients and constructs a standard voiceprint library specifically for the reducer using the DBSCAN clustering algorithm. The residual between the real-time collected voiceprints and the feature vectors in the library is calculated to identify abnormal noises from the reducer bearings and output an early warning.

[0036] In this embodiment, the cloud-based intelligent analysis layer has a built-in AI early warning sub-model for the reducer. The early warning sub-model is an LSTM long short-term memory network. It takes multi-dimensional features such as vibration, lubricating oil temperature, and motor current under the corresponding working conditions as input, configures independent fault alarm thresholds according to different working conditions, dynamically calculates the equipment deterioration trend, and predicts reducer failures in advance.

[0037] In this embodiment, the cloud-based intelligent analysis layer has a built-in device lifespan calculation module. The module counts the number of daily operating cycles and the single load weight to generate a load spectrum for the entire machine. Combined with the total allowable number of operating cycles designed for the crane, and deducting the historical cumulative cycle volume, the remaining service life of the crane is automatically calculated.

[0038] In this embodiment, the third-party sensing unit is arranged at the input and output shafts of the planetary reducer and the secondary reducer to synchronously collect three-axis vibration data, lubricating oil inlet and outlet temperatures, and equipment operating noise. All sensing data are transmitted to the vehicle switch via the MODBUS TCP protocol and aggregated to the 5G-CPE.

[0039] The overall technical solution of this invention is as follows: I. For example Figure 1 The system network topology diagram mainly consists of a remote operation and maintenance system, a vehicle control PLC, a safety monitoring system, a video system, and third-party intelligent devices. This embodiment presents a four-layer overall system architecture: terminal-network-edge-cloud. 1) Terminal data acquisition layer (vehicle-mounted terminal): This includes three main categories of data acquisition devices: ①Traffic control unit: ABB main hoist PLC and CMS control system, outputting hoisting weight, trolley / cab speed, running gear, and direction OPC signals; ② Safety monitoring unit: Siemens safety PLC, multiple high-definition industrial cameras, infrared temperature measurement equipment, outputting safety interlock signals and multiple video streams; ③ Third-party sensing unit: planetary / secondary reducer temperature and vibration integrated sensor, contact sound collector, which outputs raw data of triaxial vibration, lubricating oil inlet and outlet temperature, and equipment operation sound pattern through MODBUS TCP.

[0040] 2) Aggregation Access Layer (Vehicle-mounted 5G-CPE Core Gateway) The 5G-CPE is the only wireless exit for the vehicle and has five built-in functional modules: multi-protocol parsing module, working condition tag generation module, hierarchical data encapsulation module, private network security control module, and local cache unit. All vehicle terminals are uniformly connected to the 5G-CPE Ethernet port through a switch. The entire journey relies on the factory's 5G base station for seamless switching to maintain a continuous wireless link.

[0041] 3) Edge processing layer (ground inspection engineer station): Deploy iba PDA high-speed acquisition system and local visual operation and maintenance front end, receive 5G-CPE transparent data packets, complete protocol unpacking, real-time vehicle status display, real-time video playback, support issuing inspection and speed control commands to the vehicle, cache real-time time sequence data of the vehicle in a short period of time, and generate daily work cycle and cumulative load statistical reports.

[0042] 4) Cloud-based intelligent analysis layer (integrated smart operation and maintenance platform): Equipped with a working condition classification database, a reducer AI early warning model, a soundprint clustering model, and an equipment lifespan calculation module; receives full data with working condition tags synchronized from the edge station, realizing load spectrum statistics, multi-dimensional trend analysis of vibration / temperature / soundprint, early warning of reducer faults, calculation of remaining working cycle life of the vehicle, and supports remote WEB viewing, alarm information push, and historical data traceability and export.

[0043] II. 5G-CPE has built-in core functional modules, such as Figure 3 This is a dedicated network for this embodiment. 1) The multi-protocol parsing module is compatible with OPC UA (PLC control signals), MODBUS TCP (temperature, vibration and sound sensors), and video RTSP protocols. It can parse the raw data of the three types of heterogeneous devices in parallel and synchronously, and unify and standardize them into time-series structured data frames to eliminate device protocol barriers.

[0044] 2) The working condition label generation module reads the lifting weight, mechanism speed, trolley / cart gear position, and running direction signals from the PLC in real time, and automatically distinguishes 12 standard working conditions, including heavy load / light load / no load, full speed rise / fall, trolley left and right, and trolley forward and backward. The working condition code label is embedded in the header of each structured data frame to achieve one-to-one binding of data with the operating condition.

[0045] 3) The hierarchical data encapsulation module divides the data with working condition tags into three levels of transmission priority: Level 1 (highest) is safety control signals, which are allocated a low-latency 5G dedicated channel; Level 2 is high-frequency sensor data of temperature, vibration and sound of reducer; Level 3 is multi-channel high-definition video streaming, with high bandwidth time-division transmission, taking into account both control real-time performance and video integrity.

[0046] 4) Private Network Security Control Module: The factory's 5G private network is physically completely isolated from the enterprise's public office network, configured with a three-layer directional access policy: only allowing the vehicle-mounted CPE to unidirectionally upload data to the edge station; the edge station only issues legitimate inspection commands to the crane; the cloud only reads historical stored data and prohibits reverse writing to the crane PLC, retaining data interaction logs throughout the entire link. 5) Local Cache Unit: When the crane's 5G signal is temporarily weak and disconnected, short-term data is cached, and the transmission resumes after the signal is restored, avoiding the loss of operating condition and sensor data.

[0047] 5) Implementation of 5G private networks and vehicle control systems, such as Figure 2 The diagram shown is of the vehicle control system. The control system for the #5 450T crane consists of an ABB PLC and a frequency converter. The main hardware system is an ABB 800M PLC with IP address 192.168.100.81. A 5G device interface is added to the CMS system client, enabling wireless access via a 5G CPE, and port settings are configured within the 5G private network system.

[0048] 6) Implementation of 5G private networks and vehicle safety monitoring systems, such as Figure 3 The diagram shows the network topology of the remote safety monitoring system. The safety monitoring system for the 5#450T crane consists of a safety monitoring controller, a Siemens PLC, and a video system. The video system is networked, and the safety monitoring signals and video signals are interfaced and transmitted over the network, reaching the ground system via a 5G private network.

[0049] like Figure 4 This is a data tagging diagram of the crane. It automatically reads relevant data from the PLC and the operating handle positions, and combines this with real-time load data collection to accurately determine the crane's real-time load condition and direction of travel. The video monitoring system intelligently monitors and follows the hook, the load, and the hoisting mechanism components, while simultaneously recording operating signals and system operation signals. It reads the operating data of each mechanism in the system, automatically calculates the crane's operating frequency, load status, working time, and effective working cycles, and compiles the crane's operating load spectrum.

[0050] III. Core Functions of the Cloud-based AI Intelligent Analysis Module 1) Voiceprint auscultation sub-model: MFCC feature extraction + DBSCAN clustering is used to build a dedicated voiceprint library for reducers. The residual of newly collected sounds and features in the library is calculated to identify abnormal bearing noise. 2) Multi-parameter early warning sub-model for reducer: Based on LSTM long short-term memory network, it inputs multi-dimensional features of vibration, temperature and motor current under different working conditions, dynamically matches independent alarm thresholds for each working condition, and greatly reduces false alarms of fixed thresholds; 3) Load spectrum and life calculation module: Accumulate the number of daily working cycles and load weight, and combine them with the total number of design cycles of the crane to automatically calculate the remaining service life of the equipment.

[0051] This invention also provides a method for wireless transparent transmission of multi-source data in vehicles based on 5G-CPE, which, when applied to the above system, includes the following steps: Step S1: The terminal acquisition layer collects raw data from the vehicle's main control system, safety monitoring, video, vibration / temperature / acoustic sensor in parallel, and sends it to the vehicle-mounted 5G-CPE; Step S2: The multi-protocol adaptation and acquisition module inside the 5G-CPE uniformly parses and transforms the raw data of multiple protocols, and outputs standardized structured data frames; Step S3: Synchronously collect real-time operating parameters of vehicle load, speed, and orientation; the operating condition label generation module generates operating condition labels and embeds them into structured data frames. Step S4: The unified data encapsulation module encapsulates tagged data frames into data packets according to service priority, and transmits them wirelessly to the ground edge processing layer through the physically isolated industrial 5G private network in a hierarchical manner. Step S5: The edge server unpacks and splits the data packets, caches real-time driving data locally, and issues vehicle control commands to the 5G-CPE according to the on-site operation requirements; Step S6: The edge server periodically uploads all multi-source data with operating condition tags to the cloud-based intelligent analysis platform. Step S7: The cloud platform stores the entire lifecycle of vehicle driving data. The AI ​​fault diagnosis model combines the corresponding operating condition thresholds to analyze multi-source data, identify potential equipment faults, and push early warning information. Step S8: The security access control module executes the three-level directed access policy throughout the process, records all data interaction logs, and achieves secure and traceable transmission.

[0052] Preferably, in step S4, during industrial 5G private network transmission, a low-latency dedicated channel is allocated to the highest priority vehicle safety control signal, and a large-bandwidth time-division transmission channel is allocated to the low-priority high-definition video stream, so as to simultaneously complete the parallel backhaul of multiple types of data.

[0053] Preferably, in step S7, the cloud-based AI fault diagnosis model distinguishes different operating scenarios such as no-load, light-load, heavy-load, high-speed travel, and low-speed inspection based on the operating condition labels carried by the data. Each scenario is independently configured with fault alarm thresholds, and the analysis standards are dynamically matched to reduce the probability of false alarms and missed alarms.

[0054] This invention integrates a 5G-CPE unified architecture for steel plant mobile heavy-duty cranes, with a single gateway compatible with three types of heterogeneous devices: PLC industrial control, multi-channel video, and reducer temperature, vibration, and sound. It unifies multi-protocol parsing, completely solving the problem of data silos between multiple systems and replacing multiple independent data acquisition host computers. It designs an automatic tag binding mechanism for 12 standardized operating conditions of the crane, with each sensor and control data line accompanied by a real-time operating condition code. The cloud matches differentiated early warning thresholds according to the operating conditions, solving the industry problem of false alarms and missed alarms with fixed thresholds. It adopts a physically isolated local 5G private network + three-level directed access security mechanism, so that the crane monitoring transmission is not affected by the public network outage, eliminating the risk of control command tampering and data leakage.

[0055] The system adopts a 5G data hierarchical priority transmission strategy to differentiate channel resources for safety control, high-frequency sensing, and video data streams, simultaneously meeting the needs of low-latency hoisting monitoring and multi-channel high-definition video transmission. It integrates DBSCAN voiceprint clustering and LSTM multi-dimensional trend AI model to achieve predictive fault warnings for the reducer based on data with working condition labels. At the same time, it automatically calculates the remaining service life of the crane by statistically analyzing the load spectrum, replacing manual inspections at height and improving safety and maintenance efficiency. The 5G-CPE has a built-in local breakpoint caching mechanism to prevent the loss of key data on working conditions and equipment sensing when the crane's movement signal fluctuates.

[0056] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of the invention. All equivalent changes and modifications made in accordance with the scope of the claims of this invention should fall within the technical scope of this invention.

Claims

1. A wireless transparent transmission system for multi-source data in vehicles based on 5G-CPE, characterized in that, include: The four-tiered architecture of terminal-network-edge-cloud consists of terminal acquisition layer, aggregation and access layer, edge processing layer, and cloud intelligent analysis layer. The terminal acquisition layer includes a crane main control PLC, a safety monitoring PLC, video acquisition equipment, and a multi-dimensional sensing terminal, which are used to output crane control data, safety monitoring data, video streams, and raw data from equipment operating condition sensors. The aggregation access layer is a vehicle-mounted 5G-CPE gateway, and the 5G-CPE has a built-in multi-protocol adaptation and acquisition module, a working condition tag generation module, a unified data encapsulation module, and a secure access control module. The 5G-CPE serves as the sole data aggregation node for the vehicle's mobile terminal and communicates wirelessly with the industrial 5G private network, which is physically isolated from the public network. The edge processing layer includes a ground edge server and an engineer inspection workstation, used to receive and unpack data packets uploaded by 5G-CPE, cache real-time driving data, and send control commands to the vehicle. The cloud-based intelligent analysis layer is a multi-source data fusion AI analysis platform with a built-in time-series database and AI fault diagnosis model. It is used to receive tagged driving data synchronized from the edge layer and complete multi-source data fusion storage, operating condition matching analysis and equipment fault early warning. The 5G-CPE reads the multi-protocol raw data output by heterogeneous devices at the terminal acquisition layer through the multi-protocol adaptation acquisition module. After standardization and conversion, the data is bound to the real-time operating condition tag of the vehicle by the operating condition tag generation module, and then encapsulated into a unified data packet by the unified data encapsulation module. This data is then transmitted hierarchically to the edge processing layer through the industrial 5G private network and finally synchronized to the cloud intelligent analysis platform to complete AI intelligent diagnosis. The security access control module is configured with a three-level targeted access policy to restrict illegal cross-level data interaction.

2. The 5G-CPE-based wireless transparent transmission system for multi-source data in vehicles according to claim 1, characterized in that: The multi-protocol adaptation acquisition module is compatible with MODBUS TCP protocol, OPC UA protocol, and sensor device private communication protocol. It can parse four types of heterogeneous data in parallel: vehicle control, safety monitoring, video, and vibration / temperature / acoustic sensing data, and output structured data frames in a unified format.

3. The 5G-CPE-based wireless transparent transmission system for multi-source data in vehicles according to claim 1, characterized in that: The working condition label generation module collects working condition parameters such as vehicle load, travel speed, trolley / carriage travel direction, and lifting height in real time, encodes the working condition parameters to generate unique working condition labels, and embeds them into the header of each structured data frame to bind the data with the real-time operating status of the vehicle.

4. The 5G-CPE-based wireless transparent transmission system for multi-source data in vehicles according to claim 1, characterized in that: The unified data encapsulation module classifies data frames with operating condition labels into different levels and prioritizes them for transmission: driving safety control signals have the highest priority, equipment sensor operating condition data have the medium priority, and high-definition video streams have the lowest priority. The encapsulated data packets are marked with service priority identifiers, and the 5G private network dynamically allocates transmission bandwidth and latency resources based on these identifiers.

5. The 5G-CPE-based wireless transparent transmission system for multi-source data in vehicles according to claim 1, characterized in that: The secure access control module is configured with a three-level directional access policy, including: Level 1, only allowing the vehicle-mounted 5G-CPE to upload data unidirectionally to the edge processing layer; Level 2, only allowing the edge processing layer to issue legitimate inspection, speed adjustment, and shutdown control commands to the vehicle terminal; Level 3, the cloud-based intelligent analysis platform can only read historically stored data synchronized by the edge layer and prohibits reverse writing to the vehicle control system; all cross-level data interactions automatically retain access logs.

6. The 5G-CPE-based wireless transparent transmission system for multi-source data in vehicles according to claim 1, characterized in that: The cloud-based intelligent analysis platform incorporates AI fault diagnosis models, including vibration spectrum analysis models, equipment acoustic signature recognition models, and temperature trend prediction models. The AI ​​fault diagnosis models match specific fault alarm thresholds for corresponding operating conditions based on the operating condition tags carried in the data frames, and identify equipment faults such as motor overheating, bearing wear, and abnormal track noise based on multi-source fusion data, automatically generating maintenance early warning work orders.

7. The 5G-CPE-based wireless transparent transmission system for multi-source data in vehicles according to claim 1, characterized in that: The industrial 5G private network is equipped with 5G base stations in the factory area, and the vehicle-mounted 5G-CPE supports seamless and rapid switching of 5G cells during large-scale vehicle movement, maintaining a stable wireless communication link throughout the process and avoiding data transmission interruptions and lags when the vehicle is moving.

8. A method for wireless transparent transmission of multi-source data for vehicle operation based on 5G-CPE, applied to the system described in any one of claims 1-7, characterized in that, Includes the following steps: Step S1: The terminal acquisition layer collects raw data from the vehicle's main control system, safety monitoring, video, vibration / temperature / acoustic sensor in parallel, and sends it to the vehicle-mounted 5G-CPE; Step S2: The multi-protocol adaptation and acquisition module inside the 5G-CPE uniformly parses and transforms the raw data of multiple protocols, and outputs standardized structured data frames; Step S3: Synchronously collect real-time operating parameters of vehicle load, speed, and orientation; the operating condition label generation module generates operating condition labels and embeds them into structured data frames. Step S4: The unified data encapsulation module encapsulates tagged data frames into data packets according to service priority, and transmits them wirelessly to the ground edge processing layer through the physically isolated industrial 5G private network in a hierarchical manner. Step S5: The edge server unpacks and splits the data packets, caches real-time driving data locally, and issues vehicle control commands to the 5G-CPE according to the on-site operation requirements; Step S6: The edge server periodically uploads all multi-source data with operating condition tags to the cloud-based intelligent analysis platform. Step S7: The cloud platform stores the entire lifecycle of vehicle driving data. The AI ​​fault diagnosis model combines the corresponding operating condition thresholds to analyze multi-source data, identify potential equipment faults, and push early warning information. Step S8: The security access control module executes the three-level directed access policy throughout the process, records all data interaction logs, and achieves secure and traceable transmission.

9. The method for wireless transparent transmission of multi-source data for vehicle operation based on 5G-CPE according to claim 8, characterized in that: In step S4, during industrial 5G private network transmission, a low-latency dedicated channel is allocated to the highest priority vehicle safety control signal, and a high-bandwidth time-division transmission channel is allocated to the low-priority high-definition video stream, so as to simultaneously complete the parallel backhaul of multiple types of data.

10. The method for wireless transparent transmission of multi-source data for vehicle operation based on 5G-CPE according to claim 8, characterized in that: In step S7, the cloud-based AI fault diagnosis model distinguishes different operating scenarios such as no-load, light-load, heavy-load, high-speed travel, and low-speed inspection based on the operating condition tags carried by the data. Each scenario is independently configured with fault alarm thresholds, and the analysis standards are dynamically matched to reduce the probability of false alarms and missed alarms.

Images