Plant road traffic safety intelligent management and control system

Abstract

The present application belongs to the technical field of road management and control system, especially to a factory road traffic safety intelligent management and control system, comprising the following function modules and steps: data acquisition layer: multi-modal sensor array deployed at the key nodes of the factory road; the present application comprehensively and real-time collects the factory road traffic data through the multi-modal sensor array, the edge computing layer quickly and accurately detects and tracks the traffic target, identifies the illegal behavior, the central control platform dynamically adjusts the signal light timing, predicts the congestion trend, realizes the comprehensive analysis and intelligent regulation and control of the factory traffic. At the same time, the green wave pass of the special vehicle is automatically triggered, the accident is quickly located, the resources are dispatched and the area is controlled, the illegal behavior is effectively handled, the congestion is prevented in advance, the rapid and safe pass of the special vehicle is ensured, the emergency response ability is improved, the influence of the accident on the traffic is reduced, the rescue work is smoothly carried out, and the safety and efficiency of the factory road traffic are comprehensively improved.

Description

Technical Field

[0001] This invention relates to the field of road control system technology, and in particular to an intelligent control system for traffic safety on factory roads. Background Technology

[0002] With the rapid development of industry, the scale of various factories is constantly expanding, and the flow of vehicles and personnel within these areas is becoming increasingly frequent, making road traffic safety issues more and more prominent. Factory roads, as a crucial infrastructure for production and operation, directly impact production efficiency, material transportation, and personnel safety. Currently, many factories still employ traditional methods for road traffic management, which are ill-suited to the complex and ever-changing traffic demands of modern factories.

[0003] Incomplete and unreal-time data collection: Traditional factory traffic management relies mainly on manual patrols and a limited number of fixed monitoring devices, which cannot comprehensively and in real-time obtain various traffic data of factory roads, such as traffic flow, vehicle speed, pedestrian location, and violations. This makes it difficult for managers to grasp road traffic conditions in a timely manner, hindering their ability to make scientific and reasonable decisions, leading to frequent traffic congestion and safety accidents.

[0004] Untimely identification and handling of violations: Under the traditional management model, the discovery of violations mainly relies on manual methods, which leads to problems such as untimely discovery and insufficient evidence. Moreover, the handling process for violations is cumbersome and inefficient, failing to effectively deter violators and thus making it difficult to curb violations, seriously affecting traffic order and safety in the factory area.

[0005] Poor emergency response capability: When emergencies such as traffic accidents or hazardous chemical leaks occur in the factory area, traditional traffic management systems cannot quickly and accurately locate the accident site, dispatch rescue resources, or adjust traffic signals in a timely manner to guide vehicles and personnel to avoid dangerous areas, which hinders emergency rescue work and may cause greater casualties and property losses.

[0006] Lack of priority protection for special vehicles: Special vehicles such as hazardous chemical transport vehicles and fire trucks in the factory area need to pass quickly when performing tasks, but the traditional traffic management system cannot provide them with priority passage. Traffic congestion can easily cause delays for special vehicles, affecting emergency rescue and production safety.

[0007] Delayed monitoring and handling of equipment malfunctions: Traditional factory traffic management equipment lacks an effective health monitoring mechanism, making it impossible to detect equipment malfunctions in a timely manner. Once a malfunction occurs, it often requires manual inspection to detect, which can lead to interruption of data acquisition or failure of signal control, affecting the normal operation of the entire traffic management system.

[0008] Low communication reliability: Traditional communication methods are prone to signal interference and interruptions in complex factory environments, leading to unstable data transmission. Once communication is interrupted, the traffic management system will be unable to obtain real-time data or remotely control traffic equipment, severely impacting traffic control effectiveness.

[0009] Poor system stability: Traditional traffic management systems have inadequate software architecture design and lack effective fault tolerance and recovery mechanisms. When a module in the system fails, it may cause the entire system to crash, requiring a long time to recover and causing great inconvenience to traffic management in the factory area. Summary of the Invention

[0010] To address the aforementioned problems, this invention proposes an intelligent management and control system for factory road traffic safety, which more accurately solves the problems mentioned in the background section.

[0011] This invention is achieved through the following technical solution: This invention proposes a smart traffic safety management system for factory roads, comprising the following functional modules and steps: Data acquisition layer: A multimodal sensor array deployed at key nodes of factory roads, including a geomagnetic vehicle detector, an infrared pedestrian detector, a radar speedometer, and a high-definition camera supporting H.265 encoding. Each sensor communicates with the edge computing unit via an RS485 / Ethernet interface to collect real-time data on traffic flow, vehicle speed, pedestrian location, and violations, with a sampling frequency ≥10Hz. The high-definition camera has a wide dynamic range (WDR) function, a dynamic contrast ratio ≥120dB, and supports clear capture of license plate information in strong light / backlight environments. Edge computing layer: An embedded system integrated into the traffic signal control box. The computing unit is equipped with a lightweight YOLOv8-tiny object detection model accelerated by TensorRT and an improved SORT multi-object tracking algorithm, with a model inference speed of ≤30ms / frame. It has localized violation recognition capabilities, independently determining behaviors such as running red lights, driving against traffic, speeding (accuracy ±3%), and occupying emergency lanes, and generating data packets containing timestamps, location coordinates, and violation types. The central control platform is a microservice cluster built on a Kubernetes containerized architecture, including a traffic flow analysis subsystem, a signal control subsystem, and an emergency command subsystem. Each subsystem achieves low-latency (≤50ms) data interaction via the gRPC protocol. Dynamic signal control is also supported. Modules: Employs an improved Webster timing algorithm, dynamically adjusting traffic light phases and green light ratios based on real-time traffic flow data, supporting tidal lane direction switching; Violation tracing module: Links multi-camera detection results through spatiotemporal data fusion technology, utilizes the Hungarian algorithm to reconstruct vehicle trajectories, and automatically generates an electronic evidence chain compliant with the "Regulations on the Procedures for Handling Road Traffic Safety Violations"; Congestion prediction module: Based on an LSTM neural network model, predicts congestion trends for the next 15-30 minutes with an accuracy rate ≥85%; Interactive terminal cluster: Control terminal: An industrial-grade touchscreen all-in-one machine deployed in the guardhouse, supporting real-time map visualization, remote signal control, and equipment status monitoring; Mobile enforcement APP: Integrated The system includes: a traffic violation on-site processing function that allows law enforcement officers to quickly retrieve historical violation records by reading vehicle RFID tags (if applicable) via NFC and generate a penalty decision document with an electronic signature; an LED guidance screen using a P3-level full-color display that supports dynamic zone display, with a warning mode highlighting detour routes in red; an energy management module with a dual power supply system of solar and mains power, including an MPPT controller and lithium iron phosphate battery pack, supporting operation in a wide temperature range of -20℃ to +60℃; a three-level energy scheduling strategy that prioritizes solar power, allocates remaining power according to device priority (signal lights > cameras > guidance screens), and automatically activates the heating film in low-temperature environments; and a secure communication module using TLS 1 for the transmission layer.3. Protocol encryption ensures data transmission latency ≤20ms; deploy an intrusion detection system based on the Suricata rule engine to monitor abnormal access behavior (such as port scanning and brute-force attacks) in real time and trigger alarms, with a response time ≤1s.

[0012] Preferably, the special vehicle priority strategy of the signal control subsystem is implemented through the following technologies: RFID readers identify electronic tags affixed to hazardous chemical transport vehicles and fire trucks, with the tags storing vehicle type, affiliated unit, and emergency contact information; when a special vehicle is detected entering a preset range, a green wave traffic signal is automatically triggered: the traffic lights at the three intersections ahead are adjusted to green, and the green light duration is extended until the special vehicle can pass; the guidance screen displays the text "Special vehicle, please give way" and an arrow indicating this; and a notification is simultaneously pushed to the factory's safety management department terminal.

[0013] Preferably, the accident handling process of the emergency command subsystem includes the following steps: Accident location: By fusing spatiotemporal data from cameras and geomagnetic sensors with a time synchronization error ≤5ms, and combining this with the Dijkstra algorithm, the path distance from the accident point to the nearest rescue point is calculated; Resource scheduling: Rescue resources are automatically matched according to the accident type, and a dispatch order containing vehicle number, departure time, and estimated arrival time is generated; Area control: The accident-related intersections are closed, with the closure area being two intersections upstream and downstream of the accident point; Activation of alternative route signal control plan: The traffic flow of the original accident route is diverted to parallel roads on both sides at a ratio of 3:2; Detour prompts are issued through guidance screens and mobile law enforcement APP.

[0014] Preferably, the violation tracing function enhances the evidentiary effect through the following technologies: multi-camera collaboration: calling video data from two cameras upstream and downstream of the accident point to ensure coverage of the complete process from 30 seconds before to 10 seconds after the violation; data tamper-proofing: using blockchain technology to store electronic evidence (hash value on the chain) to ensure that video clips and metadata cannot be tampered with; judicial connection: the evidence chain format complies with the requirements of the "Administrative Case Procedure Regulations" and can be directly used as the basis for administrative penalties.

[0015] Preferably, the system is deployed in a private cloud environment and adopts the following security mechanisms: Identity authentication: multi-factor authentication based on the OAuth2.0 protocol, including password, SMS verification code and dynamic token; Access control: adopting the RBAC model, defining three types of roles: administrator, operator and auditor, with permission granularity refined to the functional module level; Log auditing: recording all system operation logs, with a storage period of ≥180 days, and supporting multi-dimensional retrieval by time, user and operation type.

[0016] Preferably, the system supports integration with third-party systems, providing the following standardized interfaces: Device layer interface: Modbus TCP protocol for connecting to geomagnetic detectors and radar speedometers; ONVIF protocol for integrating third-party cameras; GB / T 28181 protocol for interfacing with video private networks; Platform layer interface: RESTful API, supporting HTTP / HTTPS protocols for data querying and control command issuance; MQTT protocol for real-time data push, with a QoS level ≥1; Kafka message queue for high-concurrency data stream processing, with a throughput ≥100,000 messages / second; Business layer interface: providing an SDK development package to support integration with factory ERP and OA systems, enabling automatic synchronization of violation data to employee performance files.

[0017] Preferably, the system has self-diagnosis and fault tolerance capabilities, including the following mechanisms: device health monitoring: the validity of sensor data is periodically detected by the edge computing unit, with a fault detection cycle of ≤1 minute; communication redundancy design: when the main communication link (5G / fiber) fails, it automatically switches to the backup link, with a switching time of ≤3 seconds; software watchdog: each microservice process of the central control platform is equipped with an independent watchdog timer, and the service is automatically restarted if no heartbeat signal is received after the timeout.

[0018] Preferably, the system is applicable to the following factory scenarios: Chemical industrial parks: managing hazardous chemical transport vehicles through dedicated lanes and linking with gas leak detection systems; Logistics parks: supporting priority passage for container trucks and optimizing signal timing around loading and unloading areas; Manufacturing parks: dynamically adjusting traffic light timing during peak hours based on shift data (such as morning / evening shifts) to reduce employee commuting waiting time.

[0019] Preferably, the system reduces deployment costs through the following technologies: hardware reuse: installing cameras and geomagnetic detectors on existing factory street light poles to reduce the construction of independent poles; software-defined traffic: managing traffic signals of different brands and models through a central control platform to avoid equipment replacement costs; cloud-edge collaboration: migrating non-real-time computing tasks to the cloud to reduce the hardware configuration requirements of edge computing units.

[0020] Preferably, the system enhances user experience through the following technologies: Mobile-friendly design: The mobile law enforcement APP interface adopts Material Design specifications and supports dark mode and voice input; Accessibility: The guidance screen supports voice broadcasting, making it convenient for visually impaired people to obtain information; Multilingual support: The central control platform and the mobile law enforcement APP provide bilingual interfaces in Chinese and English to meet the needs of foreign employees.

[0021] Compared with the prior art, the present invention provides a smart management and control system for traffic safety in factory areas, which has the following beneficial effects: This intelligent traffic safety management system for the factory area comprehensively collects real-time traffic data through a multimodal sensor array. The edge computing layer quickly and accurately detects and tracks traffic targets and identifies violations. The central control platform dynamically adjusts traffic light timings and predicts congestion trends, enabling comprehensive analysis and intelligent control of traffic within the factory area. Simultaneously, it automatically triggers green wave traffic signals for special vehicles, quickly locates and dispatches resources to control the area in the event of an accident, effectively handles violations, prevents congestion in advance, ensures the rapid and safe passage of special vehicles, improves emergency response capabilities, reduces the impact of accidents on traffic, ensures the smooth progress of rescue work, and comprehensively enhances the safety and efficiency of road traffic within the factory area.

[0022] The intelligent traffic safety management system for the factory area employs multiple technologies to ensure stable operation. Equipment health monitoring promptly detects the validity of sensor data and issues fault alarms; communication redundancy automatically switches to a backup link in the event of a primary communication link failure; and a software watchdog automatically restarts services when microservice processes fail. These measures improve system reliability and stability, reduce downtime caused by equipment failures or communication interruptions, and ensure continuous and stable operation under various conditions, providing reliable support for traffic management within the factory area.

[0023] This intelligent traffic safety management system for the factory area reuses existing streetlights and installation equipment, saving construction costs and reducing environmental impact. Software-defined traffic management unifies the management of different brands and models of traffic signals, avoiding equipment replacement costs and improving system flexibility and scalability. Cloud-edge collaboration reduces the hardware configuration requirements of edge computing units, improving overall system performance and cost-effectiveness, and lowering deployment costs. Furthermore, its mobile-friendly design, accessibility features, and multilingual support meet the needs of different users in various scenarios, enhancing user experience and making the system more user-friendly and international. Attached Figure Description

[0024] Figure 1 This is a flowchart illustrating a smart management and control system for factory road traffic safety proposed in this invention. Detailed Implementation

[0025] To more clearly and completely illustrate the technical solution of the present invention, the present invention will be further described below with reference to the accompanying drawings.

[0026] like Figure 1As shown in the figure, an embodiment of the present invention proposes a smart traffic safety management system for factory roads. The data acquisition layer involves deploying multimodal sensor arrays at key nodes on factory roads, such as entrances / exits, intersections, and curves. Specifically, this includes geomagnetic vehicle detectors for real-time sensing of vehicle passage and traffic flow; infrared pedestrian detectors for precise pedestrian location; radar speedometers for measuring vehicle speed; and high-definition cameras supporting H.265 encoding with wide dynamic range (WDR) and a dynamic contrast ratio of 120dB, capable of clearly capturing license plate information in strong light or backlight conditions. Each sensor communicates with the edge computing unit via RS485 or Ethernet interfaces, collecting real-time data on traffic flow, vehicle speed, pedestrian location, and traffic violations at a sampling frequency of at least 10Hz. This system comprehensively, in real-time, and accurately acquires various types of traffic data from factory roads, providing a foundation for subsequent analysis and management. Edge Computing Layer: An embedded computing unit is integrated within the traffic signal control box, equipped with a lightweight YOLOv8-tiny object detection model accelerated by TensorRT and an improved SORT multi-object tracking algorithm, ensuring that the model inference speed does not exceed 30ms / frame. This computing unit has localized violation recognition capabilities, independently determining behaviors such as running red lights, driving against traffic, speeding (accuracy ±3%), and occupying emergency lanes, and generating data packets containing timestamps, location coordinates, and violation types. It quickly and accurately detects and tracks traffic targets, promptly identifying violations and providing timely decision-making basis for traffic management. Central Control Platform: A microservice cluster built on a Kubernetes containerized architecture includes a traffic flow analysis subsystem, a signal control subsystem, and an emergency command subsystem. Each subsystem achieves low-latency (not exceeding 50ms) data interaction through the gRPC protocol. Dynamic Signal Control Module: Employing an improved Webster timing algorithm, it dynamically adjusts the signal light phase and green light ratio based on real-time traffic flow data, supporting tidal lane direction switching. For example, during morning rush hour, the direction of traffic in tidal flow lanes is automatically adjusted based on changes in traffic flow, improving road efficiency. The violation tracing module uses spatiotemporal data fusion technology to link multiple camera detection results and employs a Hungarian algorithm to reconstruct vehicle trajectories. When a violation occurs, an electronic evidence chain conforming to the "Regulations on the Procedures for Handling Road Traffic Safety Violations" is automatically generated, including the time, location, vehicle information, and video footage of the violation.

[0027] Congestion Prediction Module: Based on an LSTM neural network model, this module uses historical and real-time traffic data for training and prediction to forecast congestion trends 15-30 minutes in advance, achieving an accuracy rate of over 85%. For example, it can predict potential congestion on a specific road segment during a specific time period and take timely measures to alleviate it. This enables comprehensive analysis and intelligent control of traffic within the factory area, effectively handling violations, preventing congestion in advance, and ensuring road traffic safety and smooth flow.

[0028] Interactive Terminal Cluster: Management Terminal: An industrial-grade touch screen all-in-one machine is deployed in the guard room, supporting real-time map visualization, which can intuitively display information such as traffic conditions and equipment locations in the factory area; it supports remote signal control, allowing managers to remotely adjust traffic light timings; it supports equipment status monitoring, enabling real-time understanding of the operating status of each sensor and device.

[0029] Mobile enforcement app: Integrates on-site violation processing functions. Enforcement officers can quickly retrieve historical violation records by reading vehicle RFID tags (if applicable) via NFC and generate penalty decisions with electronic signatures. For example, when investigating a vehicle for a violation, enforcement officers only need to bring their mobile phones close to the vehicle's RFID tag to quickly obtain vehicle information, improving enforcement efficiency.

[0030] LED guidance screen: Utilizing a P3-level full-color display, it supports dynamic zone display. In warning mode, it highlights detour routes in red to guide vehicles away from congested or accident-prone areas. This facilitates operation and monitoring by management and law enforcement personnel, providing drivers with timely traffic information guidance.

[0031] Energy Management Module: Employs a dual solar-mains power supply system, including an MPPT controller and lithium iron phosphate battery pack, supporting operation over a wide temperature range of -20℃ to +60℃. It features a three-level energy dispatch strategy: prioritizing solar power; allocating remaining power according to device priority (signal lights > cameras > guidance screens); and automatically activating the heating film in low-temperature environments to ensure normal equipment operation. This reduces energy consumption, improves energy efficiency, and ensures stable system operation under various environmental conditions.

[0032] Secure communication module: The transport layer uses TLS 1.3 encryption to ensure data transmission security, with a data transmission latency of no more than 20ms. An intrusion detection system based on the Suricata rule engine is deployed to monitor abnormal access behavior (such as port scanning and brute-force attacks) in real time. When abnormal behavior is detected, an alarm is triggered within 1 second to notify administrators for timely handling. This ensures the security and stability of system communication and prevents data leakage and unauthorized access.

[0033] In this invention, electronic tags are affixed to special vehicles such as hazardous chemical transport vehicles and fire trucks on factory roads. These tags store information such as vehicle type, affiliated unit, and emergency contact information. RFID readers are installed at key locations on factory roads. When a special vehicle enters a preset range, the RFID reader reads the electronic tag information and transmits it to the signal control subsystem. The signal control subsystem automatically triggers a green wave traffic signal: adjusting the traffic lights at the three intersections ahead to green and extending the green light duration until the special vehicle can pass, based on its speed and distance. The system displays the message "Special vehicle, please give way" along with arrows on guidance screens to guide other vehicles to give way. A notification is simultaneously pushed to the factory safety management department's terminal, allowing them to be promptly informed of the special vehicle's passage. This ensures the rapid and safe passage of special vehicles within the factory area and improves emergency response capabilities.

[0034] In this invention, the following steps are implemented: Accident Location: When an accident occurs, spatiotemporal data from cameras and geomagnetic sensors are fused, with a time synchronization error not exceeding 5ms. The Dijkstra algorithm is used to calculate the path distance from the accident point to the nearest rescue point, determining the accident location and rescue route. Resource Scheduling: Rescue resources, such as fire trucks, ambulances, and rescue personnel, are automatically matched based on the accident type (e.g., vehicle collision, fire). A dispatch order containing vehicle number, departure time, and estimated arrival time is generated and sent to rescue personnel and relevant departments. Area Control: The accident-related intersections are closed, with the closure area consisting of two intersections upstream and downstream of the accident point to prevent other vehicles from entering the accident area and causing secondary accidents. Alternative Route Signal Control Plan: Traffic flow from the original accident route is diverted to parallel roads on both sides at a 3:2 ratio to alleviate traffic pressure. Detour prompts are issued through guidance screens and a mobile enforcement app to guide vehicles away from the accident section. This rapid and effective handling of accidents minimizes their impact on traffic and ensures the smooth progress of rescue operations.

[0035] In this invention, multiple cameras work together: when a traffic violation occurs, video data from two cameras upstream and downstream of the accident site is retrieved, ensuring coverage of the entire process from 30 seconds before to 10 seconds after the violation, comprehensively recording the entire course of the violation. Data tamper-proofing: Blockchain technology is used to store electronic evidence (hash values ​​are uploaded to the blockchain). Video clips and metadata are encrypted and stored on the blockchain, ensuring that the video clips and metadata cannot be tampered with, guaranteeing the authenticity and reliability of the evidence. Judicial linkage: the evidence chain format complies with the requirements of the "Administrative Case Procedure Regulations" and can be directly used as the basis for administrative penalties. When generating the electronic evidence chain, it is organized and stored according to the prescribed format and requirements for convenient use by authorities. This enhances the effectiveness of evidence of traffic violations and provides strong legal support for traffic violation penalties.

[0036] In this invention, identity authentication is implemented using the OAuth 2.0 protocol for multi-factor authentication. When logging into the system, users need to enter a password, receive an SMS verification code, and use a dynamic token for verification, ensuring that only authorized users can access the system. Access control adopts the RBAC model, defining three roles: administrator, operator, and auditor. Permission granularity is refined to the functional module level. For example, administrators have the highest system privileges and can operate on all functional modules; operators can only perform specific business operations, such as data queries and device control; auditors are responsible for auditing and supervising system operations. Log auditing records all system operation logs, including user login time, operation content, and operation results, with a storage period of no less than 180 days. Multi-dimensional retrieval by time, user, and operation type is supported, facilitating auditing and tracing of system operations by administrators. This ensures system security and compliance, prevents unauthorized access and operation, and promptly detects and handles security issues.

[0037] In this invention, the device layer interface employs the Modbus TCP protocol to connect to devices such as geomagnetic detectors and radar speedometers for data acquisition and transmission. It uses the ONVIF protocol to integrate third-party cameras, facilitating system access to cameras of different brands and models. The GB / T 28181 protocol is used for interfacing with a dedicated video network, enabling video data sharing and interaction. The platform layer interface provides a RESTful API supporting HTTP / HTTPS protocols for data querying and control command issuance. Other systems can obtain system data or send control commands to the system by calling the API. The MQTT protocol is used for real-time data push, with a QoS level of at least 1 to ensure reliable data transmission. For example, real-time traffic data can be pushed to relevant application systems. A Kafka message queue is used for high-concurrency data stream processing, with a throughput of at least 100,000 messages per second. This ensures efficient system operation when processing large amounts of traffic data. The business layer interface provides an SDK development package supporting integration with factory ERP and OA systems. By calling the SDK interface, violation data can be automatically synchronized to employee performance records, facilitating enterprise management and assessment of employee traffic violations. To achieve seamless integration between the system and third-party systems, improve system compatibility and scalability, and meet the needs of different users.

[0038] In this invention, device health monitoring involves periodically checking the validity of sensor data via an edge computing unit, with a fault detection cycle of no more than one minute. For example, data from sensors such as geomagnetic vehicle detectors and infrared pedestrian detectors are checked every minute to determine if the sensors are functioning correctly. If abnormal sensor data is detected, an alarm is promptly issued to notify maintenance personnel for repair. Communication redundancy design: The main communication link uses 5G or fiber optic communication. When the main communication link fails, the system automatically switches to a backup link, such as 4G or satellite communication, with a switching time of no more than three seconds, ensuring uninterrupted system communication. Software watchdog: An independent watchdog timer is set in each microservice process on the central control platform. If a heartbeat signal is not received within the timeout period, the service is automatically restarted. For example, each microservice process sends a heartbeat signal to the watchdog at regular intervals. If the watchdog does not receive a heartbeat signal within the specified time, it considers the microservice process to have failed and automatically restarts the service, ensuring stable system operation. This improves system reliability and stability, reducing system downtime caused by device failures or communication interruptions.

[0039] In this invention, for chemical industrial parks: dedicated lanes for hazardous chemical transport vehicles are established, and the system manages these lanes to restrict the entry of other vehicles. Simultaneously, a gas leak detection system is integrated; when a gas leak is detected, traffic signals are adjusted promptly to guide vehicles and personnel away from the danger zone. For logistics parks: priority passage for container trucks is supported by a signal control subsystem, reducing their waiting time. Signal timing around loading and unloading areas is optimized, dynamically adjusting signal timing based on operational conditions to improve traffic efficiency. For manufacturing parks: based on shift data (e.g., morning / night shifts), the commuting time and traffic flow of employees on different shifts are analyzed, and signal timing during peak hours is dynamically adjusted to reduce employee commuting wait times and improve travel efficiency. This addresses the specific needs of different plant scenarios, improving the targeting and effectiveness of plant traffic management.

[0040] This invention features several key advantages: Hardware reuse: Cameras and geomagnetic detectors are installed on existing factory streetlights, reducing the need for independent pole construction. For example, installing high-definition cameras and geomagnetic vehicle detectors on streetlights saves construction costs and reduces the impact of poles on the factory environment. Software-defined traffic: A central control platform manages different brands and models of traffic signals, developing a unified control interface and protocol to ensure system compatibility with various signals. This avoids costs associated with equipment replacement and improves system flexibility and scalability. Cloud-edge collaboration: Non-real-time computing tasks, such as traffic flow analysis and historical data storage, are migrated to the cloud, reducing the hardware requirements of edge computing units. Edge computing units are primarily responsible for real-time data acquisition and preliminary processing, uploading the processed data to the cloud for further analysis and storage, improving overall system performance and cost-effectiveness. This reduces system deployment costs, improves resource utilization efficiency, and makes the system more economical and feasible.

[0041] This invention features a mobile-friendly design: the mobile law enforcement app interface adopts Material Design specifications, making it simple, aesthetically pleasing, and easy to use. It supports a dark mode, facilitating use by law enforcement officers at night or in low-light environments; it also supports voice input, allowing officers to quickly input information via voice, improving efficiency. Accessibility features include: the guidance screen supports voice broadcasting; when important information needs to be highlighted, the screen not only displays text information but also delivers it to visually impaired individuals via voice broadcast, facilitating information access for them. Multilingual support is provided: the central control platform and mobile law enforcement app offer bilingual (Chinese and English) interfaces to meet the needs of foreign staff. Both Chinese and English versions are provided in the interface design and operation prompts to facilitate understanding and use of the system by foreign staff. This enhances the user experience, meets the needs of different users, and makes the system more user-friendly and international.

[0042] Finally, it should be noted that the basic concepts have been described above. Obviously, for those skilled in the art, the detailed disclosure above is merely illustrative and does not constitute a limitation of this specification. Although not explicitly stated herein, those skilled in the art may make various modifications, improvements, and corrections to this specification. Such modifications, improvements, and corrections are suggested in this specification, and therefore remain within the spirit and scope of the exemplary embodiments of this specification. Furthermore, this specification uses specific terms to describe embodiments of this specification. For example, "an embodiment," "one embodiment," and / or "some embodiments" refer to a feature, structure, or characteristic associated with at least one embodiment of this specification. Therefore, it should be emphasized and noted that "an embodiment," "one embodiment," or "an alternative embodiment" mentioned twice or more in different locations in this specification do not necessarily refer to the same embodiment. In addition, certain features, structures, or characteristics in one or more embodiments of this specification can be appropriately combined. Moreover, unless expressly stated in the claims, the order of processing elements and sequences, the use of numbers and letters, or other names described in this specification are not intended to limit the order of the processes and methods of this specification.

[0043] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A smart management and control system for traffic safety on factory roads, characterized in that, Includes the following functional modules and steps: Data Acquisition Layer: The multimodal sensor array deployed at key nodes of the factory roads includes a geomagnetic vehicle detector, an infrared pedestrian detector, a radar speed detector, and a high-definition camera that supports H.265 encoding. Each sensor communicates with the edge computing unit through an RS485 / Ethernet interface to collect real-time data on traffic flow, vehicle speed, pedestrian location, and violations, with a sampling frequency ≥10Hz. The high-definition camera has a wide dynamic range (WDR) function and a dynamic contrast ratio of ≥120dB, which supports clear capture of license plate information in strong light / backlight environments. Edge computing layer: The embedded computing unit, integrated into the traffic signal control box, is equipped with a lightweight YOLOv8-tiny target detection model based on TensorRT acceleration and an improved SORT multi-target tracking algorithm, with a model inference speed of ≤30ms / frame; It has the ability to identify localized traffic violations, and can independently determine behaviors such as running red lights, driving against traffic, speeding (accuracy ±3%), and occupying emergency lanes, and generate data packets containing timestamps, location coordinates and violation types; Central control platform: The microservice cluster built on the Kubernetes container architecture includes a traffic flow analysis subsystem, a signal control subsystem, and an emergency command subsystem. Each subsystem achieves low-latency (≤50ms) data interaction through the gRPC protocol. Dynamic signal control module: It adopts an improved Webster timing algorithm and dynamically adjusts the phase and green light ratio of traffic lights in combination with real-time traffic flow data, and supports tidal lane direction switching; Violation tracing module: By using spatiotemporal data fusion technology to associate the detection results of multiple cameras, the Hungarian algorithm is used to reconstruct vehicle trajectories and automatically generate an electronic evidence chain that complies with the "Regulations on the Procedures for Handling Road Traffic Safety Violations". Congestion prediction module: Based on an LSTM neural network model, it predicts congestion trends for the next 15-30 minutes with an accuracy of ≥85%. Interactive terminal cluster: Control terminal: An industrial-grade touch screen all-in-one machine deployed in the guard room, supporting real-time map visualization, remote signal control and equipment status monitoring; Mobile enforcement APP: integrates on-site violation processing function, supports law enforcement officers to quickly retrieve historical violation records by reading vehicle RFID tags via NFC (if applicable), and generate penalty decision letters with electronic seals; LED guidance screen: It adopts a P3 level full-color display screen and supports dynamic display of zones. In the warning mode, the detour route is displayed in red. Energy Management Module: The dual power supply system of solar and mains includes an MPPT controller and a lithium iron phosphate battery pack, supporting operation in a wide temperature range of -20℃ to +60℃. It features a three-level energy dispatch strategy: prioritizing solar power, allocating remaining power according to equipment priority (signal lights > cameras > guidance screens), and automatically activating the heating film in low-temperature environments. Secure communication module: The transport layer uses TLS 1.3 encryption, and the data transmission latency is ≤20ms; Deploy an intrusion detection system based on the Suricata rule engine to monitor abnormal access behavior (such as port scanning and brute-force attacks) in real time and trigger alarms with a response time of ≤1s.

2. The system according to claim 1, characterized in that, The special vehicle priority strategy of the signal control subsystem is implemented through the following technologies: The electronic tags affixed to hazardous chemical transport vehicles and fire trucks are identified by RFID readers. The tags store the vehicle type, the unit to which the vehicle belongs, and emergency contact information. When a special vehicle is detected entering the preset range, a green wave traffic signal is automatically triggered: Adjust the traffic lights at the three intersections ahead to green and extend the green light duration until special vehicles can pass; The guidance screen displays the text "Special vehicles are passing, please give way" and arrow indicators; The notification was simultaneously pushed to the terminal of the factory's safety management department.

3. The system according to claim 1, characterized in that, The accident handling process of the emergency command subsystem includes the following steps: Accident location: By fusing spatiotemporal data from cameras and geomagnetic sensors with a time synchronization error of ≤5ms, and combining this with the Dijkstra algorithm, the path distance from the accident point to the nearest rescue point is calculated. Resource scheduling: Automatically matches rescue resources based on the type of accident and generates a dispatch order that includes vehicle number, departure time and estimated arrival time; Regional control: Close the intersections associated with the accident, covering two intersections upstream and two downstream of the accident site; Activate the alternative route signal control plan: divert traffic from the original accident route to parallel roads on both sides at a 3:2 ratio; Detour notices were issued via navigation screens and mobile law enforcement apps.

4. The system according to claim 1, characterized in that, The violation tracing function enhances the evidentiary value through the following technologies: Multi-camera collaboration: It calls on video data from two cameras upstream and downstream of the accident site to ensure coverage of the entire process from 30 seconds before the violation occurred to 10 seconds after. Data tamper-proof: Blockchain technology is used to store electronic evidence (hash values ​​are recorded on the chain) to ensure that video clips and metadata cannot be tampered with.

5. The system according to claim 1, characterized in that, The system is deployed in a private cloud environment and employs the following security mechanisms: Identity authentication: Implements multi-factor authentication based on the OAuth 2.0 protocol, including password, SMS verification code and dynamic token; Access control: The RBAC model is adopted, defining three types of roles: administrator, operator, and auditor, with permission granularity refined to the functional module level; Log auditing: Records all system operation logs, with a storage period of ≥180 days, and supports multi-dimensional retrieval by time, user, and operation type.

6. The system according to claim 1, characterized in that, The system supports integration with third-party systems and provides the following standardized interfaces: Device layer interface: The Modbus TCP protocol is used to connect geomagnetic detectors and radar speedometers. The ONVIF protocol is used for integrating third-party cameras; The GB / T 28181 protocol is used for interfacing with private video networks; Platform layer interface: A RESTful API, supporting HTTP / HTTPS protocols, is used for data querying and control command issuance. The MQTT protocol is used for real-time data push, with a QoS level of ≥1. Kafka message queues are used for high-concurrency data stream processing with a throughput of ≥100,000 messages / second. Business layer interface: An SDK development package is provided to support integration with the factory's ERP and OA systems, enabling automatic synchronization of violation data to employee performance records.

7. The system according to claim 1, characterized in that, The system possesses self-diagnosis and fault tolerance capabilities, including the following mechanisms: Equipment health monitoring: The validity of sensor data is checked periodically by the edge computing unit, with a fault detection cycle of ≤1 minute; Communication redundancy design: When the main communication link (5G / fiber) fails, it automatically switches to the backup link with a switching time of ≤3 seconds; Software watchdog: Each microservice process on the central control platform is equipped with an independent watchdog timer. If no heartbeat signal is received within the timeout period, the service will be automatically restarted.

8. The system according to claim 1, characterized in that, The system is suitable for the following factory scenarios: Chemical industrial parks: Controlled through dedicated lanes for hazardous chemical transport vehicles, and linked to gas leak detection systems; Logistics parks: Priority passage for container trucks and optimized signal timing around loading and unloading areas; Manufacturing Park: Dynamically adjust traffic light timings during peak commuting hours based on shift data (such as morning / evening shifts) to reduce employee commuting waiting time.

9. The system according to claim 1, characterized in that, The system reduces deployment costs through the following technologies: Hardware reuse: Cameras and geomagnetic detectors are installed on existing streetlights in the factory area, reducing the need for independent pole construction; Software-defined traffic: A central control platform manages traffic signals of different brands and models in a unified manner, avoiding equipment replacement costs; Cloud-edge collaboration: Migrate non-real-time computing tasks to the cloud and reduce the hardware configuration requirements of edge computing units.

10. The system according to claim 1, characterized in that, The system enhances user experience through the following technologies: Mobile-friendly design: The mobile law enforcement app interface adopts Material Design guidelines and supports dark mode and voice input; Accessibility features: The guidance screen supports voice broadcasting, making it easier for visually impaired people to obtain information; Multilingual support: The central control platform and mobile law enforcement app provide bilingual interfaces in Chinese and English to meet the needs of foreign employees.

Images