A safety cleaning system and method for a first pass of a beneficiation belt

Abstract

The application discloses a beneficiation belt first round safety cleaning system and method, and belongs to the technical field of beneficiation intelligent safety protection. The system is composed of a multi-source sensing component, a vehicle-mounted control interaction unit, a cleaning platform sound and light emergency component, a cloud data platform and a dual-path uninterrupted power supply module, and fuses triple sensing signals of millimeter wave radar ranging, AI visual recognition and UWB personnel positioning. The controller weighted fusion algorithm is used to accurately identify the cleaning operation personnel and the risk of cross collision of the grab bucket. A four-level grading linkage protection mechanism of early warning, danger, personnel anomaly and emergency is set. When the distance meets the standard, sound and light reminders, driving limits, grab lock and whole machine shutdown are executed in sequence. A supporting on-site emergency call button realizes one-key forced shutdown of personnel in distress. The cloud platform completely stores operation and early warning data for safety review, and the standby power module ensures the continuous and effective protection function under power failure conditions. The application solves the defects of traditional single sensing protection, such as false triggering, missed identification, no emergency protection, no data traceability and power failure, etc. While ensuring the personal safety of cleaning personnel and eliminating driving collision accidents, the application reduces the interference of the protection mechanism on the continuous production of beneficiation, and the system is modularized and easy to transform, and is suitable for various beneficiation workshops with cross operation of driving and manual cleaning, and promotes the digital upgrading of plant safety management.

Description

Technical Field

[0001] This invention belongs to the field of intelligent safety protection technology in mineral processing production, specifically involving a multi-source sensing hierarchical linkage safety cleaning system and supporting safety protection control method for the first round of ore accumulation cleaning operation on conveyor belts in mineral processing plants, which can prevent collisions between grab bucket cranes and cleaning personnel. Background Technology

[0002] Iron concentrate easily accumulates and spills at the first wheel of the belt conveyor in the ore dressing plant, requiring operators to periodically climb the cleaning platform to remove the accumulated ore. Due to the limited height of the plant space, the trajectory of the grab crane overlaps with the cleaning area of ​​the first wheel of the belt conveyor, greatly increasing the risk of collisions between personnel and the crane / grab, posing a significant safety hazard of personal injury or death. Existing safety measures have obvious shortcomings.

[0003] 1. Passive and singular warning protection leads to serious misjudgments and omissions: Relying solely on fixed warning signs and ordinary infrared sensors cannot distinguish between personnel, ore, equipment, and debris, easily triggering unwarranted shutdowns and causing production stoppages, or failing to identify workers, creating blind spots in protection; relying solely on fixed areas with blanket prohibitions cannot balance safety and continuous production.

[0004] 2. Lack of personnel status monitoring and proactive emergency response mechanism: The existing protection system only passively monitors approaching vehicles and cannot identify abnormal states such as personnel slipping, being trapped, or remaining stationary for extended periods; there is no one-button emergency call device on site, so personnel in distress cannot actively trigger the highest level of shutdown protection, resulting in a delayed rescue response.

[0005] 3. Lack of closed-loop linkage in control logic: The sensing, judgment, early warning, and equipment braking processes are independent and isolated, lacking a unified controller to integrate multi-dimensional signals. Early warning, braking, and audible / visual alarms cannot be synchronized, resulting in delayed protection response.

[0006] 4. Lack of operational data traceability: The system cannot store operational periods, vehicle trajectories, warning records, or equipment anomaly data. Safety management relies solely on manual ledgers, making it impossible to identify high-frequency risk points, optimize operational processes, and provide data-driven support for safety control.

[0007] 5. Insufficient power supply reliability: Conventional protection systems rely solely on the plant's main power supply. After a power outage, the protection function completely fails, and the entire cleaning operation loses its safety barrier.

[0008] In summary, existing technologies lack a closed-loop intelligent safety cleaning and protection solution that integrates multi-source fusion sensing, hierarchical dynamic protection, personnel anomaly monitoring, one-click emergency shutdown, cloud data traceability, and uninterrupted backup power supply, making it difficult to meet the safety management and control needs of complex cross-operation scenarios in ore dressing plants. Summary of the Invention

[0009] To address the shortcomings of existing conveyor belt cleaning operations, such as high false trigger rate, low risk identification accuracy, lack of personnel anomaly monitoring, inadequate emergency support, lack of data traceability, and power outage protection failure, this paper proposes a safety cleaning system and method for the first stage of conveyor belt cleaning. It achieves accurate risk identification through multi-source perception fusion algorithms using millimeter-wave radar, AI vision cameras, and UWB personnel positioning, establishing a distance-level early warning and braking protection mechanism. It also includes functions for identifying personnel stationary anomalies and forcing shutdown via on-site emergency calls. A cloud-based data management platform enables full-process data storage and review, and a backup power module ensures continuous protection during power outages. This forms a complete closed-loop control system encompassing multi-source perception, intelligent analysis, graded protection, emergency response, and data traceability. While eliminating the risk of collisions between personnel and the grab crane, it minimizes the interference of the protection mechanism on continuous mineral processing production, thereby improving the level of intelligent safety management in the plant area.

[0010] To achieve the above objectives, the technical solution of the present invention is as follows: A safety cleaning system for the first wheel of a mineral processing conveyor belt, comprising plant foundation equipment, multi-source sensing components, control and interaction components, on-site audible and visual warning components, emergency operation components, cloud management components, and uninterruptible power supply components; the plant foundation equipment includes a grab bucket trolley, a grab bucket, the first wheel of the conveyor belt, and a cleaning platform; the cleaning platform is located below the first wheel of the conveyor belt and serves as the work station for cleaning accumulated ore; the grab bucket trolley carries the grab bucket and moves along the plant track, with its running path intersecting the working area of ​​the cleaning platform; the multi-source sensing components include a millimeter-wave radar sensor, a high-definition AI camera, a UWB personnel positioning tag, an area calibration module, and a signal receiver; the area calibration module, millimeter-wave radar sensor, high-definition AI camera, UWB personnel positioning tag, and signal receiver work together to construct a multi-dimensional sensing network; the control and interaction components are a display screen and a controller, with built-in multi-source data fusion algorithms and hierarchical protection logic programs, which are communicatively connected to the multi-source sensing components, the grab bucket trolley electronic control system, the audible and visual warning components, the emergency operation components, and the cloud management components; The on-site audible and visual warning components include a driver's cab warning device and a cleaning platform audible and visual warning device; the emergency operation component integrates a switch box and an emergency call button, which are uniformly installed on the side wall of the cleaning platform; the switch box is equipped with a main and backup power automatic switching module; the cloud management component is a cloud data management platform that communicates with the controller via industrial Ethernet; the uninterruptible power supply component is a backup power module that is connected in parallel to all electrical equipment in the system, and works with the switch box power switching module to achieve uninterrupted power supply switching in the event of a main power failure.

[0011] Installation and signal interaction of various components: A high-definition AI camera, signal receiver, and millimeter-wave radar sensor are installed on the exterior of the grab bucket trolley cab; a display screen, controller, and cab warning device are fixed inside the cab; from top to bottom, the side wall of the cleaning platform is equipped with an area calibration module, audible and visual warning device, switch box, and emergency call button; operators wear UWB personnel positioning tags, which, together with the area calibration module and signal receiver, form a UWB positioning network; the backup power module, the switch box power switching module, and the main power supply of the plant are electrically connected to provide dual power supply for the main control, sensing, warning, and emergency equipment; the controller integrates millimeter-wave radar distance and speed data, AI camera image recognition data, and UWB personnel positioning coordinate data to output graded warnings, vehicle braking, grab bucket locking, cloud push, and audible and visual alarm control commands.

[0012] System core hierarchical control logic

[0013] (1) Multi-source perception layer: Millimeter-wave radar collects the distance and speed between the grab bucket trolley and the cleaning platform in real time; high-definition AI camera equipped with image recognition algorithm distinguishes personnel, ore, and foreign objects from equipment; UWB positioning tag uploads the coordinates of the workers in real time, and the area calibration module delineates the cleaning operation boundary; the three types of perception data are transmitted to the controller synchronously. (2) Intelligent judgment layer: The controller uses a weighted fusion algorithm to analyze multi-source signals and divides them into three types of working conditions: ① Normal cleaning operation: switch box closed + personnel positioning signal falls into the calibration area; ② Graded risk working condition: the distance between the trolley and the cleaning area is ≤8m and the warning state is determined, ≤5m and the dangerous state is determined; ③ Abnormal emergency working condition: the personnel are stationary in the working area for more than 30s, or the emergency call button is pressed and the highest level of emergency state is determined. (3) Hierarchical dynamic protection layer: Warning state: The low-frequency sound and light alarm of the cab warning device and the real-time display of the driving distance and personnel position are displayed on the screen; only prompting and not restricting the driving operation; Dangerous state: The controller outputs the command to prohibit right movement to the grab bucket crane, and locks the lifting and opening and closing actions of the grab bucket to prevent the crane from approaching the work area; Abnormal state: The high-frequency sound and light alarm of the cleaning platform warning device and the controller pushes abnormal warning information to the cloud; Emergency state: The grab bucket crane is forced to stop, the high-frequency sound and light alarm of the cab and platform are activated, and the cloud pushes an emergency notification with on-site video to the mobile terminal of the management personnel. (4) Emergency linkage layer: After the emergency call button is triggered, the system immediately implements the highest level of protection; when the main power is cut off, the switch box switching module switches to the backup power within 10ms to ensure the continuous operation of all protection functions. (5) Data traceability layer: The cloud data management platform stores the start and stop time of operation, personnel information, driving trajectory, warning records at all levels, and equipment operating parameters in real time, and supports historical query, risk statistics, and safety report generation.

[0014] A safety cleaning control method based on the above-mentioned safety cleaning system at the first stage of the mineral processing conveyor belt includes the following steps:

[0015] Step 1: System self-test standby: After the device is started, the controller automatically checks the millimeter-wave radar, AI camera, positioning network, communication link, and main and backup power supply. If there are no faults in the self-test, it enters standby mode.

[0016] Step 2, Personnel preparation before operation: Cleanup personnel wear UWB personnel positioning tags and confirm that the emergency call button and audible and visual warning devices are in good working order;

[0017] Step 3: Activate the cleaning operation mode: Personnel board the cleaning platform and close the switch box switch. The system starts the operation mode, and the calibration module, radar, camera, and positioning tag are activated simultaneously. Operation data recording is started in the cloud.

[0018] Step 4, Tiered Dynamic Risk Protection: During vehicle movement, the radar continuously measures the distance. If the distance is ≤8m, a low-frequency warning is triggered in the cab; if the distance is ≤5m, the rightward movement of the vehicle and the grab bucket action are locked.

[0019] Step 5, Automatic Anomaly Detection: The controller monitors the personnel coordinates in real time. If a person remains stationary in the work area for more than 30 seconds, the platform will automatically trigger a high-frequency audible and visual alarm and upload an alert to the cloud.

[0020] Step 6, Emergency Response: When personnel slip or become trapped, press the emergency call button. The system will control the grab bucket crane to stop, trigger a dual-point audible and visual high-frequency alarm, and push emergency notifications and on-site videos to the cloud.

[0021] Step 7: Operation completed and protection removed: After cleaning, disconnect the switch box knife, the system exits the operation mode, the restrictions on the movement of the crane and grab bucket are lifted, and the radar and camera go into sleep mode; all data of this operation is archived in the cloud.

[0022] Step 8: Post-event data review: Managers log in to the cloud data management platform, retrieve historical operation, early warning, and equipment operation data, and analyze high-frequency risk points to optimize operation control processes.

[0023] Sensing equipment parameters: High-definition AI camera 1080P industrial infrared type, personnel recognition accuracy ≥99%; 24GHz millimeter wave radar detection distance 0.1~10m, ranging accuracy ±5cm; UWB positioning tag positioning accuracy ≤10cm, battery life ≥72h;

[0024] Controller hardware: 10.1-inch industrial touchscreen, equipped with an embedded industrial control motherboard, built-in Linux system, configured with multi-source data weighted fusion algorithm, with visual weight 30%, radar weight 40%, and UWB positioning weight 30%;

[0025] Power supply configuration: The backup power supply uses a 12V / 100Ah lithium battery, which can support the system to run continuously for 4 hours after a power outage; the switching time between the main and backup power supplies is ≤10ms.

[0026] Communication architecture: Industrial Ethernet transmits high-definition video and big data, while wireless transmission transmits signals from positioning tags and emergency buttons, with dual links ensuring stable communication;

[0027] Protection threshold parameters: warning distance 8m, dangerous braking distance 5m, personnel static abnormal judgment threshold 30s, which can be calibrated and modified on-site according to the factory site dimensions.

[0028] Compared with the prior art, the present invention has the following advantages:

[0029] 1) Multi-source fusion perception significantly reduces false alarms and missed detections: The fusion of radar ranging, AI visual recognition, and UWB precise positioning allows for the correction of defects identified by a single device by the other two types of signals, accurately distinguishing between personnel, ore, and debris, avoiding unnecessary downtime or blind spots, and balancing safety and production continuity.

[0030] 2) Tiered flexible protection, balancing safety and production capacity: Abandoning the one-size-fits-all approach of complete traffic restrictions, a two-tiered protection system is implemented: early warning and emergency braking. During the early warning phase, only audible and visual alerts are provided without restricting vehicle movement; equipment is only locked when a dangerous distance is reached. This minimizes the impact of the protection mechanism on the transportation of mineral processing materials.

[0031] 3) Dual protection of personnel status monitoring and proactive emergency response: Real-time monitoring of personnel remaining stationary for extended periods and automatic early warning; a one-button emergency call system is installed on-site, which can force the entire machine to shut down in case of an emergency, achieving a dual safety barrier of passive risk interception and proactive rescue, and solving the hidden dangers of personnel being trapped or slipping unnoticed.

[0032] 4) Cloud-based data traceability throughout the entire process enables digital safety management: Complete retention of operation, early warning, and equipment operation data; automatic statistics of high-frequency risk points; upgrading safety management from manual experience-based control to data-driven intelligent control; facilitating system optimization and accident tracing.

[0033] 5) Uninterrupted dual power supply design, seamless protection: Automatically switches to backup lithium battery in the event of a main power failure. Warning, braking, alarm, and positioning functions all function normally during power outages, eliminating the risk of protection failure due to power outages.

[0034] 6) Modular retrofitting, adaptable to existing ore dressing plants: All sensing, control, and warning components are installed using bolted modular installation, eliminating the need for large-scale modifications to existing grab bucket cranes and cleaning platform equipment. The retrofitting process is simple, adaptable to various first-round cleaning scenarios for similar ore dressing conveyor belts, and has strong versatility for widespread adoption.

[0035] 7) More precise and reliable safety protection: Through multi-source fusion sensing and intelligent judgment, the system effectively avoids misjudgment and missed judgment, accurately identifies safety risks and implements graded protection, greatly reducing the safety hazards of collisions between operators and the grab crane, and ensuring the personal safety of workers.

[0036] 8) Balancing safety and production efficiency: The tiered dynamic protection mechanism avoids unnecessary production stoppages, minimizing the impact on the normal operation of the grab crane while ensuring safety, thus improving the overall efficiency of ore beneficiation production.

[0037] 9) Significantly enhanced emergency response capabilities: The integration of personnel status monitoring and emergency call functions enables early detection and warning of abnormal situations and rapid response to emergencies, providing dual safety guarantees for workers.

[0038] 10) Facilitating Intelligent Upgrades in Safety Management: The cloud-based data management platform enables the traceability and analysis of safety data, facilitating managers to accurately identify high-frequency risk points, optimize work processes, and improve the standardization and intelligence of safety management in mineral processing production.

[0039] 11) High system reliability and strong adaptability: The uninterrupted power supply design improves the system's reliability in complex production environments; the system structure is simple, and new components are compatible with existing production equipment, making it easy to upgrade and modify in existing mineral processing environments, with a wide range of applications. Attached Figure Description

[0040] Figure 1. Schematic diagram of the overall system structure layout.

[0041] Figure 2. Schematic diagram of the installation of the on-board components of the grab bucket crane.

[0042] Figure 3. Schematic diagram of the equipment installation on the side wall of the cleaning platform.

[0043] Figure 4. System control logic principle block diagram.

[0044] The diagram shows: 1. Grab bucket crane, 2. Cleaning platform, 3. Belt pulley, 4. Cab warning device, 5. Display screen and controller, 6. High-definition AI camera, 7. Signal receiver, 8. Grab bucket, 9. Switch box, 10. Audible and visual warning device, 11. Calibration module, 12. Millimeter-wave radar sensor, 13. UWB personnel positioning tag, 14. Emergency call button, 15. Cloud data management platform, 16. Backup power module. Detailed Implementation

[0045] To enhance understanding of the present invention, the embodiments will be described in detail below with reference to the accompanying drawings.

[0046] Example 1: This invention provides a safety cleaning system and method for the first pulley of a mineral processing conveyor belt, combined with... Figures 1 to 4 As can be seen, the system includes basic components and newly added intelligent components, specifically: 1. Grab bucket trolley; 2. Cleaning platform; 3. Belt pulley; 4. Warning device; 5. Display screen and controller; 6. High-definition camera; 7. Receiver; 8. Grab bucket; 9. Switch box; 10. Audible and visual warning device; 11. Calibration module; 12. Millimeter-wave radar sensor; 13. Personnel positioning tag; 14. Emergency call button; 15. Cloud data management platform; 16. Backup power module.

[0047] The installation and connection relationships of each component are as follows:

[0048] Basic component installation: The grab bucket trolley 1 is installed inside the plant; the cleaning platform 2 is installed to the lower left of the belt pulley 3; the warning device 4 is fixed to the top of the grab bucket trolley cab with bolts; the display screen and controller 5 are fixed to the upper right of the grab bucket trolley cab with bolts, so that the operator can view the video and data; the high-definition camera 6 is installed at the bottom right outside the grab bucket trolley cab to monitor the status of the cleaning platform 2; the receiver 7 is installed at the top right outside the grab bucket trolley cab to receive signals from the calibration module 11, the millimeter-wave radar sensor 12 and the personnel positioning tag 13; the left wall of the cleaning platform 2 is equipped with the calibration module 11, the audible and visual warning device 10, the switch box 9 and the emergency call button 14 from top to bottom.

[0049] New intelligent component installations: Millimeter-wave radar sensor 12 is installed on the lower right edge of the grab bucket trolley, and its detection range covers the entire cleaning operation area and the necessary path for the grab bucket trolley to approach the area; personnel positioning tags 13 are worn by the operators and adopt UWB ultra-wideband positioning technology to form a positioning network with calibration module 11 and receiver 7; cloud data management platform 15 establishes a communication connection with display screen and controller 5, calibration module 11 and millimeter-wave radar sensor 12 through industrial Ethernet; backup power module 16 is connected to each power component of the system to provide uninterrupted power supply to the system; the switch box 9 integrates a power switching module to realize automatic switching between main power and backup power.

[0050] The core working logic of the system is "multi-source perception - intelligent judgment - dynamic protection - emergency response - data traceability", and the specific implementation method is as follows:

[0051] (1) Multi-source perception layer: Multi-dimensional data collection is achieved through visual perception by high-definition camera 6, distance and motion state perception by millimeter-wave radar sensor 12, precise positioning of workers by personnel positioning tag 13, and area boundary calibration by calibration module 11. It can simultaneously identify the position of workers, the operating status and speed of grab bucket crane, distance, whether there are foreign objects in the cleaning area, and the activity trajectory of workers. Among them, high-definition camera 6 is equipped with AI image recognition algorithm, which can accurately distinguish between personnel, equipment and irrelevant objects, avoiding misjudgment by a single perception method.

[0052] (2) Intelligent judgment layer: The display screen and controller 5 have a built-in core control algorithm to perform fusion analysis on the data collected by the multi-source perception layer and judge whether there is a safety risk in the current state: ① When the operator enters the cleaning area, the personnel positioning tag 13 signal enters the area defined by the calibration module 11 and the switch box 9 is turned on, it is judged as "normal cleaning operation state"; ② When the grab bucket crane approaches the cleaning area, the distance is ≤8 meters, it is judged as "warning state" and the distance is ≤5 meters, it is judged as "danger state"; ③ When the switch box 9 is not turned on but the personnel positioning tag 13 signal enters the cleaning area, or the operator stays still in the area for more than a preset threshold such as 30 seconds, it is judged as "abnormal state".

[0053] (3) Dynamic protection layer: Based on the results of the intelligent judgment layer, graded protection measures are implemented: ① Warning state: The warning device 4 in the grab bucket crane cab emits a low-frequency audible and visual alarm, and the display screen and controller 5 display the real-time distance between the grab bucket crane and the cleaning area and the position of the workers; ② Danger state: The control system automatically outputs a prohibition command, the grab bucket crane is prohibited from moving to the right, and the action of the grab bucket 8 is locked until the risk is eliminated; ③ Abnormal state: The audible and visual warning device 10 emits a high-frequency audible and visual alarm, the display screen and controller 5 push abnormal prompts, and at the same time send warning information to the cloud data management platform 15.

[0054] (4) Emergency linkage layer: If the operator encounters an emergency such as slipping or being trapped, he / she can press the emergency call button 14 on the cleaning platform 2. The system will immediately trigger the highest level of warning: the grab bucket crane will be forced to stop, the sound and light alarm 10 and the alarm 4 will simultaneously sound a high frequency alarm, the display screen and controller 5 will display the location of the emergency call, and the cloud data management platform 15 will push the emergency notification to the mobile terminal of the management personnel; the backup power module 16 will be automatically put into use when the main power is interrupted to ensure that the emergency protection measures continue to be effective.

[0055] (5) Data traceability layer: The cloud data management platform 15 stores data such as the entry / exit time of operators, the running trajectory and early warning records of grab bucket cranes, and the operating status of system equipment in real time. It supports historical data query, risk point statistics and report generation, providing data support for the safety management of mineral processing production.

[0056] The specific execution process of the cleanup operation:

[0057] 1. Preparation before operation: Operators should wear personnel positioning tags 13 and check whether the sound and light warning devices 10, millimeter-wave radar sensors 12, and emergency call buttons 14 of the system are in normal working order;

[0058] 2. Start the operation mode: After the operator arrives at the cleaning platform 2, he closes the switch box 9 and the system automatically activates the "cleaning operation mode". The calibration module 11, the sound and light alarm 10, and the personnel positioning tag 13 are started at the same time, and the cloud data management platform 15 begins to record the operation data.

[0059] 3. Dynamic protection execution: During the operation of the grab bucket crane 1, the receiver 7 receives signals from each sensing component in real time, and the display screen and controller 5 display real-time data; the millimeter-wave radar sensor 12 senses the current distance in real time. When the grab bucket crane is ≤8 meters away from the cleaning area, the warning device 4 alarms at a low frequency to remind the crane operator to pay attention; when the distance is ≤5 meters, the grab bucket crane is automatically prohibited from moving to the right, and the grab bucket 8 is locked.

[0060] 4. Abnormal and Emergency Handling: If the operator remains stationary in the area for more than 30 seconds, the system will determine it as an abnormality, triggering a high-frequency alarm on the audible and visual warning device 10 and sending a warning to the cloud. In case of an emergency, the operator can press the emergency call button 14, which will trigger a forced shutdown and a system-wide alarm, notifying management personnel for handling.

[0061] 5. Operation completed: After the cleaning operation is completed, the operator will disconnect the switch box 9, the system will turn off the "cleaning operation mode", the restriction on the grab bucket crane will be lifted, and the grab bucket crane will return to normal operation; the cloud data management platform 15 will complete the archiving of the data of this operation.

[0062] System components:

[0063] 1. High-definition camera 6: It uses an industrial-grade infrared camera with a resolution of 1920×1080, supports shooting in low-light environments, and is equipped with Hikvision's AI image recognition algorithm, which can achieve accurate identification of people and equipment with an accuracy rate of ≥99%;

[0064] 2. Millimeter-wave radar sensor 12: Selects a 24GHz millimeter-wave radar with a detection range of 0.1-10 meters and a detection angle of ±60°, with a ranging accuracy of ≤±5cm, and can output the target's distance and speed information in real time;

[0065] 3. Personnel positioning tag 13: UWB positioning tag is selected, with positioning accuracy ≤10cm, communication distance ≥50m, battery life ≥72 hours, and low battery alarm is supported;

[0066] 4. Display and Controller 5: 10.1-inch industrial-grade touchscreen, equipped with an Intel Celeron J1900 processor, built-in Linux operating system, supporting multi-source data fusion analysis and command output;

[0067] 5. Cloud-based data management platform 15: Adopting Alibaba Cloud Industrial Internet platform, it supports real-time data uploading, storage, querying and report generation, and can be connected to mobile APP to realize remote early warning notification;

[0068] 6. Backup power module 16: Uses a 12V / 100Ah lithium battery pack, equipped with a charging management module, which can support the system to work continuously for ≥4 hours.

[0069] Specific technical solutions

[0070] 1. Multi-source data fusion algorithm: The core algorithm built into the display screen and controller 5 adopts a weighted average fusion strategy to fuse visual data from high-definition cameras, distance data from millimeter-wave radar, and location data from UWB positioning. The weight allocation is: visual data 30%, radar data 40%, and positioning data 30%, ensuring accurate risk identification even in complex environments such as dust and insufficient light.

[0071] 2. Graded protection parameter settings: The warning distance is set to 8 meters, which is the safe braking distance of the grab bucket vehicle, ensuring that the vehicle has enough time to decelerate after receiving the warning; the danger distance is set to 5 meters, which is the safe boundary of the clearing area, ensuring that the grab bucket cannot reach the workers; the abnormal stillness threshold is set to 30 seconds, which can effectively distinguish between normal work stoppage and accidental entrapment of workers.

[0072] 3. Communication Link Design: The system adopts a dual-link design of "industrial Ethernet + wireless communication". High-definition video and large data transmission are realized through industrial Ethernet, while the signals of personnel positioning tags and emergency call buttons are transmitted through wireless communication to ensure the stability and real-time performance of data transmission.

[0073] 4. Power supply protection design: The switch box 9 integrates a power monitoring module to monitor the main power supply voltage status in real time. When the main power supply voltage is lower than the preset threshold such as 220V±10% or when the power is cut off, the power switching module automatically switches to the backup power supply module 16 for power supply. The switching time is ≤10ms, ensuring uninterrupted operation of the system.

[0074] 5. Installation and commissioning: All components of the system adopt a modular design and are fixed with bolts, making installation convenient. During the commissioning phase, the cleaning area boundary is delineated through calibration module 11, the positioning tag information of the operators is entered, and the graded protection parameters are calibrated on-site to ensure that they meet the actual production environment requirements.

[0075] Specific Work Process

[0076] 1. System initialization: Before production, the management personnel start the system through the cloud data management platform 15. The system automatically completes the self-test of each component, including the detection of the high-definition camera 6, millimeter-wave radar sensor 12, personnel positioning tag 13, communication link and power status. After the self-test is passed, it enters the standby state.

[0077] 2. Personnel preparation: Personnel receive and wear personnel positioning tags 13, confirm tag activation via mobile APP, and check that emergency call button 14 is working properly;

[0078] 3. Start the operation mode: After the operator arrives at the cleaning platform 2, close the power switch in the switch box 9. After the system receives the start signal, it automatically activates the cleaning operation mode: the calibration module 11 outputs the area boundary signal, the sound and light alarm 10 emits a low-frequency prompt sound to indicate that the operation has started, the millimeter-wave radar sensor 12 starts detection, the high-definition camera 6 starts shooting and performing AI recognition, and the personnel positioning tag 13 sends the location information in real time.

[0079] 4. Grab bucket crane linkage protection: When the grab bucket crane 1 is running, the receiver 7 receives signals from each sensing component in real time and transmits them to the display screen and controller 5; when the grab bucket crane moves towards the cleaning area and the distance reaches the 8-meter warning threshold, the display screen and controller 5 control the warning device 4 to issue a low-frequency audible and visual alarm, and the display screen displays the prompt "Approaching the work area, please pay attention to avoid" and the real-time distance; when the distance drops to the 5-meter danger threshold, the display screen and controller 5 output a command to prohibit right movement to the grab bucket crane control system, and at the same time lock the lifting and opening and closing actions of the grab bucket 8, the grab bucket crane stops approaching, and the display screen displays "No passage, people in the work area";

[0080] 5. Abnormal Status Handling: If the worker remains stationary in the cleaning area for more than 30 seconds, the system will determine it as abnormal. The audible and visual alarm 10 will switch to a high-frequency audible and visual alarm, and the display screen and controller 5 will display "Worker's status is abnormal". At the same time, the cloud data management platform 15 will push the warning information to the mobile terminal of the management personnel. The management personnel can view the on-site situation through the high-definition camera. If the worker is accidentally trapped, the grab bucket crane can be remotely triggered to stop and rescue can be arranged.

[0081] 6. Emergency Response: If workers encounter an emergency such as slipping or being buried by ore, they can press the emergency call button 14. The system will immediately trigger the highest level of emergency response: the audible and visual alarms 10 and 4 will simultaneously issue high-frequency audible and visual alarms, the grab crane will be forced to stop, the display screen and controller 5 will display the location and personnel information of the emergency call, and the cloud data management platform 15 will push an emergency notification containing a link to the on-site video to the mobile terminal of the management personnel. The management personnel can view the on-site situation in real time and organize rescue efforts.

[0082] 7. Operation Completion: After the cleanup operation is completed, the operators check that there are no residual mineral deposits or safety hazards in the cleaned area. Then, they disconnect the power switch in switch box 9. After the system receives the stop signal, it shuts down the cleanup operation mode, the audible and visual alarm 10 stops alarming, the millimeter-wave radar sensor 12 and the high-definition camera 6 stop working, the restrictions on the grab crane are lifted, and the grab crane resumes normal operation. The cloud data management platform 15 automatically archives the data of this operation, including the start / end time of the operation, operator information, grab crane warning records, system operating status, etc.

[0083] 8. Data traceability and analysis: Managers can query historical operation data through the cloud data management platform, generate safety operation reports, analyze high-frequency risk points such as the frequency of grab bucket cranes approaching the work area and the number of abnormal warnings, and optimize operation processes and protection parameters.

[0084] Specific Implementation Example 1: System Hardware Installation and Wiring

[0085] 1. Onboard component assembly of the crane: A display screen and controller 5 are fixed on the upper right side of the grab crane cab, and a warning device 4 is installed on the top of the cab; a millimeter-wave radar sensor 12 and a high-definition AI camera 6 are installed on the lower right side of the exterior of the cab, and a signal receiver 7 is installed in the upper right corner; all onboard equipment cables are uniformly connected to the controller, and the controller communication port is connected to the industrial Ethernet of the plant area.

[0086] 2. Cleaning Platform Side Wall Equipment Installation: On the left side wall of the cleaning platform 2, from top to bottom, the area calibration module 11, the audible and visual alarm 10, the switch box 9, and the emergency call button 14 are fixed in sequence; the switch box integrates a main and backup power automatic switching module, which is connected to the plant's 220V main power supply and backup power supply module 16 respectively; the calibration module, the audible and visual alarm, and the emergency call button interact with the vehicle receiver 7 via wireless signal.

[0087] 3. Personnel Support Equipment: Operators are equipped with UWB personnel positioning tags 13. The tags have a built-in wireless transmission module that continuously uploads real-time coordinates to the calibration module and signal receiver.

[0088] 4. Cloud Deployment: The cloud data management platform 15 is deployed on an industrial cloud server and establishes a long connection with the controller via Ethernet to receive sensing data, early warning records, and operation logs in real time.

[0089] Specific Implementation Example 2: System Parameter Debugging and Algorithm Configuration

[0090] 1. Area calibration: The operation boundary of the cleaning platform is defined by calibration module 11 and entered into the controller for storage, serving as the benchmark for UWB positioning risk assessment;

[0091] 2. Threshold setting: The controller has built-in preset parameters for warning distance of 8m, dangerous braking distance of 5m, and abnormal personnel stillness duration of 30s, which can be modified on-site via touch screen;

[0092] 3. Fusion Algorithm Configuration: The controller has a built-in weighted fusion algorithm with radar ranging data weighting at 40%, AI visual recognition weighting at 30%, and UWB positioning coordinate weighting at 30%. Multi-signal cross-verification eliminates interference errors from a single device.

[0093] 4. Power switching test: Disconnect the main power supply to simulate a power outage scenario, verify that the switch box switching module switches to the backup lithium battery within 10ms, and all equipment can work continuously for more than 4 hours.

[0094] Specific Implementation Example 3: On-site Standardized Operation Control Process

[0095] 1. Power-on self-test: Upon system startup, the controller automatically completes self-tests of the radar, camera, positioning network, power supply, and communication. If no faults are found, it enters standby mode.

[0096] 2. Operation Activation: Personnel wearing positioning tags board the cleaning platform, close the switch box switch, the system starts the cleaning operation mode, and data recording is initiated in the cloud;

[0097] 3. Tiered protection operation: As the grab bucket trolley moves towards the cleaning area, the radar measures the distance in real time; at a distance of 8m, a low-frequency alarm in the cab reminds the driver to slow down; at a distance of 5m, the controller outputs a locking command, prohibiting the trolley from moving to the right and rendering the grab bucket unable to move;

[0098] 4. Abnormal identification and handling: If personnel remain stationary for more than 30 seconds, the platform's audible and visual alarm will sound at high frequency, a warning message will be pushed to the cloud, and management personnel can remotely retrieve camera footage to confirm the situation on site;

[0099] 5. Emergency rescue trigger: When personnel press the emergency call button in distress, the entire grab bucket crane is forced to stop, the cab and platform simultaneously trigger high-frequency audible and visual alarms, and the management personnel receive an emergency notification with a video link on their mobile phones;

[0100] 6. Reset upon completion of operation: After cleaning, disconnect the switch box knife, exit the operation mode, release the restrictions on the movement of the crane and grab bucket, and put the radar and camera into sleep mode; all operation data will be automatically archived in the cloud.

[0101] 7. Data review: Safety management personnel log in to the cloud platform to query operation duration, number of warnings, and frequency of vehicle approach, and analyze high-risk areas to optimize operation scheduling and protection parameters.

[0102] It should be noted that the above embodiments are not intended to limit the scope of protection of the present invention. Equivalent transformations or substitutions made based on the above technical solutions all fall within the scope of protection of the claims of the present invention.

Claims

1. A safety cleaning system for the first pulley of a mineral processing conveyor belt, characterized in that, The system includes basic plant equipment, multi-source sensing components, control and interaction components, on-site audio-visual warning components, emergency operation components, cloud management components, and uninterruptible power supply components. The basic plant equipment includes a grab bucket crane, a grab bucket, a conveyor belt pulley, and a cleaning platform. The cleaning platform is located below the conveyor belt pulley as a ore accumulation cleaning station. The grab bucket crane carries the grab bucket and its operating path intersects with the cleaning platform's work area. The multi-source sensing components include millimeter-wave radar sensors, high-definition AI cameras, UWB personnel positioning tags, area calibration modules, and signal receivers. The millimeter-wave radar sensors, high-definition AI cameras, and signal receivers are installed on the exterior of the grab bucket crane, while the area calibration module is installed on the side wall of the cleaning platform. The UWB... Personnel positioning tags are worn by operators to jointly construct a multi-dimensional perception network; the control and interaction components are displays and controllers, with built-in multi-source data fusion algorithms and hierarchical protection logic programs, communicating with multi-source perception components, grab bucket trolley electronic control system, audible and visual warning components, emergency operation components, and cloud management components respectively; the on-site audible and visual warning components include a warning device installed in the grab bucket trolley cab and an audible and visual warning device installed on the side wall of the cleaning platform; the emergency operation components include a switch box and an emergency call button, both fixed to the side wall of the cleaning platform, with the switch box integrating a main and backup power automatic switching module; the cloud management component is a cloud data management platform that establishes data interaction with the displays and controllers via industrial Ethernet; the uninterruptible power supply component is a backup power module connected in parallel to all electrical equipment in the system, working with the switch box power switching module to achieve uninterrupted power switching when the main power supply fails; the displays and controllers integrate distance and speed data from millimeter-wave radar, high-definition AI camera image recognition data, and UWB... The personnel positioning coordinate data is used for intelligent analysis to classify the working conditions into four levels: early warning, danger, abnormality and emergency. Corresponding outputs include graded audible and visual alarms, grab bucket travel limit braking, grab bucket action locking, forced shutdown of the whole machine, and cloud-based early warning push control commands. The cloud-based data management platform stores all data on the operation of personnel, operation time, vehicle trajectory, early warning at all levels, and equipment operation in real time, and supports historical query, risk statistics, and safety report generation.

2. The safety cleaning system at the first pulley of a mineral processing conveyor belt according to claim 1, characterized in that, The high-definition AI camera is a 1080P industrial infrared camera equipped with a personnel image recognition algorithm, with a personnel recognition accuracy of ≥99%; the millimeter-wave radar is a 24GHz radar with a detection range of 0.1~10m and a ranging accuracy of ≤±5cm; the UWB personnel positioning tag has a positioning accuracy of ≤10cm and a battery life of ≥72h.

3. The safety cleaning system at the first pulley of a mineral processing conveyor belt according to claim 1, characterized in that, The controller has a built-in weighted multi-source data fusion algorithm, with visual recognition data weighted at 30%, millimeter-wave radar ranging data weighted at 40%, and UWB positioning coordinate data weighted at 30%. The system has preset graded protection thresholds: when the distance between the grab bucket and the cleaning platform is ≤8m, it is in a warning state; when the distance is ≤5m, it is in a dangerous braking state; and when personnel remain stationary in the cleaning area for more than 30 seconds, it is considered an abnormal personnel state.

4. The safety cleaning system at the first pulley of a mineral processing conveyor belt according to claim 1, characterized in that, The backup power module uses a 12V / 100Ah lithium battery pack. When the main power fails, the switching time of the power switching module in the switch box is ≤10ms. The backup power can support the continuous operation of the entire system for no less than 4 hours.

5. The safety cleaning system at the first pulley of a mineral processing conveyor belt according to claim 1, characterized in that, The system adopts dual-link communication of industrial Ethernet and wireless. High-definition video and large-volume operation data are transmitted through industrial Ethernet, while signals from UWB positioning tags, emergency call buttons, and area calibration modules are transmitted wirelessly.

6. The safety cleaning system at the first pulley of a mineral processing conveyor belt according to claim 1, characterized in that, The logic for tiered protection is as follows: Warning status: Low-frequency audible and visual alarm in the cab, display screen shows driving distance and personnel position, no restriction on grab bucket operation; Dangerous situation: The controller outputs a command to prohibit the vehicle from moving towards the cleanup area, and simultaneously locks the lifting, lowering, opening and closing actions of the grab bucket; Abnormal personnel status: The high-frequency audible and visual alarm on the cleaning platform triggers an alarm, and the controller pushes an abnormal warning to the cloud data management platform; Emergency call status: The grab bucket crane is forced to stop, and high-frequency audible and visual alarms are triggered simultaneously at both the cab and the cleaning platform. The cloud pushes an emergency notification with on-site video to the mobile terminal of the management personnel.

7. A safety cleaning control method based on the safety cleaning system at the first wheel of the mineral processing conveyor belt as described in any one of claims 1 to 6, characterized in that, Includes the following steps: Step 1: System self-test standby: After the system is powered on, the controller automatically checks the millimeter-wave radar, high-definition AI camera, UWB positioning network, communication link, and main and backup power supply. If there are no faults in the self-test, it enters standby mode. Step 2, Personnel Preparation: Cleanup personnel wear UWB personnel positioning tags and check that the emergency call button and audible and visual warning devices are in good working order; Step 3: Activate the cleaning operation mode: Personnel board the cleaning platform and close the switch box switch. The system starts the cleaning operation mode. The area calibration module, millimeter-wave radar, high-definition AI camera, and UWB positioning tag are activated simultaneously. The cloud data management platform starts recording operation data. Step 4, Tiered Dynamic Risk Protection: During the movement of the grab bucket, the millimeter-wave radar continuously measures the distance. If the distance between the grab bucket and the cleaning platform is ≤8m, a low-frequency warning is triggered in the cab; if the distance is ≤5m, the grab bucket operation and the approach of the grab bucket are locked. Step 5: Automatic monitoring of personnel anomalies: The controller reads the UWB positioning coordinates in real time. If a person remains stationary in the cleaning area for more than 30 seconds, the high-frequency sound and light alarm of the cleaning platform is automatically triggered and an anomaly warning is pushed to the cloud. Step 6, Emergency Response: If personnel encounter danger, press the emergency call button. The system will force the entire grab crane to stop, and the cab and cleaning platform will simultaneously trigger high-frequency audible and visual alarms. The cloud will push an emergency notification with on-site video to the mobile terminal of the management personnel. Step 7, Disarming After Operation: After the ore accumulation is cleared, disconnect the switch box switch. The system will exit the cleaning operation mode, remove the restrictions on the grab crane and grab movement, and put the millimeter-wave radar and high-definition AI camera into sleep mode. The cloud-based data management platform archives all data from this assignment. Step 8: Post-event data review: Managers log in to the cloud data management platform to retrieve historical operation, early warning, and equipment operation data, identify high-frequency risk points, and optimize on-site operation control flow.

Images