A method, robot and system for automatic identification and intelligent firefighting of welding job fires

Abstract

The application discloses a kind of welding operation fire automatic identification and intelligent fire fighting method, robot and system, the method mainly includes by the controller of patrol robot according to the temperature field distribution data t of real-time acquisition welding operation construction site of long-range infrared temperature sensor judges whether welding operation occurs fire: when t≤T1, it is judged that no fire occurs;When t>T2, it is judged that fire occurs;When T1<t≤T2, it is judged that temperature is abnormal, and image data collected in real time is introduced visible light camera to judge multi-modal fusion, when the image of the temperature abnormal area is identified to at least one of smoke and flame by the large model after training, it is judged that fire occurs, otherwise no fire occurs;And when fire occurs, control patrol robot navigates to fire occurrence area and controls fire extinguishing equipment to automatically execute fire extinguishing operation.The application can realize the closed-loop processing of early "identification and extinguishing" of welding operation fire.

Description

Technical Field

[0001] This invention relates to the field of welding technology, and in particular to an automatic fire identification and intelligent firefighting method, robot, and system for welding operations. Background Technology

[0002] Fires caused by welding operations at construction sites are on the rise in construction projects. The high-temperature welding slag and sparks generated during welding can easily ignite surrounding combustibles such as formwork, safety nets, and insulation materials. Due to the complex environment, confined space, and scattered storage of combustibles at construction sites, once a fire breaks out, it spreads rapidly and has serious consequences.

[0003] Conventional fire prevention management primarily relies on manual inspections by patrol personnel. Upon discovering a fire, these personnel must rush to the fire-fighting equipment locations to retrieve handheld fire extinguishers and other firefighting gear before returning to the fire site to extinguish it. Manual inspections suffer from problems such as response delays, limited coverage, and low efficiency during nighttime patrols. Fixed smoke and heat detectors are typically installed inside buildings under construction. While they can trigger alarms after a fire breaks out, in open or semi-open construction environments with frequent welding operations, the rapid spread of smoke and significant background interference often make accurate early warnings difficult in the early stages of a fire (smoldering or localized high-temperature stages). Furthermore, even after an alarm is triggered, it still requires waiting for nearby workers to travel back and forth to retrieve fire-fighting equipment, delaying the fire's spread and potentially allowing it to escalate, ultimately hindering timely fire suppression.

[0004] In fire early warning systems, while existing fixed-point video surveillance systems at construction sites can provide manual monitoring or system-based early warning of fires, their accuracy in identifying early flames and smoke is low under complex lighting conditions at construction sites. Furthermore, they lack intelligent linkage mechanisms with fire extinguishing equipment, failing to achieve a closed-loop response of "early detection and immediate extinguishing" of fires. Existing inspection robots, primarily used in the security field, provide early warnings based on their inspection results, but also lack intelligent linkage mechanisms with fire extinguishing equipment, thus failing to achieve a closed-loop response of "early detection and immediate extinguishing" of fires. Summary of the Invention

[0005] The purpose of this invention is to provide an automatic identification and intelligent fire-fighting method, robot, and system for welding operation fires, in order to solve the problems of existing fire prevention and management methods lacking an intelligent linkage mechanism with fire-fighting equipment, failing to achieve early "identification and extinguishing" closed-loop processing of fires, and the delay caused by manual back-and-forth handling of fire prevention and extinguishing equipment, which leads to the spread of fires.

[0006] To address the aforementioned technical problems, this invention provides an automatic identification and intelligent firefighting method for welding operation fires, which involves equipping an inspection robot with a lidar and a visible light camera with a long-range infrared temperature sensor and a fire extinguishing device connected to its built-in controller.

[0007] The controller uses LiDAR to scan the 3D data of the welding site in real time to build an environmental map and automatically plans the inspection path based on the environmental map to control the inspection robot to autonomously avoid obstacles and navigate for inspection.

[0008] The controller uses a long-range infrared temperature sensor to acquire real-time temperature field distribution data of the welding operation site along the inspection path in a non-contact manner according to a scanning matrix to determine whether a fire has occurred during the welding operation.

[0009] When the temperature field distribution data within the spatial range of a certain scan matrix is ​​not greater than the first threshold, it is determined that the temperature is normal and no fire has occurred.

[0010] When the temperature field distribution data within the spatial location range of a certain scan matrix is ​​greater than the second threshold, it is determined that the temperature is high-risk and a fire has occurred, where the second threshold is greater than the first threshold.

[0011] When the temperature field distribution data within a certain spatial range of a scanning matrix falls between a first threshold and a second threshold, a temperature anomaly is determined. Simultaneously, the controller uses a multimodal fusion method, combining images captured in real-time by a visible light camera with temperature field distribution data acquired in real-time by a long-range infrared temperature sensor, to determine whether a fire has occurred during the welding operation.

[0012] When the controller detects an abnormal temperature through a long-range infrared temperature sensor, and at the same time, the controller identifies smoke and flame in real-time images captured by a visible light camera through a trained large model, and the spatial locations of the two overlap and match, the controller determines that a fire has occurred; otherwise, it determines that no fire has occurred.

[0013] In the event of a fire, the controller directs the inspection robot to navigate to the fire area and controls the fire extinguishing equipment to automatically perform fire extinguishing operations.

[0014] Furthermore, the automatic identification and intelligent fire-fighting method for welding operation fires provided by the present invention, when performing fire-fighting operations, the controller continuously monitors the fire-fighting situation in the fire-occurring area through a long-distance infrared temperature sensor. When the maximum value of the temperature field distribution data in the fire-occurring area drops below the third threshold and remains above the set time period, the fire-fighting operation is determined to be successful; otherwise, the fire-fighting operation is determined to be unsuccessful, wherein the third threshold is less than the first threshold.

[0015] If firefighting operations fail, the controller will instruct the fire extinguishing equipment to execute a secondary firefighting procedure.

[0016] Furthermore, the automatic identification and intelligent fire-fighting method for welding operation fires provided by the present invention also includes an alarm device connected to the controller in the inspection robot;

[0017] When the controller determines that the temperature is high through the long-distance infrared temperature sensor, it identifies it as a high-risk fire alarm and sends out a high-risk fire alarm signal at the construction site through the alarm device.

[0018] When the controller detects an abnormal temperature through a long-range infrared temperature sensor, and at the same time, the controller identifies smoke and flame in real-time images captured by a visible light camera through a trained large model, and the spatial locations of the two overlap and match, it determines that it is a level one fire alarm and issues a level one fire alarm signal at the construction site through the alarm.

[0019] When the controller detects an abnormal temperature through a long-range infrared temperature sensor, and at the same time, the controller identifies smoke and flame in real-time images captured by a visible light camera through a trained large model, but the spatial location ranges of the two do not overlap or match, it determines that it is a level two fire alarm and controls the inspection robot to stop navigation and continue monitoring.

[0020] When the controller determines that the temperature is normal through a long-range infrared temperature sensor, and the controller identifies at least one of smoke and flame in the real-time images captured by the visible light camera through a trained large model, it determines that no fire has occurred and simultaneously determines that it is a level three fire alarm. The controller of the inspection robot records and maintains observation and inspection.

[0021] Furthermore, the automatic identification and intelligent fire-fighting method for welding operation fires provided by the present invention includes a fire-fighting equipment comprising a water tank and a connected water pump and an adjustable water nozzle with horizontal omnidirectional rotation and pitch angle adjustment functions. The water pump and the adjustable water nozzle are connected to a controller, which controls the start and stop of the water pump and, in conjunction with the real-time image signal from a visible light camera, controls the horizontal omnidirectional rotation angle and pitch angle of the adjustable water nozzle to align with the fire area.

[0022] Furthermore, the automatic identification and intelligent fire-fighting method for welding operation fires provided by the present invention further includes a water level sensor installed inside the water tank and extending outside the water tank and connected to the controller.

[0023] To address the aforementioned technical problems, this invention also provides an automatic identification and intelligent firefighting robot for welding operations. The aforementioned automatic identification and intelligent firefighting method for welding operations includes an inspection robot equipped with a lidar and a visible light camera, as well as a long-range infrared temperature sensor and fire extinguishing equipment mounted on it. The controller built into the inspection robot is connected to the lidar, visible light camera, long-range infrared temperature sensor, and fire extinguishing equipment.

[0024] Furthermore, the welding operation fire automatic identification and intelligent fire-fighting robot provided by the present invention includes a skeleton and its walking mechanism, battery, power management circuit and locator. The controller is connected to the battery through the power management circuit and the walking mechanism is a tracked walking mechanism, a wheeled walking mechanism or a quadrupedal walking mechanism.

[0025] To address the aforementioned technical problems, the present invention further provides an automatic identification and intelligent fire suppression system for welding operations, comprising:

[0026] The welding operation fire automatic identification and intelligent fire-fighting robot is the aforementioned welding operation fire automatic identification and intelligent fire-fighting robot, which also includes a wireless communication module connected to its controller;

[0027] The cloud server is wirelessly connected to the welding operation fire automatic identification and intelligent fire-fighting robot via a wireless communication module. It is used to receive and push information on whether a fire has occurred in the welding operation stored in the controller of the welding operation fire automatic identification and intelligent fire-fighting robot.

[0028] The terminal device is connected to the cloud server via a wireless communication module to receive information from the cloud server regarding whether a fire has occurred during the welding operation.

[0029] Furthermore, the automatic identification and intelligent fire suppression system for welding operations provided by this invention...

[0030] When firefighting operations fail, the controller sends a manual firefighting request and pushes it to the terminal device via the cloud server.

[0031] Furthermore, the welding operation fire automatic identification and intelligent fire-fighting system provided by the present invention includes multiple welding operation fire automatic identification and intelligent fire-fighting robots. When a welding operation fire automatic identification and intelligent fire-fighting robot automatically identifies a welding operation fire and controls its fire-fighting equipment to automatically perform fire-fighting operations in the fire area, it sends location information and reinforcement instructions to other welding operation fire automatic identification and intelligent fire-fighting robots. Upon receiving the location information and reinforcement instructions, the welding operation fire automatic identification and intelligent fire-fighting robot automatically navigates to the fire area and performs automatic fire-fighting operations alongside the original vehicle.

[0032] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0033] The present invention provides an automatic fire identification and intelligent firefighting method, robot, and system for welding operations. By equipping a traditional inspection robot with lidar, a long-range infrared temperature sensor, a visible light camera, and fire extinguishing equipment, an automatic fire identification and intelligent firefighting robot for welding operations is formed. The controller automatically plans the inspection path based on the environmental map created by the lidar in real time, controlling the inspection robot to autonomously navigate and avoid obstacles within the welding operation site. During the navigation and inspection process, the controller, on the one hand, uses the long-range infrared temperature sensor to acquire real-time temperature field distribution data of the welding operation site along the inspection path according to a scanning matrix to determine whether a fire has occurred. On the other hand, the controller uses the real-time temperature field distribution data of the welding operation site acquired by the long-range infrared temperature sensor and the multimodal fusion of images of the welding operation site captured in real time by the visible light camera, based on a trained large model, to determine whether a fire has occurred. In the event of a fire, the system controls inspection robots to navigate to the fire area and automatically directs firefighting equipment to extinguish the fire. This intelligent firefighting system automates the identification and extinguishing of fires occurring during welding operations, creating a smart linkage mechanism between fire identification and firefighting equipment execution. This achieves a closed-loop process of early fire detection and extinguishing, facilitating early identification and control of welding fires. It avoids the problem of significant property damage caused by post-fire identification and handling, and avoids the risks associated with existing fire prevention and management methods that rely solely on alarm signals to alert personnel to manually retrieve firefighting equipment, leading to delays or untimely responses that could cause the fire to spread. It also reduces the probability of personal injury caused by personnel approaching the fire area to extinguish the fire when it spreads. Attached Figure Description

[0034] Figure 1 This is a flowchart of an automatic fire detection and intelligent fire suppression method for welding operations;

[0035] Figure 2 This is a flowchart of a welding operation fire automatic identification and intelligent fire-fighting robot performing navigation inspection, automatic fire identification and fire alarm level judgment and fire extinguishing operations at a welding operation construction site.

[0036] Figure 3 This is a schematic diagram showing the structural relationship between automatic fire identification and intelligent fire protection in welding operations.

[0037] Figure 4 This is a schematic diagram showing the structural composition of the controller, its power management system, and the fire extinguishing equipment.

[0038] Figure 5 This is a block diagram of an automatic fire detection and intelligent fire protection system for welding operations.

[0039] As shown in the figure:

[0040] 100. Automatic fire identification and intelligent firefighting robot for welding operations; 101. Controller; 102. LiDAR; 103. Visible light camera; 104. Long-range infrared temperature sensor, visible light camera; 105. Fire extinguishing equipment; 1051. Water level sensor; 1052. Water tank; 1053. Adjustable water nozzle; 1054. Solenoid valve; 1055. Water level sensor; 106. Alarm; 107. Positioner; 108. Frame; 109. Battery; 110. Walking mechanism; 111. Power management circuit; 112. Wireless communication module.

[0041] 200. Cloud server;

[0042] 300. Terminal equipment. Detailed Implementation

[0043] The present invention will now be described in detail with reference to the accompanying drawings. The advantages and features of the present invention will become clearer from the following description. It should be noted that the drawings are all in a very simplified form and use non-precise proportions, and are only used to facilitate and clarify the illustration of the embodiments of the present invention.

[0044] Example 1

[0045] Please refer to Figures 1 to 3 This invention provides an automatic identification and intelligent fire suppression method for welding operation fires, comprising:

[0046] Step S1: Install a long-range infrared temperature sensor 104 and a fire extinguishing device 105 on the inspection robot equipped with a lidar 102 and a visible light camera 103, connected to its built-in controller 101. The inspection robot includes, but is not limited to, security inspection robots and quadruped robots. Quadruped robots are also known as robot dogs.

[0047] In step S2, the controller 101 scans the three-dimensional data of the construction site of the welding operation in real time through the lidar 102 to construct an environment map, and automatically plans the inspection path according to the environment map to control the inspection robot to perform autonomous obstacle-avoiding navigation inspection. The inspection robot generates the optimal inspection path from the current position to the arranged inspection task point based on the environment map constructed by the lidar 102 and the simultaneous localization and mapping (SLAM) algorithm, and controls the inspection robot to move to the target position along the inspection path.

[0048] In step S3, the controller 101 uses the long-distance infrared temperature sensor 104 to scan the scanning matrix in a non-contact manner along the inspection path in a sliding window mode to obtain the temperature field distribution data of the welding operation construction site in real time (i.e., infrared temperature measurement) to determine whether a fire occurs in the welding operation, that is, temperature judgment:

[0049] When the temperature field distribution data t within the spatial position range of a certain scanning matrix is not greater than the first threshold T1, that is, t≤T1, it is judged that the temperature is normal and no fire occurs. For example: T1=150°C. Where t represents the maximum temperature value in the temperature field distribution data, that is, t=T max .

[0050] When the temperature field distribution data t within the spatial position range of a certain scanning matrix is greater than the second threshold T2, that is, t>T2, it is judged that the temperature is high-risk and a fire occurs, where T1<T2. For example: T2=300°C.

[0051] When the temperature field distribution data t within the spatial position range of a certain scanning matrix is between the first threshold and the second threshold, that is, T1<t≤T2, it is judged that the temperature is abnormal, and at the same time, the controller 101 performs multimodal fusion judgment on whether a fire occurs in the welding operation through the images collected in real time by the visible light camera 103 and the temperature field distribution data obtained in real time by the long-distance infrared temperature sensor 104:

[0052] When the controller 101 judges that there is a temperature abnormality through the long-distance infrared temperature sensor 104, and at the same time the controller 101 identifies at least one of smoke and flame in the images collected in real time by the visible light camera 103 through the trained large model, and their spatial position ranges overlap and match, that is, when T1<t≤T2 and there are smoke and flame in the same spatial position range, it is judged that a fire occurs, otherwise it is judged that no fire occurs. In order to realize the automatic identification of smoke and flame, a deep learning target detection model (such as the YOLO series) can be adopted for the large model, and the deep learning target detection model is used to identify whether there is a flame or smoke target in the image, and generate a target bounding box and a confidence score. When the confidence score exceeds a preset threshold (for example, 0.75), the existence of the flame or smoke target is confirmed.

[0053] In step S4, when a fire occurs, the controller 101 controls the inspection robot to navigate to the fire occurrence area (for example, stop moving forward at a position 2-3 meters away from the fire source point), and controls the fire extinguishing equipment 105 to aim at the fire occurrence area to automatically perform fire extinguishing operations.

[0054] The long-distance infrared temperature sensor 104 and the visible light camera 103 can statically collect data at each inspection task point, so as to improve the accuracy of data collection, and further improve the accuracy of judging whether a fire occurs in welding operations.

[0055] Please refer to Figure 2 , in order to avoid re-ignition in the fire occurrence area after fire extinguishing is performed, in the automatic welding operation fire identification and intelligent fire protection method provided by the embodiment of the present invention, when performing the fire extinguishing operation (that is, starting the fire extinguishing procedure), the controller 101 continuously monitors the fire extinguishing situation in the fire occurrence area through the long-distance infrared temperature sensor 104 (that is, performs infrared temperature detection). When the maximum value of the temperature field distribution data of the fire occurrence area drops below the third threshold T3 (for example, T3=80°C) and remains for more than a set time period (for example, 30 seconds), that is, the maximum temperature of the target area drops to T3 within 30 seconds, it is determined that the fire extinguishing operation is successful; otherwise, it is determined that the fire extinguishing operation fails, wherein the third threshold is smaller than the first threshold, that is, T3<T1. When the fire extinguishing operation fails, the controller 101 controls the fire extinguishing equipment 105 to execute a secondary fire extinguishing procedure, and continues to judge whether the maximum temperature of the target area drops to T3 within 30 seconds, and if so, it indicates that the fire extinguishing is successful.

[0056] Please refer to Figures 2 to 3 , in order to realize early warning for the operating personnel at the construction site when a fire is detected and judge the degree of fire danger, in the automatic welding operation fire identification and intelligent fire protection method provided by the embodiment of the present invention, the inspection robot is further provided with an alarm 106 connected to the controller 101. In this case:

[0057] When the controller 101 judges that the temperature is at a high risk level through the long-distance infrared temperature sensor 104, it is determined as a high-risk fire alarm, a fire extinguishing instruction is triggered, and the alarm 106 sends out a high-risk fire alarm signal at the construction site to remind the operating personnel at the construction site to evacuate urgently.

[0058] When controller 101 detects an abnormal temperature through long-range infrared temperature sensor 104, and simultaneously identifies smoke and flame in the real-time image captured by visible light camera 103 using a trained large model (e.g., YOLO v8 model), and their spatial locations overlap and match (i.e., spatial matching), a level one fire alarm is determined. A fire extinguishing command is triggered, and an alarm signal is issued at the construction site via alarm 106 to remind workers to evacuate safely. In other words, a level one fire alarm is determined when the confidence level of visible light detection of at least one of smoke and flame is >0.75 and spatial matching occurs.

[0059] When controller 101 detects an abnormal temperature through long-range infrared temperature sensor 104, and simultaneously identifies smoke and flame in real-time images captured by visible light camera 103 using a trained large model, but the spatial locations of the two do not overlap (i.e., only the temperature is abnormal, the spatial location is mismatched), a level two fire alarm is determined, and the inspection robot is controlled to stop navigation and continue monitoring. For example, if no fire occurs after 30 minutes of continuous observation and monitoring, the alarm is deactivated and navigation and inspection resume. Alternatively, if the level two fire alarm changes to a level one fire alarm or a high-risk fire alarm during continuous observation, corresponding actions are taken.

[0060] When the controller 101 determines that the temperature is normal through the long-range infrared temperature sensor 104, and the controller 101 identifies at least one of smoke and flame in the real-time image captured by the visible light camera 103 using the trained large model, it determines that no fire has occurred, and simultaneously classifies it as a level three fire alarm. The controller 101 of the inspection robot then records and maintains observation and inspection. Navigation does not stop at this point.

[0061] The alarm system (106) sends out alarm signals during high-risk and level 1 fire alarms to alert workers at the construction site. The severity of the fire hazard is assessed by classifying fire alarms as high-risk, level 1, level 2, and level 3, allowing for different responses to different situations and preventing excessive alarm interference that could disrupt welding operations.

[0062] Please refer to Figure 4To improve fire extinguishing effectiveness, this invention provides an automatic identification and intelligent fire suppression method for welding operation fires. The fire extinguishing equipment 105 includes a water tank 1051 and a connected water pump 1052, and an adjustable water nozzle 1053 with horizontal omnidirectional rotation and pitch angle adjustment functions. The water pump 1052 and the adjustable water nozzle 1053 are connected to a controller 101. The controller 101 controls the start and stop of the water pump 1052 and, in conjunction with the real-time image signal from a visible light camera 103, controls the horizontal omnidirectional rotation angle and pitch angle of the adjustable water nozzle 1053 to align with the fire area. The adjustable water nozzle 1053 can be a fire-fighting water tank. The fire extinguishing equipment 105 also includes a solenoid valve 1053 connected between the water pump 1052 and the adjustable water nozzle 1053 to control the flow of fire-extinguishing water. In this case, the solenoid valve 1053 is connected to the controller 101.

[0063] Please refer to Figure 3 To determine the remaining water level in the water tank 1051 after firefighting operations, the automatic identification and intelligent fire suppression method for welding fires provided in this embodiment of the invention includes a water level sensor 1055 installed inside the water tank 1051 and extending outside the water tank 1051, connected to the controller 101. The water level sensor 1055 monitors the water level in the water tank 1051, allowing for timely water replenishment and preventing the water supply from running out, thus ensuring continued fire suppression.

[0064] Example 2

[0065] Please refer to Figures 2 to 3 This invention provides an automatic fire detection and intelligent firefighting robot 100 (hereinafter referred to as the firefighting robot 100) for welding operations based on the above-described method, including an inspection robot and a long-range infrared temperature sensor 104 and a fire extinguishing device 105 mounted on it. The inspection robot includes, but is not limited to, security inspection robots and quadruped robots (robot dogs).

[0066] The inspection robot includes a skeleton 108 and a controller 101, a lidar 102, a visible light camera 103, a locator 107, a battery 109, and a walking mechanism 110 on it. The lidar 102, the visible light camera 103, the locator 107, the battery 109, and the walking mechanism 110 are all connected to the controller 101.

[0067] The lidar 102 is mounted on top of the inspection robot.

[0068] The visible light camera 104 can be a high-definition wide-angle camera, which can be installed side by side with the long-distance infrared temperature sensor 104 at the front of the inspection robot.

[0069] The locator 107 can be a GPS or a gyroscope for real-time positioning.

[0070] The battery 109 includes, but is not limited to, rechargeable batteries such as lithium batteries, which are used to provide power to the entire machine. In this embodiment, a large-capacity lithium battery pack is used, with a rated voltage of 24V and a capacity of not less than 20Ah, which can provide the fire-fighting robot 100 with a continuous working endurance of not less than 4 hours.

[0071] The walking mechanism 110 can be a tracked walking mechanism, a wheeled walking mechanism, or a quadrupedal walking mechanism. When the walking mechanism 110 is a tracked walking mechanism, the inspection robot constitutes a tracked vehicle, possessing a certain obstacle-crossing ability and good terrain adaptability, enabling it to travel stably on uneven ground, gravel ground, or environments with small obstacles at construction sites. When the walking mechanism 110 is a wheeled walking mechanism, the inspection robot constitutes a wheeled vehicle, suitable for flat roads. When the walking mechanism 110 is a quadrupedal walking mechanism, the inspection robot constitutes a quadrupedal robot, possessing good obstacle-crossing ability and capable of walking on all terrains. When the walking mechanism is one of the above two vehicle modes, the frame 108 is the chassis.

[0072] To optimize power supply, the inspection robot also includes a circuit management circuit 111, in which the controller 101, the power management circuit 111, and the battery 109 are connected in a closed loop.

[0073] To improve the computing power of the inspection robot, the controller 101 is embedded with an AI chip or connected to a graphics processing unit (GPU). Image recognition processing is then performed using a large model trained on the AI ​​chip or GPU.

[0074] To enable alarms, the inspection robot also includes an alarm 106 installed on the skeleton 108 or its housing and connected to the controller 101, wherein the alarm 106 includes, but is not limited to, an audible and visual alarm 106.

[0075] To enable data exchange, the inspection robot also includes a wireless communication module 112 mounted on the skeleton 108 and connected to the controller 101. The wireless communication module 112 includes, but is not limited to, a 4G / 5G communication module and a WIFI module.

[0076] The fire extinguishing equipment 105 includes a water tank 1051 and a connected water pump 1052, as well as an adjustable water nozzle 1053 with horizontal omnidirectional rotation and pitch angle adjustment functions. The water pump 1052 and the adjustable water nozzle 1053 are connected to a controller 101. The controller 101 controls the start and stop of the water pump 1052 and, in conjunction with the real-time image signal from the visible light camera 103, controls the horizontal omnidirectional rotation angle and pitch angle of the adjustable water nozzle 1053 to aim at the fire area. The water tank 1051 is used to store fire extinguishing water or liquid, and the water pump 1052 is connected to the water tank 1051 and the adjustable water nozzle 1054 via water pipes. To control the flow of fire extinguishing water, the fire extinguishing equipment 105 may also include a solenoid valve 1053 connected between the water pump 1052 and the adjustable water nozzle 1053, wherein the solenoid valve 1053 is connected to the controller 101. The adjustable water nozzle 1053 can be mounted on the inspection robot via a gimbal bracket. In this case, a gimbal bracket with omnidirectional horizontal rotation and pitch adjustment functions can be used instead of the adjustable water nozzle 1053, and the adjustable water nozzle can be replaced with a fixed water nozzle. The adjustable water nozzle 1053 can adjust the spray direction and angle according to the location of the fire source. When a fire extinguishing command is received, the fire extinguishing controller opens the solenoid valve, starts the water pump, and controls the water nozzle to accurately spray and extinguish the fire at the fire source.

[0077] The automatic fire detection and intelligent firefighting robot 100 for welding operations provided in Embodiment 2 of the present invention works on the following principle:

[0078] The inspection robot autonomously roams along a preset inspection route or within the welding site, continuously collecting infrared temperature data and visible light image data via a long-range infrared temperature sensor 104 and a visible light camera 103, respectively. When high-temperature welding slag or sparks from the welding operation cause a temperature rise in a certain area, a temperature assessment is first performed. If the temperature is below a first threshold, it indicates normal temperature and no fire has occurred during the welding operation. If the temperature exceeds a second threshold, it indicates a high-risk temperature and a fire has occurred, automatically initiating firefighting operations. If the temperature exceeds the first threshold but not the second threshold, a large model is used to detect flames and smoke in the visible light image of the area. Multimodal data fusion is used to determine the fire alarm level and whether a fire has occurred. When a level one fire alarm is detected, a fire is identified, a firefighting task command is automatically generated, a navigation path is planned, and the robot autonomously navigates to the fire source. The inspection robot moves to the target area. Upon arrival, the firefighting equipment adjusts the water nozzles to point at the fire source for precise spraying. After extinguishing the fire, continuous temperature monitoring confirms the effectiveness of the firefighting, completing a closed-loop process.

[0079] Example 3

[0080] Please refer to Figure 5 and combined Figures 1 to 4 This invention also provides an automatic identification and intelligent fire protection system for welding operations, comprising:

[0081] The welding operation fire automatic identification and intelligent fire-fighting robot 100 is the welding operation fire automatic identification and intelligent fire-fighting robot 100 of the above embodiment 2. The welding operation fire automatic identification and intelligent fire-fighting robot 100 includes a wireless communication module 112 connected to its controller 101.

[0082] The cloud server 200 is wirelessly connected to the welding operation fire automatic identification and intelligent fire-fighting robot 100 via the wireless communication module 112. It receives and pushes information (raw, process, and result information) stored in the controller 101 of the welding operation fire automatic identification and intelligent fire-fighting robot 100 regarding whether a fire has occurred during the welding operation. The pushed information includes raw temperature and image data of the inspection process, temperature anomaly information, fire alarm information, inspection robot status information, and whether a fire has occurred. For example, the cloud server 200 periodically or in real-time receives status information from the fire-fighting robot 100, such as the current location of the inspection robot, battery level, water tank volume, inspection path, and temperature distribution. In the event of a high-risk fire alarm, a level-one fire alarm, or both, it pushes alarm information and on-site images in real-time.

[0083] The terminal device 300 is connected to the cloud server 200 via a wireless communication module 112, and is used to receive information pushed by the cloud server 200 regarding whether a fire has occurred during welding operations. The terminal device 300 includes, but is not limited to, smartphones, tablets, and monitoring computers. The terminal device 300 can also send remote control commands, including but not limited to assigning inspection task points, to the fire-fighting robot 100 through the cloud server 200. This enables data interaction and communication between the terminal device 300 and the fire-fighting robot 100.

[0084] To prevent fire hazards from persisting due to failed firefighting operations, the welding operation fire automatic identification and intelligent fire protection system provided in this embodiment of the invention sends a manual firefighting request message to the controller 101 and pushes it to the terminal device 300 via the cloud server 200 when firefighting operations fail. This alerts on-site workers to promptly carry out manual firefighting.

[0085] Please refer to Figure 5To improve firefighting efficiency and ensure successful fire suppression, this invention provides an automatic fire detection and intelligent firefighting system for welding operations. Multiple welding operation fire detection and intelligent firefighting robots 100 are included. When a welding operation fire detection and intelligent firefighting robot 100 automatically detects a fire at a welding operation and controls its fire extinguishing equipment 105 to automatically perform firefighting operations towards the fire area, it sends location information and reinforcement commands to other welding operation fire detection and intelligent firefighting robots 100. Upon receiving the location information and reinforcement commands, the welding operation fire detection and intelligent firefighting robots 100 automatically navigate to the fire area and perform automatic firefighting operations alongside the original vehicle. The figure illustrates two firefighting robots 100, but is not limited to two. Any two firefighting robots 100 can communicate directly or indirectly through a cloud server 200.

[0086] The welding operation fire automatic identification and intelligent firefighting method, robot, and system provided in the above embodiments of the present invention form a welding operation fire automatic identification and intelligent firefighting robot 100 by equipping a traditional inspection robot with a lidar 102, a visible light camera 103, a long-range infrared temperature sensor 104, and fire extinguishing equipment 105. The controller 101 automatically plans the inspection path based on the environmental map scanned in real time by the lidar 102, controlling the inspection robot to autonomously navigate and avoid obstacles within the welding operation site. During the navigation and inspection process, the controller 101, on the one hand, uses the temperature field distribution data of the welding operation site obtained in real time along the inspection path by the long-range infrared temperature sensor 104 according to the scanning matrix to determine whether a fire has occurred in the welding operation; on the other hand, the controller 101 uses the temperature field distribution data of the welding operation site obtained in real time by the long-range infrared temperature sensor 104 and the images of the welding operation site collected in real time by the visible light camera 103, combined with the trained large model, to construct the fire detection system. Multimodal fusion is used to determine whether a fire has occurred during welding operations. In the event of a fire, an inspection robot is guided to the fire area, and fire extinguishing equipment 105 is automatically directed to extinguish the fire. This intelligent firefighting system automates the identification and extinguishing of welding fires, creating a smart linkage mechanism between fire identification and fire extinguishing. This achieves a closed-loop process of early "identification and extinguishing" of welding fires, facilitating early identification and control. It avoids the problem of significant property damage caused by post-fire identification and handling, and avoids the risks associated with existing fire prevention management methods that rely solely on alarm signals to alert personnel to manually retrieve fire extinguishing equipment 105 for manual extinguishing, leading to delays or untimely responses that could cause the fire to spread. Furthermore, it reduces the probability of personal injury caused by personnel approaching the fire area to extinguish the fire when it spreads.

[0087] The automatic identification and intelligent fire suppression method, robot, and system for welding operations provided in the above embodiments of the present invention can integrate early detection of potential fire sources such as welding sparks and high-temperature welding slag at the construction site, multi-level fire alarm identification, autonomous path planning, and precise fire suppression into a closed-loop process. This solves the problems of untimely identification of fire hazards in welding operations at construction sites, low accuracy of early warning, and delayed fire suppression response in existing technologies. Furthermore, the all-terrain walking advantage of the quadruped robot overcomes the poor adaptability of single-wheeled walking mechanisms in complex terrain.

[0088] The automatic identification and intelligent fire-fighting method, robot and system for welding operations provided in the above embodiments of the present invention are particularly suitable for scenarios and working conditions where welding operations are frequent and combustible materials are complexly distributed, such as construction sites and steel structure construction sites.

[0089] This invention is not limited to the specific embodiments described above. Obviously, the embodiments described above are only a part of the embodiments of this invention, not all of them. All other embodiments obtained by those skilled in the art based on the described embodiments of this invention are within the scope of protection of this invention. Those skilled in the art can make other modifications and variations to this invention. Therefore, if these modifications and variations of this invention fall within the scope of the claims of this invention, then this invention also intends to include these modifications and variations.

Claims

1. A method for automatic identification and intelligent fire suppression in welding operations, characterized in that, Long-range infrared temperature sensors and fire extinguishing equipment are installed on inspection robots equipped with lidar and visible light cameras and connected to their built-in controllers. The controller uses LiDAR to scan the 3D data of the welding site in real time to build an environmental map and automatically plans the inspection path based on the environmental map to control the inspection robot to autonomously avoid obstacles and navigate for inspection. The controller uses a long-range infrared temperature sensor to acquire real-time temperature field distribution data of the welding operation site along the inspection path in a non-contact manner according to a scanning matrix to determine whether a fire has occurred during the welding operation. When the temperature field distribution data within the spatial range of a certain scan matrix is ​​not greater than the first threshold, it is determined that the temperature is normal and no fire has occurred. When the temperature field distribution data within the spatial location range of a certain scan matrix is ​​greater than the second threshold, it is determined that the temperature is high-risk and a fire has occurred, where the second threshold is greater than the first threshold. When the temperature field distribution data within a certain spatial range of a scanning matrix falls between a first threshold and a second threshold, a temperature anomaly is determined. Simultaneously, the controller uses a multimodal fusion method, combining images captured in real-time by a visible light camera with temperature field distribution data acquired in real-time by a long-range infrared temperature sensor, to determine whether a fire has occurred during the welding operation. When the controller detects an abnormal temperature through a long-range infrared temperature sensor, and at the same time, the controller identifies smoke and flame in real-time images captured by a visible light camera through a trained large model, and the spatial locations of the two overlap and match, the controller determines that a fire has occurred; otherwise, it determines that no fire has occurred. In the event of a fire, the controller directs the inspection robot to navigate to the fire area and controls the fire extinguishing equipment to automatically perform fire extinguishing operations.

2. The automatic identification and intelligent fire suppression method for welding operation fires according to claim 1, characterized in that, During firefighting operations, the controller continuously monitors the firefighting situation in the fire area through a long-range infrared temperature sensor. When the maximum value of the temperature field distribution data in the fire area drops below the third threshold and remains above the set time period, the firefighting operation is deemed successful; otherwise, the firefighting operation is deemed to have failed. The third threshold is less than the first threshold. If firefighting operations fail, the controller will instruct the fire extinguishing equipment to execute a secondary firefighting procedure.

3. The automatic identification and intelligent fire suppression method for welding operation fires according to claim 1, characterized in that, The inspection robot is also equipped with an alarm that is connected to the controller; When the controller determines that the temperature is high through the long-distance infrared temperature sensor, it identifies it as a high-risk fire alarm and sends out a high-risk fire alarm signal at the construction site through the alarm device. When the controller detects an abnormal temperature through a long-range infrared temperature sensor, and at the same time, the controller identifies smoke and flame in real-time images captured by a visible light camera through a trained large model, and the spatial locations of the two overlap and match, it determines that it is a level one fire alarm and issues a level one fire alarm signal at the construction site through the alarm. When the controller detects an abnormal temperature through a long-range infrared temperature sensor, and at the same time, the controller identifies smoke and flame in real-time images captured by a visible light camera through a trained large model, but the spatial location ranges of the two do not overlap or match, it determines that it is a level two fire alarm and controls the inspection robot to stop navigation and continue monitoring. When the controller determines that the temperature is normal through a long-range infrared temperature sensor, and the controller identifies at least one of smoke and flame in the real-time images captured by the visible light camera through a trained large model, it determines that no fire has occurred and simultaneously determines that it is a level three fire alarm. The controller of the inspection robot records and maintains observation and inspection.

4. The automatic identification and intelligent fire suppression method for welding operation fires according to claim 1, characterized in that, The fire extinguishing equipment includes a water tank and a connected water pump, and an adjustable water nozzle with horizontal omnidirectional rotation and pitch angle adjustment functions. The water pump and the adjustable water nozzle are connected to a controller. The controller controls the start and stop of the water pump and, in conjunction with the real-time image signal from a visible light camera, controls the horizontal omnidirectional rotation angle and pitch angle of the adjustable water nozzle to aim at the fire area.

5. The automatic identification and intelligent fire-fighting method for welding operation fires according to claim 4, wherein the fire-fighting equipment further includes a water level sensor installed inside the water tank and extending outside the water tank and connected to the controller.

6. An automatic fire detection and intelligent firefighting robot for welding operations, characterized in that, The welding operation fire automatic identification and intelligent fire protection method according to any one of claims 1-5 includes an inspection robot with lidar and visible light camera, and a long-range infrared temperature sensor and fire extinguishing equipment mounted on it. The controller built into the inspection robot is connected to the lidar, visible light camera, long-range infrared temperature sensor and fire extinguishing equipment.

7. The automatic fire identification and intelligent firefighting robot for welding operations according to claim 6, characterized in that, The inspection robot also includes a skeleton and its walking mechanism, battery, power management circuit and locator. The controller is connected to the battery through the power management circuit. The walking mechanism is a tracked walking mechanism, a wheeled walking mechanism or a quadrupedal walking mechanism.

8. An automatic identification and intelligent fire suppression system for welding operations, characterized in that, include: The welding operation fire automatic identification and intelligent fire-fighting robot is the welding operation fire automatic identification and intelligent fire-fighting robot as described in claim 6, and the welding operation fire automatic identification and intelligent fire-fighting robot further includes a wireless communication module connected to its controller; The cloud server is wirelessly connected to the welding operation fire automatic identification and intelligent fire-fighting robot via a wireless communication module. It is used to receive and push information on whether a fire has occurred in the welding operation stored in the controller of the welding operation fire automatic identification and intelligent fire-fighting robot. The terminal device is connected to the cloud server via a wireless communication module to receive information from the cloud server regarding whether a fire has occurred during the welding operation.

9. The automatic identification and intelligent fire suppression system for welding operations according to claim 8, characterized in that, When firefighting operations fail, the controller sends a manual firefighting request and pushes it to the terminal device via the cloud server.

10. The automatic identification and intelligent fire protection system for welding operations according to claim 8, characterized in that, The welding operation fire automatic identification and intelligent fire-fighting robot consists of multiple robots. When a welding operation fire automatic identification and intelligent fire-fighting robot automatically identifies a welding operation fire and controls its fire-fighting equipment to automatically perform fire-fighting operations in the fire area, it sends location information and reinforcement instructions to other welding operation fire automatic identification and intelligent fire-fighting robots. Upon receiving the location information and reinforcement instructions, the welding operation fire automatic identification and intelligent fire-fighting robot automatically navigates to the fire area and performs automatic fire-fighting operations alongside the original vehicle.

Images