Unmanned aerial vehicle intelligent fire-fighting inspection system and method
Through the intelligent fire inspection system of the drone, the drone is equipped with sensors and cameras, the inspection path is planned, the image and video data are captured and transmitted, the fire status is identified and the alarm is issued, solving the problems of low efficiency and major safety hazards in traditional fire inspection methods, and achieving fast, safe and efficient fire inspection and emergency response.
Patent Information
- Application Number
- CN202510246860.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-04
- Publication Date
- 2025-06-03
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Traditional fire inspection methods are inefficient and have great safety hazards, making it difficult to detect fire hazards in a timely manner, and the emergency response speed is slow, which affects the efficiency of rescue decision-making and execution.
Design an intelligent fire inspection system for drones, including design modules, shooting modules, transmission modules, identification modules and alarm modules. Through the drone, it is equipped with sensors and cameras, it plans patrol paths, shoots and transmits images and video data, recognizes fire status and issues alarms.
It has achieved rapid, safe and efficient fire inspections, which can promptly detect fire hazards, respond quickly to fire incidents, reduce the risks of on-site personnel, improve the efficiency of fire rescue, and ensure the safety of life and property.
Smart Images
Figure CN120088918A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of fire inspection, and particularly to an intelligent fire inspection system and method for unmanned aerial vehicles (UAVs). Background Art
[0002] In modern society, sudden disaster events such as fires pose a serious threat to people's lives and property. Traditional fire inspection methods mainly rely on manual inspections, which have many limitations. On the one hand, the efficiency of manual inspections is relatively low, making it difficult to comprehensively and timely cover large areas. On the other hand, in complex terrains or dangerous environments, there are significant safety hazards for inspection personnel, and it is difficult to accurately and quickly detect fire hazards. In addition, the emergency response speed of traditional fire protection systems is also relatively slow, and it is impossible to obtain detailed information on the scene in a timely manner, thus affecting the formulation and implementation efficiency of rescue decisions.
[0003] The rapid development of technologies such as UAV technology, artificial intelligence technology, and communication technology provides new ideas and methods for solving the above problems. UAVs have the advantages of being flexible, able to quickly reach designated areas, and can carry a variety of sensors, enabling efficient inspections of target areas. At the same time, artificial intelligence technology has made breakthrough progress in the field of image recognition and can accurately and quickly identify fire characteristics in images. The application of these technologies in fire inspection systems provides strong support for early fire warning and emergency response. Summary of the Invention
[0004] To solve the technical problems in the above background, the present invention proposes an intelligent fire inspection system and method for UAVs, aiming to construct an efficient, intelligent, and reliable fire inspection solution to improve the efficiency and effectiveness of fire prevention and emergency rescue and ensure people's lives and property safety.
[0005] To achieve the above object, the present invention provides an intelligent fire inspection system for UAVs, including: a design module, a shooting module, a transmission module, an identification module, and an alarm module;
[0006] The design module is used to plan and design the inspection path of the UAV;
[0007] The shooting module is used to conduct inspections based on the designed inspection path, shoot inspection scene images and videos, and temporarily store the image and video data;
[0008] The transmission module is used to transmit the image and video data to the command center;
[0009] The identification module is used to identify the fire status in the image and video data and lock the target area;
[0010] The alarm module is used to issue an alarm and trigger a fire warning disposal mechanism when a fire occurs.
[0011] Preferably, the design module includes: a data acquisition unit, a scene construction unit, and a path planning unit;
[0012] The data acquisition unit is used to collect the site information of the area to be inspected;
[0013] The scene construction unit is used to construct a simulated scene of the area to be inspected based on the site information;
[0014] The path planning unit is used to plan the inspection path based on the simulated scene.
[0015] Preferably, the A-star algorithm is used for path planning, and its expression is as follows:
[0016] f(n) = g(n) + h(n)
[0017] Among them, g(n) represents the actual cost from the starting point to the current node n; h(n) represents the heuristic estimated cost from the current node n to the target node; f(n) represents the total evaluation cost, which is used to determine the priority of the node.
[0018] Preferably, the scene construction unit uses a pre-trained convolutional neural network model to construct a fire recognition model; key features of the fire image are extracted through multi-layer convolution and pooling operations. During the training process of the convolutional neural network, according to the characteristics of the fire recognition task, the number of network layers, the size of the convolutional kernel, the activation function are set, and the cross-entropy loss function and the stochastic gradient descent optimization algorithm are used for model optimization.
[0019] Preferably, the A-star algorithm is used for path planning, and the structure of the fire recognition model includes: the first convolutional layer uses 32 convolutional kernels with a size of 3x3; the second and third convolutional layers use 64 convolutional kernels with a size of 3x3; the window size of the pooling layer is 3x3; the ReLU function is used as the activation function.
[0020] The present invention also provides a method for intelligent fire inspection of drones. The method is used to implement the above system, and the steps include:
[0021] Plan and design the inspection path of the drone;
[0022] Conduct inspections according to the designed inspection path, and capture inspection images and videos, and temporarily store the image and video data;
[0023] Transmit the image and video data to the command center;
[0024] Identify the fire fighting status in the video data and lock the target area.
[0025] Preferably, the A-star algorithm is used for path planning, and its expression is as follows:
[0026] f(n) = g(n) + h(n)
[0027] Among them, g(n) represents the actual cost from the starting point to the current node n; h(n) represents the heuristic estimated cost from the current node n to the target node; f(n) represents the total evaluation cost, which is used to determine the priority of the node.
[0028] Preferably, a pre-trained convolutional neural network model is used to construct a fire recognition model; key features of the fire image are extracted through multi-layer convolution and pooling operations. During the training process of the convolutional neural network, according to the characteristics of the fire recognition task, the number of network layers, the size of the convolutional kernel, and the activation function are set, and the cross-entropy loss function and the stochastic gradient descent optimization algorithm are used for model optimization.
[0029] Compared with the prior art, the beneficial effects of the present invention are as follows:
[0030] The present invention can cooperate with an automated takeoff and landing platform to be on standby for 7×24 hours, quickly respond to fire incidents, reduce the risks of on-site personnel, improve the overall efficiency of fire fighting and rescue, and provide strong support for protecting people's lives and property. BRIEF DESCRIPTION OF THE DRAWINGS
[0031] In order to more clearly illustrate the technical solutions of the present invention, the accompanying drawings required for use in the embodiments will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other accompanying drawings can be obtained based on these drawings without creative efforts.
[0032] Figure 1 It is a schematic diagram of the system structure of the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0033] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some of the embodiments of the present invention, rather than all of them. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.
[0034] In order to make the above objects, features, and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.
[0035] Embodiment 1
[0036] As shown Figure 1 in the system structure flowchart of this embodiment, it includes: a design module, a shooting module, a transmission module, an identification module, and an alarm module; the design module is used to plan and design the inspection path of the drone; the shooting module is used to conduct inspections based on the designed inspection path, and shoot inspection scene images and videos, and temporarily store the image and video data; the transmission module is used to transmit the image and video data to the command center; the identification module is used to identify the fire situation in the image and video data, and lock the target area; the alarm module is used to issue an alarm and trigger the fire warning disposal mechanism when a fire occurs.
[0037] Next, in combination with this embodiment, it will be described in detail how the present invention solves the technical problems in actual work.
[0038] First, the data acquisition unit in the design module acquires the site information of the area to be inspected, including: surrounding environment information, as well as terrain and geographical information. Among them, the surrounding environment information includes: vegetation conditions (such as trees, grasslands, etc.), shadow conditions (specifically referring to areas where light is dim or missing due to the occlusion of objects or terrain), and obstacles (such as buildings, walls, poles, etc.). These factors have an important impact on the flight path and obstacle avoidance of the drone, and help the planning module to conduct path planning and flight safety assessment.
[0039] And the terrain and geographical information includes: terrain height information, ground flatness, and terrain undulation information. Considering the flight height and flight speed of the drone during the inspection process, understanding the slope and undulation of the site can avoid unnecessary shaking or impact risks of the drone during takeoff, landing, or flight.
[0040] After that, the scene construction unit constructs a simulated scene of the area to be inspected based on the acquired site information, and through importing the above parameters into BIM software, conducts three-dimensional modeling to obtain the simulated scene.
[0041] After that, the path planning unit uses the A-star algorithm to conduct path planning according to the constructed scene, and its expression is as follows:
[0042] f(n) = g(n) + h(n)
[0043] Among them, g(n) represents the actual cost from the starting point to the current node n; h(n) represents the heuristic estimated cost from the current node n to the target node; f(n) represents the total evaluation cost, which is used to determine the priority of the node.
[0044] After the path planning is completed, the drone conducts inspections according to the planned path. The shooting module of this embodiment is a high-definition camera mounted on the drone, and the shooting module is used to shoot high-definition images or videos of the inspection area.
[0045] After shooting, the transmission module is used to transmit the images and videos to the command center. The transmission module adopts 5G technology to enable the command center to obtain the drone images and video footages in real time.
[0046] After that, a fire recognition model is constructed through the recognition module to recognize fires in the collected image and video streams or image data. Specifically, in this embodiment, a pre-trained convolutional neural network model is used to construct the fire recognition model. In the convolutional neural network, key features of fire images are extracted through multi-layer convolution and pooling operations. During the training process of the convolutional neural network, according to the characteristics of the fire recognition task, hyperparameters such as the appropriate number of network layers, convolution kernel size, and activation function are set, and the cross-entropy loss function and stochastic gradient descent optimization algorithm are used to optimize the model. The performance of the trained convolutional neural network model is evaluated using the test set samples. By calculating metrics such as recognition accuracy and misrecognition rate, the generalization ability and robustness of the model under different lighting conditions are judged. If the model performance does not meet the expectations, the model is improved by adjusting the network structure, optimization algorithm, increasing training data, etc., and the above training and evaluation process is repeated until a fire recognition model that meets the robustness requirements is obtained.
[0047] Specifically, the model structure of this embodiment is as follows:
[0048] The first convolutional layer uses 32 convolutional kernels with a size of 3x3 to extract low-level features such as edges and textures of fire images. The second and third convolutional layers use 64 convolutional kernels with a size of 3x3 to extract more advanced features, such as smoke and open fire flames. The window size of the pooling layer is 3x3, which can reduce the dimension of the feature map, reduce the amount of calculation, and improve the robustness of the model. At the same time, it can also capture a larger receptive field, and the ReLU function is used as the activation function.
[0049] By calculating metrics such as recognition accuracy and misrecognition rate, the generalization ability and robustness of the model under different lighting conditions are judged. For example, if the recognition accuracy of the model on the test set reaches more than 95% and the misrecognition rate is less than 5%, the model performance is considered good. If the model performance does not meet the expectations, the model needs to be improved.
[0050] After the recognition module completes the recognition, the alarm module activates the corresponding fire warning disposal mechanism according to the severity and destructiveness of the fire, and notifies the firefighters for task arrangement and dispatching.
[0051] In addition, the fire-fighting drone of this embodiment can be deployed in cooperation with an automated takeoff and landing platform. This platform supports the drone to be on standby 7x24 hours. It can be quickly dispatched within 5 minutes after receiving an instruction, greatly shortening the response time. The drone can penetrate into narrow and dangerous areas to detect the environment, reducing the risk of on-site personnel; it can transmit image information in real time to provide reference for rescue decision-making; and it can carry various rescue equipment such as fire-fighting water tanks and transport cabins to carry out operations.
[0052] Embodiment 2
[0053] This embodiment also provides a method for intelligent fire-fighting inspection of drones. The method is used to implement the above system. The steps include: designing the inspection path of the drone; performing inspections according to the designed inspection path, and taking pictures and videos of the inspection scenes, and temporarily storing the image and video data; transmitting the image and video data to the command center; identifying the fire-fighting status in the image and video data, and locking the target area.
[0054] Among them, the A-star algorithm is used for path planning, and its expression is as follows:
[0055] f(n) = g(n) + h(n)
[0056] Among them, g(n) represents the actual cost from the starting point to the current node n; h(n) represents the heuristic estimated cost from the current node n to the target node; f(n) represents the total evaluation cost, which is used to determine the priority of the node.
[0057] A pre-trained convolutional neural network model is used to construct a fire recognition model; key features of the fire image are extracted through multi-layer convolution and pooling operations. During the training process of the convolutional neural network, according to the characteristics of the fire recognition task, the number of network layers, the size of the convolutional kernel, and the activation function are set, and the cross-entropy loss function and the stochastic gradient descent optimization algorithm are used for model optimization.
[0058] The embodiments described above are only descriptions of the preferred embodiments of the present invention, and do not limit the scope of the present invention. Without departing from the design spirit of the present invention, various deformations and improvements made by those of ordinary skill in the art to the technical solutions of the present invention shall fall within the protection scope determined by the claims of the present invention.
Claims
1. An unmanned aerial vehicle intelligent fire inspection system, characterized in that: include: Design module, shooting module, transmission module, recognition module and alarm module; The design module is used to design the inspection path of the drone; The shooting module is used to perform inspections based on the designed inspection path, shoot scenes, and generate and temporarily store image and video data; The transmission module is used to transmit the image and video data to a command center; The recognition module is used to recognize the firefighting status in the image and video data and lock the target area; The alarm module is used to sound an alarm and trigger a fire warning and disposal mechanism when a fire occurs.
2. The UAV intelligent fire inspection system according to claim 1 is characterized in that: The design module includes: a data acquisition unit, a scene construction unit and a path planning unit; The data collection unit is used to collect site information of the area to be inspected; The scene construction unit is used to construct a simulation scene of the area to be inspected based on the inspection site information; The path planning unit is used to plan an inspection path based on the simulation scenario.
3. The UAV intelligent fire inspection system according to claim 2 is characterized in that: The A-star algorithm is used for path planning, and its expression is as follows: f(n)=g(n)+h(n) Among them, g(n) represents the actual cost from the starting point to the current node n; h(n) represents the heuristic estimated cost from the current node n to the target node; f(n) represents the total evaluation cost, which is used to determine the priority of the node.
4. The UAV intelligent fire inspection system according to claim 2 is characterized in that: The scene construction unit uses a pre-trained convolutional neural network model to build a fire recognition model; extracts key features of the fire image through multi-layer convolution and pooling operations; during the convolutional neural network training process, the number of network layers, convolution kernel size, and activation function are set according to the characteristics of the fire recognition task, and the cross entropy loss function and stochastic gradient descent optimization algorithm are used to optimize the model.
5. The UAV intelligent fire inspection system according to claim 4 is characterized in that: The A-star algorithm is used for path planning. The structure of the fire recognition model includes: the first convolution layer uses 32 convolution kernels of size 3x3; the second and third convolution layers use 64 convolution kernels of size 3x3; the window size of the pooling layer is 3x3; and the activation function uses the ReLU function.
6. A method for intelligent fire inspection by unmanned aerial vehicles, the method being used to implement the system according to any one of claims 1 to 5, characterized in that the steps include: Design the inspection path of the drone; Carry out inspections according to the designed inspection routes, take images and videos of the inspection scenes, and temporarily store the image and video data; transmitting the image and video data to a command center; The fire situation in the image and video data is identified, and the target area is locked.
7. The UAV intelligent fire inspection method according to claim 6 is characterized in that: The A-star algorithm is used for path planning, and its expression is as follows: f(n)=g(n)+h(n) Among them, g(n) represents the actual cost from the starting point to the current node n; h(n) represents the heuristic estimated cost from the current node n to the target node; f(n) represents the total evaluation cost, which is used to determine the priority of the node.
8. The UAV intelligent fire inspection method according to claim 6 is characterized in that: A pre-trained convolutional neural network model is used to build a fire recognition model. The key features of the fire image are extracted through multi-layer convolution and pooling operations. During the convolutional neural network training process, the number of network layers, convolution kernel size, and activation function are set according to the characteristics of the fire recognition task. The cross entropy loss function and stochastic gradient descent optimization algorithm are used for model optimization.