Detection cabin fire-fighting monitoring and fire-fighting processing method and system based on multiple sensors

Abstract

The invention provides a multi-sensor-based detection cabin fire-fighting monitoring and fire-fighting processing method and system, and relates to the technical field of monitoring processing, and the method comprises the steps: obtaining multi-sensor real-time monitoring data, employing a deep learning model to evaluate a fire behavior state, employing a distributed edge calculation mode to carry out a fire behavior response decision based on a fire behavior evaluation result, and the movable fire extinguishing device is controlled to autonomously move to the optimal fire extinguishing position, spraying parameters are adaptively adjusted, the execution effect is monitored in real time, and the deep learning model is fed back and optimized. The fire behavior evaluation accuracy and the fire-fighting processing efficiency are improved, and the reliability and the adaptability of the system are enhanced.

Description

Technical Field

[0001] The present invention relates to monitoring and processing technology, and in particular to a multi-sensor based detection cabin fire monitoring and fire processing method and system. Background Art

[0002] With the continuous development of industrial automation and intelligent manufacturing, the safety of inspection cabins, as a vital component of industrial facilities, is receiving increasing attention. Inspection cabins typically store precision instruments and equipment, as well as flammable and explosive items. A fire accident can not only cause severe property damage but also trigger a chain reaction of safety incidents. Traditional inspection cabin fire monitoring systems primarily rely on a single type of sensor for fire detection, such as temperature sensors and smoke sensors. When abnormal data is detected, an alarm is triggered and a pre-set fire prevention process is executed. With the development of artificial intelligence technology, intelligent fire monitoring systems based on multi-sensor fusion and deep learning have gradually become a research hotspot. These systems can improve the accuracy and timeliness of fire detection by comprehensively analyzing data from multiple sensors.

[0003] However, the existing detection cabin fire monitoring technology still has some obvious defects and shortcomings. First, most of the existing technologies use a centralized computing architecture for fire assessment and decision-making. The computing resources are concentrated on a single server. When the system load is high, response delays are prone to occur, and it cannot meet the needs of real-time fire handling in complex environments. Secondly, the traditional fixed fire extinguishing device is fixed in position, and the fire extinguishing agent injection parameters are preset. It lacks the ability to adaptively adjust according to the dynamic changes of the fire source, resulting in low fire extinguishing efficiency, especially for complex and changeable fire scenes. Finally, the existing system generally lacks a real-time monitoring and feedback mechanism for the effectiveness of fire treatment. It is impossible to dynamically adjust the strategy according to the actual situation during the fire extinguishing process, nor can it continuously optimize the system's fire assessment model, making it difficult to cope with complex and changeable fire development situations. Summary of the Invention

[0004] The embodiments of the present invention provide a multi-sensor based detection cabin fire monitoring and fire treatment method and system, which can solve the problems in the prior art.

[0005] A first aspect of an embodiment of the present invention provides a multi-sensor-based detection cabin fire monitoring and fire treatment method, comprising: Acquire real-time monitoring data from multiple sensors installed in the detection cabin, and use a deep learning model to perform real-time assessment of the fire status of the detection cabin based on the real-time monitoring data of the multiple sensors to obtain a fire assessment result; A distributed edge computing method is used to make fire response decisions based on the fire assessment results. The fire assessment results are distributed to multiple edge computing nodes for parallel computing processing. Each edge computing node generates a corresponding fire handling strategy based on preset fire response rules, and votes on the fire handling strategies through a consensus protocol. The fire handling strategy with the highest number of votes is selected as the final fire response decision result. According to the fire response decision result, the fire protection system in the detection cabin is controlled to perform corresponding fire protection treatment operations. The fire protection system includes a movable fire extinguishing device. The movable fire extinguishing device can autonomously move to the optimal fire extinguishing position according to the fire source location information, and adaptively adjust the injection pressure and injection angle of the fire extinguishing agent according to the fire development level. The execution effect of the fire protection treatment operation is monitored in real time, and the monitoring results are fed back to the deep learning model to optimize the fire assessment accuracy of the deep learning model.

[0006] Real-time monitoring data from multiple sensors installed in the detection cabin is obtained. Based on the real-time monitoring data from the multiple sensors, a deep learning model is used to perform a real-time assessment of the fire status of the detection cabin. The fire assessment results obtained include: Acquire real-time monitoring data from multiple sensors in the detection chamber, the multiple sensors including a temperature sensor, a smoke sensor, and a gas concentration sensor; A deep learning model optimized by transfer learning is used to build a multimodal data fusion network based on the correlation analysis of time series features of different types of sensor data to perform feature extraction and fire situation analysis on the real-time monitoring data. According to the results of the fire situation analysis, a fire assessment result including the probability of fire occurrence, fire type determination and fire development level is output.

[0007] Based on the fire assessment results, a distributed edge computing method is used to make fire response decisions. The fire assessment results are distributed to multiple edge computing nodes for parallel computing processing. Each edge computing node generates a corresponding fire handling strategy based on preset fire response rules, including: Extracting features from the fire assessment results according to fire type, fire severity, and fire source location, and dividing the fire response task into a plurality of prioritized subtasks based on the extracted features; The subtasks are matched with resources based on the task priority and the real-time load status of the edge computing nodes, and the subtasks are allocated to the corresponding edge computing nodes using a dynamic load balancing algorithm, so that multiple edge computing nodes can synchronously process subtasks of different priorities based on a distributed parallel computing architecture; After receiving the assigned subtask, each edge computing node generates a corresponding fire handling strategy based on the fire response rules corresponding to the subtask.

[0008] A dynamic load balancing algorithm is used to assign subtasks to corresponding edge computing nodes, enabling multiple edge computing nodes to synchronously process subtasks of different priorities based on a distributed parallel computing architecture. These subtasks include: Monitor the resource utilization of each edge computing node in real time, build a node health assessment model based on the resource utilization, obtain the availability score of each node, and match the availability score with the priority of the subtask to be assigned; A task allocation priority queue is established based on the order of the availability scores. A task scheduling algorithm that takes load balancing into consideration is used to preferentially allocate priority subtasks to nodes with availability scores higher than a preset score threshold. The load level of each node is ensured to be consistent by dynamically adjusting the task allocation ratio. Each edge computing node starts task processing in the order of assigned task priority and maintains computational synchronization between nodes through a timestamp-based message passing mechanism, so that subtasks of different priorities can be coordinated and processed in parallel under a distributed architecture.

[0009] According to the fire response decision result, the fire protection system in the detection cabin is controlled to perform corresponding fire protection processing operations, wherein the fire protection system includes a movable fire extinguishing device, and the movable fire extinguishing device can autonomously move to the optimal fire extinguishing position according to the fire source location information, including: Analyzing the fire source location information and fire development parameters according to the fire response decision result, and determining the optimal fire extinguishing position based on the fire source location information and the fire development parameters; Based on the optimal fire extinguishing position and the real-time obstacle distribution in the detection cabin, a movement trajectory is planned for the movable fire extinguishing device that avoids obstacles and has the shortest path, and a motion control instruction sequence including position, speed and direction is generated; The movable fire extinguishing device is controlled to execute the motion control instruction sequence, continuously update the surrounding environment information and optimize the motion trajectory in real time during the autonomous movement process, and carry out the fire extinguishing operation after reaching the optimal fire extinguishing position.

[0010] Adaptively adjust the injection pressure and injection angle of the fire extinguishing agent according to the fire development level, monitor the execution effect of the fire treatment operation in real time, and feed the monitoring results back to the deep learning model, including: Calculating the current fire intensity index based on the fire source temperature, flame height, and spread speed, and determining the initial injection pressure and injection angle of the fire extinguishing agent according to the fire intensity index; performing a fire extinguishing operation using the initial injection pressure and the injection angle as control parameters, while continuously collecting a fire source temperature change rate and a fire extinguishing agent coverage area through a sensor network, and calculating a fire extinguishing effect evaluation index based on the fire source temperature change rate and the fire extinguishing agent coverage area; The fire extinguishing effect evaluation index is input into a deep learning model, and the initial injection pressure and the injection angle are fine-tuned in real time according to the fire extinguishing effect evaluation index. The fire extinguishing effect is improved through parameter iterative optimization, and the optimized parameter combination is stored in the deep learning model.

[0011] A second aspect of an embodiment of the present invention provides a multi-sensor-based detection cabin fire monitoring and fire processing system, comprising: The first unit is configured to obtain real-time monitoring data from multiple sensors installed in the detection cabin, and use a deep learning model to perform real-time assessment of the fire status of the detection cabin based on the real-time monitoring data of the multiple sensors to obtain a fire assessment result; The second unit is configured to make a fire response decision based on the fire assessment result using a distributed edge computing method, distribute the fire assessment result to multiple edge computing nodes for parallel computing processing, each edge computing node generates a corresponding fire handling strategy based on a preset fire response rule, and votes on the fire handling strategies through a consensus protocol, selecting the fire handling strategy with the highest number of votes as the final fire response decision result; The third unit is used to control the fire-fighting system in the detection cabin to perform corresponding fire-fighting treatment operations according to the fire response decision result. The fire-fighting system includes a movable fire-fighting device, which can autonomously move to the optimal fire-fighting position according to the fire source location information, and adaptively adjust the injection pressure and injection angle of the fire-fighting agent according to the fire development level, monitor the execution effect of the fire-fighting treatment operation in real time, and feed back the monitoring results to the deep learning model to optimize the fire assessment accuracy of the deep learning model.

[0012] According to a third aspect of an embodiment of the present invention, an electronic device is provided, including: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0013] According to a fourth aspect of an embodiment of the present invention, a computer-readable storage medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the method described above is implemented.

[0014] The beneficial effects of this application are as follows: By using a deep learning model to conduct real-time assessment of the fire status of multi-sensor data, the accuracy and timeliness of fire detection are improved, the false alarm and missed alarm problems caused by a single sensor are avoided, and the system can fully grasp the development of the fire in the detection cabin.

[0015] The use of distributed edge computing to make fire response decisions achieves efficient utilization of computing resources. Through multi-node parallel processing and consistent voting mechanisms, the scientific nature and reliability of fire handling strategies are ensured, significantly improving system response speed and decision-making quality.

[0016] The equipped movable fire extinguishing device can autonomously move to the optimal fire extinguishing position according to the location of the fire source, and adaptively adjust the fire extinguishing parameters according to the level of fire development, realizing intelligent and precise fire fighting. At the same time, it continuously optimizes the deep learning model through the real-time monitoring feedback mechanism, forming a self-improving closed-loop system, which greatly improves the efficiency and success rate of fire fighting. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 Schematic diagram of the flow of a multi-sensor-based fire monitoring and fire treatment method for a detection cabin according to an embodiment of the present invention; Figure 2 This is a flowchart of edge computing node task allocation in an embodiment of the present invention. DETAILED DESCRIPTION

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0019] The technical solution of the present invention is described in detail below with reference to specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0020] Figure 1 FIG is a flow chart of a method for fire protection monitoring and fire protection treatment in a detection cabin based on multiple sensors according to an embodiment of the present invention. Figure 1 As shown, the method includes: Acquire real-time monitoring data from multiple sensors installed in the detection cabin, and use a deep learning model to perform real-time assessment of the fire status of the detection cabin based on the real-time monitoring data of the multiple sensors to obtain a fire assessment result; A distributed edge computing method is used to make fire response decisions based on the fire assessment results. The fire assessment results are distributed to multiple edge computing nodes for parallel computing processing. Each edge computing node generates a corresponding fire handling strategy based on preset fire response rules, and votes on the fire handling strategies through a consensus protocol. The fire handling strategy with the highest number of votes is selected as the final fire response decision result. According to the fire response decision result, the fire protection system in the detection cabin is controlled to perform corresponding fire protection treatment operations. The fire protection system includes a movable fire extinguishing device. The movable fire extinguishing device can autonomously move to the optimal fire extinguishing position according to the fire source location information, and adaptively adjust the injection pressure and injection angle of the fire extinguishing agent according to the fire development level. The execution effect of the fire protection treatment operation is monitored in real time, and the monitoring results are fed back to the deep learning model to optimize the fire assessment accuracy of the deep learning model.

[0021] In an optional embodiment, real-time monitoring data from multiple sensors installed in the detection cabin is obtained, and a deep learning model is used to perform real-time assessment of the fire status of the detection cabin based on the real-time monitoring data of the multiple sensors. The fire assessment result includes: Acquire real-time monitoring data from multiple sensors in the detection chamber, the multiple sensors including a temperature sensor, a smoke sensor, and a gas concentration sensor; A deep learning model optimized by transfer learning is used to build a multimodal data fusion network based on the correlation analysis of time series features of different types of sensor data to perform feature extraction and fire situation analysis on the real-time monitoring data. According to the results of the fire situation analysis, a fire assessment result including the probability of fire occurrence, fire type determination and fire development level is output.

[0022] Real-time monitoring data is acquired from multiple sensors within the inspection chamber, including temperature sensors, smoke sensors, and gas concentration sensors. The temperature sensor collects temperature changes at various locations within the inspection chamber, with a sampling frequency of 1 Hz, a measurement range of -20°C to 150°C, and an accuracy of ±0.5°C. The smoke sensor monitors smoke particle concentration within the chamber, with a sampling frequency of 2 Hz, a detection range of 0-500 μg / m³, and an accuracy of ±3%. The gas concentration sensor monitors the concentrations of carbon monoxide, carbon dioxide, and combustible gases, with a sampling frequency of 1 Hz. The detection ranges for carbon monoxide are 0-1000 ppm, carbon dioxide 0-5000 ppm, and combustible gas 0-100%LEL, with accuracies of ±5 ppm, ±30 ppm, and ±3%LEL, respectively.

[0023] These sensors are connected to the central processing unit via wired or wireless networks. Real-time data is transmitted in JSON format, containing information such as sensor ID, timestamp, value, and unit. The system uses a data preprocessing module to clean and standardize the raw data, including outlier processing, missing value filling, and data normalization. Outlier processing uses a sliding window median filter with a window size of 5 sampling points; missing values ​​are filled using a forward fill method; and normalization uses the Min-Max method to map various sensor data to the [0, 1] interval.

[0024] For temperature data, the system specifically focuses on temperature rise rate and temperature gradient characteristics, calculating the temperature change rate within a 10-second window and the temperature difference between different locations. For smoke data, the system extracts smoke concentration trend characteristics, including short-term concentration growth rate and fluctuation frequency. For gas concentration data, the system analyzes gas ratios and concentration change patterns, particularly the ratio of carbon monoxide to carbon dioxide, which is crucial for determining fire type.

[0025] A deep learning model optimized with transfer learning is used for fire situation analysis. This model, based on a pretrained ResNet50 architecture, is adapted to the fire detection task through transfer learning. Specifically, the system pretrains the base network on a large fire dataset, then freezes the weights of the first few layers and fine-tunes subsequent layers to adapt to the specific detection chamber environment. The model input is a time series feature matrix composed of multi-sensor data, with a time window length of 30 seconds.

[0026] To achieve multimodal data fusion, this example constructs a two-stream network architecture. One stream processes temperature and smoke data, using a one-dimensional convolutional neural network to extract temporal features. The convolution kernel size is 3, the stride is 1, and there are four convolutional layers, each followed by batch normalization and a Reluctant Unit (ReLU) activation function. The other stream processes gas concentration data using a long short-term memory (LSTM) network architecture with 128 hidden layers and two layers. The features of the two streams are fused using an attention mechanism to generate a comprehensive feature representation. The attention weights are dynamically adjusted based on the confidence level of different sensor data, giving higher weight to reliable temperature sensor data and increasing the weight to gas sensor data when gas concentration fluctuates significantly.

[0027] The model training in this example uses the Adam optimizer with an initial learning rate of 0.001 and a cosine annealing strategy for dynamic learning rate adjustment. The training batch size is 64, and the total number of training rounds is 100. The training data consists of 8,000 samples collected in a simulated environment, covering various fire scenarios (such as electrical fires, oil fires, and solid combustible material fires) and non-fire scenarios. A validation set of 20% is used to monitor the training process and prevent overfitting. To enhance model robustness, data augmentation techniques are also employed, including the addition of Gaussian noise (σ = 0.02) and random time offsets (±2 seconds).

[0028] Based on the fire situation analysis results, the system outputs a three-part fire assessment. The probability of fire occurrence is calculated using a Softmax function, with the output value ranging from 0 to 1. A primary alarm is triggered when the probability exceeds 0.75, and a high-level alarm is triggered when it exceeds 0.9. Fire type determination categorizes fires into four categories: electrical fires, oil fires, solid combustible fires, and non-fires. The fire type is determined based on the probability distribution output by the model. The fire development level is divided into four levels: incubation, initial ignition, development, and peak stage, based on sensor data trends and historical patterns. This is used to predict fire development trends.

[0029] Within 10 seconds, the temperature sensor inside the inspection chamber detected a rapid rise in temperature from 25°C to 48°C, an increase in smoke concentration from 5μg / m³ to 120μg / m³, an increase in carbon monoxide from 3ppm to 85ppm, and an increase in carbon dioxide from 450ppm to 1200ppm. After inputting this data into a deep learning model, the system determined a fire probability of 0.94, identifying it as an electrical fire and a fire development stage of incipient ignition. This assessment result was calculated and output within 1.5 seconds, providing accurate and timely evidence for subsequent firefighting decisions.

[0030] Experimental verification shows that this method has an accuracy rate of 96.3% in early fire detection, a false alarm rate of less than 3%, and an average response time of less than 2 seconds. Compared with traditional threshold judgment methods, it issues an early warning more than 30 seconds earlier.

[0031] In an optional embodiment, a distributed edge computing method is used to make fire response decisions based on the fire assessment results. The fire assessment results are distributed to multiple edge computing nodes for parallel computing processing. Each edge computing node generates a corresponding fire handling strategy based on a preset fire response rule, including: Extracting features from the fire assessment results according to fire type, fire severity, and fire source location, and dividing the fire response task into a plurality of prioritized subtasks based on the extracted features; The subtasks are matched with resources based on the task priority and the real-time load status of the edge computing nodes, and the subtasks are allocated to the corresponding edge computing nodes using a dynamic load balancing algorithm, so that multiple edge computing nodes can synchronously process subtasks of different priorities based on a distributed parallel computing architecture; After receiving the assigned subtask, each edge computing node generates a corresponding fire handling strategy based on the fire response rules corresponding to the subtask.

[0032] like Figure 2 As shown, the method includes: The feature extraction module processes raw fire data. For example, for a fire in a chemical park, the extracted features include: fire type "chemical fire," fire severity "serious," fire source location "storage area A," the presence of flammable and explosive materials within a 50-meter radius, fire spread of 5 meters per minute, northeast wind direction, wind speed 4. This feature extraction utilizes a deep learning model trained on 10,000 fire data sets, achieving an accuracy rate of 95.7%.

[0033] Based on the extracted features, the system divides fire response tasks into prioritized subtasks. In the aforementioned chemical park fire scenario, the system-generated subtasks include: Subtask 1 "Evacuation" (priority 10), Subtask 2 "Isolation of Flammable and Explosive Materials" (priority 9), Subtask 3 "Chemical Fire Extinguishment" (priority 8), Subtask 4 "Monitoring Fire Spread" (priority 7), and Subtask 5 "Traffic Control" (priority 6). Priority settings range from 1 to 10, with higher values ​​indicating higher priorities.

[0034] After the task division is completed, the system needs to assign subtasks to appropriate edge computing nodes. This embodiment uses a dynamic load balancing algorithm for resource matching. The algorithm continuously monitors indicators such as CPU utilization, memory usage, network bandwidth usage, and processing power of each edge computing node. For example, there are 5 edge computing nodes in the system, of which node 1 has a CPU utilization of 30% and a memory usage of 25%; node 2 has a CPU utilization of 65% and a memory usage of 70%; node 3 has a CPU utilization of 45% and a memory usage of 50%; node 4 has a CPU utilization of 20% and a memory usage of 30%; node 5 has a CPU utilization of 55% and a memory usage of 40%.

[0035] The dynamic load balancing algorithm comprehensively considers task priorities and node load status. For the fire scenario described above, the system assigns subtask 1, "Evacuation" (priority 10), to the lightly loaded node 4; subtask 2, "Isolation of Flammable and Explosive Materials" (priority 9), to node 1; subtask 3, "Chemical Fire Extinguishment" (priority 8), to node 5; subtask 4, "Monitoring Fire Spread" (priority 7), to node 3; and subtask 5, "Traffic Control" (priority 6), to the heavily loaded node 2.

[0036] To ensure dynamic load balancing, the system reassesses the load status of each node every five seconds. When a node's CPU utilization exceeds 85% or its memory usage exceeds 90%, the system triggers a task redistribution mechanism, moving low-priority tasks from that node to a less-loaded node. For example, if Node 5's CPU utilization rises to 87%, the system will move Subtask 3, "Chemical Fire Extinguishing," to Node 1 for continued execution.

[0037] After receiving the assigned subtask, each edge computing node generates a corresponding fire response strategy based on pre-set fire response rules. These pre-set fire response rules are stored in a distributed database and include a library of response plans for different fire types, fire severity levels, and environmental conditions. For example, for chemical fires, the system has built-in dedicated fire extinguishing plans for 500 different chemicals; for building fires, the system has built-in 200 fire extinguishing strategies categorized by building structure, materials, and function.

[0038] Taking subtask 1, "Evacuation," as an example, Node 4 generated a detailed evacuation route map based on information such as the fire source location, fire spread rate, wind direction, and building layout. It identified four main exits, estimated an evacuation time of 15 minutes, and generated a timetable for evacuation in different zones and batches. The system also identified two emergency exits that were unsuitable for use due to the fire and planned specialized evacuation routes and rescue plans for people with limited mobility.

[0039] For subtask 2, "Isolating flammable and explosive items," the handling strategy generated by Node 1 includes: establishing an isolation zone 50 meters northwest of the fire source, mobilizing three specialized robots to transfer 12 flammable gas containers in Area A to a safe area, and simultaneously activating the storage area's automatic sprinkler cooling system to control the surrounding temperature below 25°C.

[0040] To ensure the consistency and coordination of distributed computing, a real-time data sharing mechanism is established between edge computing nodes. When a processing strategy generated by one node influences the decisions of other nodes, the relevant information is immediately broadcast to all nodes. For example, when Node 1 determines a route for isolating flammable and explosive materials, this information is immediately notified to Node 4, responsible for personnel evacuation, allowing it to adjust the evacuation route accordingly to avoid intersecting with routes carrying hazardous materials.

[0041] Through the above-mentioned distributed edge computing method, the system can complete the entire process from fire assessment to generating a comprehensive treatment strategy within 3 seconds. Compared with traditional centralized computing methods, the response speed is increased by 78%, and basic functions can still be maintained in the event of partial network failure, greatly improving the efficiency and reliability of fire emergency response.

[0042] In an optional embodiment, a dynamic load balancing algorithm is used to allocate subtasks to corresponding edge computing nodes, so that multiple edge computing nodes can synchronously process subtasks of different priorities based on a distributed parallel computing architecture, including: Monitor the resource utilization of each edge computing node in real time, build a node health assessment model based on the resource utilization, obtain the availability score of each node, and match the availability score with the priority of the subtask to be assigned; A task allocation priority queue is established based on the order of the availability scores. A task scheduling algorithm that takes load balancing into consideration is used to preferentially allocate priority subtasks to nodes with availability scores higher than a preset score threshold. The load level of each node is ensured to be consistent by dynamically adjusting the task allocation ratio. Each edge computing node starts task processing in the order of assigned task priority and maintains computational synchronization between nodes through a timestamp-based message passing mechanism, so that subtasks of different priorities can be coordinated and processed in parallel under a distributed architecture.

[0043] The resource utilization of each edge computing node is monitored in real time. The resource monitoring module collects data such as CPU utilization, memory utilization, network bandwidth utilization, and disk I / O load every 5 seconds. For example, the monitoring results for a particular node at a specific moment might be: CPU utilization of 65%, memory utilization of 58%, network bandwidth utilization of 43%, and disk I / O load of 32%. The system inputs these resource metrics into the node health assessment model for calculation. The node health assessment model uses a weighted average method, assigning weights based on the importance of different resource types. For example, CPU has a weight of 40%, memory has a weight of 30%, network has a weight of 20%, and disk I / O has a weight of 10%. The node availability score is calculated by calculating the weighted average of resource utilization and performing reverse mapping. The availability score ranges from 0 to 100, with higher scores indicating more available resources. Given the above resource utilization, the calculated availability score for this node is 54.

[0044] Assigned subtasks are prioritized based on factors such as business importance, time urgency, and resource requirements. Task priorities are categorized as urgent tasks (priority A, 90-100 points), high priority tasks (priority B, 75-89 points), medium priority tasks (priority C, 50-74 points), and low priority tasks (priority D, 0-49 points). The system maintains a task priority queue and matches task priorities to node availability scores.

[0045] During the task allocation phase, a task allocation priority queue is constructed based on the node availability score. The preset node availability score threshold is set to 65 points. Nodes with an availability score above 65 are given priority for assignment of Class A and Class B tasks. For example, the availability score of edge node 1 is 88 points, that of edge node 2 is 72 points, that of edge node 3 is 63 points, and that of node 4 is 52 points. In this case, the system prioritizes assigning Class A tasks to Node 1, Class B tasks to Nodes 1 and 2, Class C tasks to Nodes 2 and 3, and Class D tasks to Nodes 3 and 4.

[0046] A dynamic task scheduling algorithm is used for load balancing. This algorithm sets a target load deviation threshold of 10%, meaning that the load difference between any two nodes should not exceed 10%. When the system detects that the load difference between nodes exceeds the threshold, it triggers a task reallocation mechanism. For example, if the CPU utilization of Node 1 reaches 85% and that of Node 2 is only 45%, the system will transfer some Class B tasks originally assigned to Node 1 to Node 2 to achieve load balancing. The task allocation ratio is adjusted using a dynamic coefficient, which is determined by the ratio of each node's current load to the system average load. If a node's load is 20% higher than the system average, the node's task acceptance coefficient will be reduced by 25%. Conversely, if the load is 20% lower than the system average, the task acceptance coefficient will be increased by 25%.

[0047] After receiving assigned tasks, each edge computing node initiates processing in order of task priority. Each node maintains an internal priority queue to ensure that high-priority tasks are executed first. For Class A tasks, the system allocates exclusive computing resources; for Class B tasks, the system allocates 75% of available computing resources; for Class C tasks, the system allocates 50% of available computing resources; and for Class D tasks, the system allocates 25% of available computing resources.

[0048] To ensure computational synchronization between nodes in a distributed environment, the system implements a timestamp-based messaging mechanism. Each computational task is assigned a globally unique timestamp during execution, with millisecond-level accuracy. When a node completes a subtask, it generates a message containing the timestamp, task ID, execution results, and status information, which is then broadcast to other relevant nodes via a message queue. Receiving nodes process messages sequentially based on timestamps to ensure data consistency. The system sets a message synchronization timeout threshold of 200 milliseconds. If an expected message is not received within this timeframe, a message retransmission mechanism is triggered.

[0049] The system also implements an exception handling mechanism. If a node's resource utilization exceeds 90% for more than 30 seconds, the system marks the node as overloaded, halts new task assignments, and shifts some existing tasks to other nodes. If a node fails completely, the system triggers a task migration strategy, reallocating any unfinished tasks to other available nodes. The migration order is determined based on task priority.

[0050] Through the above implementation, the present invention realizes dynamic load balancing and task collaborative processing in an edge computing environment, which increases the overall resource utilization of the system by 32%, shortens the average response time of high-priority tasks by 45%, and increases the task processing throughput by 28%, significantly improving the performance and reliability of the edge computing system.

[0051] In an optional embodiment, controlling a fire protection system in the detection cabin to perform corresponding fire protection operations according to the fire response decision result, wherein the fire protection system includes a movable fire extinguishing device, and the movable fire extinguishing device can autonomously move to an optimal fire extinguishing position according to the fire source location information, includes: Analyzing the fire source location information and fire development parameters according to the fire response decision result, and determining the optimal fire extinguishing position based on the fire source location information and the fire development parameters; Based on the optimal fire extinguishing position and the real-time obstacle distribution in the detection cabin, a movement trajectory is planned for the movable fire extinguishing device that avoids obstacles and has the shortest path, and a motion control instruction sequence including position, speed and direction is generated; The movable fire extinguishing device is controlled to execute the motion control instruction sequence, continuously update the surrounding environment information and optimize the motion trajectory in real time during the autonomous movement process, and carry out the fire extinguishing operation after reaching the optimal fire extinguishing position.

[0052] Based on the fire response decision, the firefighting system within the detection cabin is controlled to execute the corresponding firefighting operations. The detection cabin is equipped with a firefighting system that includes a mobile fire extinguishing device. This device can autonomously move to the optimal extinguishing position based on fire source location information, enabling more efficient and accurate firefighting operations.

[0053] The mobile fire-fighting system utilizes a wheeled mobile platform equipped with a fire extinguisher, camera, lidar, processor, communication module, and drive system. Fire extinguishers can be selected from carbon dioxide, dry powder, or water-based, depending on the type of fire within the detection chamber. The camera utilizes a high-definition wide-angle lens, providing real-time images with a 160-degree field of view and a resolution of 1920 x 1080 pixels. The lidar utilizes a 16-line lidar with a scanning range of 360 degrees horizontally and 30 degrees vertically, a ranging accuracy of ±2 cm, and a scanning frequency of 10 Hz. The processor utilizes a quad-core 1.8 GHz CPU with 4 GB of RAM, capable of real-time environmental perception and path planning tasks. The communication module supports Wi-Fi 802.11ac and Bluetooth 5.0 protocols, ensuring stable communication with the fire system's main controller.

[0054] The fire response decision results are used to analyze the fire source location and fire growth parameters. These results typically include information such as the fire source coordinates, fire type, fire spread speed, and direction. For example, the decision indicates that the fire source is located at the coordinates (x=3.5m, y=2.1m, z=0.4m) within the detection chamber, that the fire type is electrical, and that the fire spreads at a rate of 0.3m² / minute, primarily toward the northeast. The system extracts these parameters and stores them in a temporary cache to prepare for subsequent calculation of the optimal fire extinguishing location.

[0055] The optimal fire extinguishing position is determined based on the fire source location information and fire development parameters. The system will comprehensively consider multiple indicators such as fire extinguishing efficiency, safety factors and equipment limitations. For the above-mentioned electrical fire case, the system considers the characteristics of electrical fires and determines that a carbon dioxide fire extinguisher should be used, and it is necessary to maintain a safe distance from the fire source to avoid the risk of electric shock. The system calculates that the optimal spray distance is 1.5 meters. At the same time, considering that the fire mainly spreads to the northeast, the fire source should be approached from the southwest. The coordinates of the optimal fire extinguishing position are finally determined to be (x=2.4m, y=1.0m, z=0.4m). This position can ensure the best fire extinguishing effect and protect equipment safety.

[0056] After determining the optimal fire-fighting position, the system plans a trajectory for the mobile fire-fighting device based on the real-time obstacle distribution within the detection chamber. The system uses lidar and cameras to obtain real-time obstacle information within the detection chamber and construct an environmental map. The system maps the locations of fixed equipment, temporary objects, and personnel within the detection chamber. For example, a 0.5m x 0.8m equipment box is detected at (x=1.8m, y=1.5m), and a row of seats is detected between (x=3.0m, y=0.8m) and (x=4.2m, y=0.8m). The system uses the A* algorithm to calculate a path from the current position (x=0.5m, y=0.5m) to the optimal fire-fighting position (x=2.4m, y=1.0m), planning a trajectory with a total length of 2.8 meters and three turning points. The system then generates a detailed sequence of motion control commands, including the position coordinates, travel speed, and direction adjustment angle for each path segment. For example: the first straight line movement is from (0.5, 0.5) to (1.5, 0.5) at a speed of 0.5 m / s; the first turn is a 45-degree clockwise rotation; the second straight line movement is from (1.5, 0.5) to (2.0, 0.75) at a speed of 0.3 m / s; and so on.

[0057] When controlling the mobile fire-extinguishing device to execute a sequence of motion control commands, the system uses a closed-loop control method, comparing the deviation between the expected trajectory and the actual position in real time and making dynamic adjustments. The drive system of the mobile fire-extinguishing device includes four independently controlled motors with a maximum speed of 1.2m / s, a minimum speed of 0.1m / s, and a steering accuracy of ±2 degrees. During autonomous movement, the lidar scans the surrounding environment at a frequency of 10Hz to detect dynamic obstacles within the range. If a new obstacle is detected, such as a person entering the movement path, the system will immediately update the environmental map and replan the avoidance path. For example, when a new obstacle is detected at the original path (x=2.0m, y=0.8m), the system calculates a new detour path, first deviating 0.3 meters to the right, bypassing the obstacle, and then returning to the original path.

[0058] When the mobile fire extinguishing device reaches the optimal fire extinguishing position, the system will perform position fine-tuning to ensure that the nozzle of the fire extinguishing device is facing the fire source and adjusted to the optimal spray angle and distance. For the above case, the system adjusts the direction of the fire extinguishing device to the northeast and sets the spray angle to 15 degrees to ensure that the carbon dioxide can effectively cover the fire source area. Then, the system activates the fire extinguisher and starts the fire extinguishing operation. During the fire extinguishing process, the system continuously monitors the temperature changes of the fire source through the thermal imaging camera. When the temperature drops to the safety threshold (such as below 60°C), the system determines that the fire is extinguished successfully, stops spraying, and reports the fire extinguishing results to the main controller of the fire protection system.

[0059] The above describes in detail the specific implementation method of the mobile fire-fighting device autonomously moving to the optimal fire-fighting position and performing fire-fighting operations. Through precise location perception, intelligent path planning and real-time environmental adaptation, efficient and accurate fire-fighting operations are achieved.

[0060] In an optional embodiment, adaptively adjusting the injection pressure and injection angle of the fire extinguishing agent according to the fire development level, monitoring the execution effect of the fire treatment operation in real time, and feeding the monitoring results back to the deep learning model include: Calculating the current fire intensity index based on the fire source temperature, flame height, and spread speed, and determining the initial injection pressure and injection angle of the fire extinguishing agent according to the fire intensity index; performing a fire extinguishing operation using the initial injection pressure and the injection angle as control parameters, while continuously collecting a fire source temperature change rate and a fire extinguishing agent coverage area through a sensor network, and calculating a fire extinguishing effect evaluation index based on the fire source temperature change rate and the fire extinguishing agent coverage area; The fire extinguishing effect evaluation index is input into a deep learning model, and the initial injection pressure and the injection angle are fine-tuned in real time according to the fire extinguishing effect evaluation index. The fire extinguishing effect is improved through parameter iterative optimization, and the optimized parameter combination is stored in the deep learning model.

[0061] Firefighting efficiency is optimized by combining fire data collection, fire intensity assessment, fire extinguishing parameter adjustment, and an effect feedback mechanism. The system first deploys a multi-sensor array, including infrared temperature sensors, optical flame detectors, and flow sensors, to collect key data such as fire source temperature, flame height, and spread speed. Temperature sensors are arranged at different heights within a range of 0.5-3 meters around the fire source, with a sampling frequency of 5 times per second to ensure the real-time and accuracy of temperature data. The optical flame detector measures the flame height using image processing technology with an accuracy of ±5 cm. The flame spread speed is calculated by the time difference between adjacent sensors detecting the flame, with typical values ​​ranging from 0.1 to 2 meters per second.

[0062] Based on the acquired data on fire source temperature, flame height, and spread speed, the system calculates the current fire intensity index. The fire intensity index uses a comprehensive scoring mechanism, assigning weight coefficients to temperature, flame height, and spread speed respectively. The temperature weight is 0.5, the flame height weight is 0.3, and the spread speed weight is 0.2. Specifically, when the fire source temperature is detected to be 400°C, the flame height is 1.2 meters, and the spread speed is 0.5 meters per second, the fire intensity index calculated by the system is 7.4 (out of 10 points). Based on this index value, the system determines the initial injection pressure to be 2.8MPa and the injection angle to be 35 degrees from the preset parameter mapping table. This parameter configuration is suitable for medium-intensity fires and can reduce the waste of fire extinguishing agents while ensuring the fire extinguishing effect.

[0063] After the injection system is activated, the control module transmits the initial injection pressure of 2.8 MPa and the injection angle of 35 degrees as control parameters to the actuator. The actuator includes a high-precision pressure regulating valve and a nozzle angle adjustment mechanism driven by a servo motor. The pressure adjustment accuracy reaches ±0.05 MPa and the angle adjustment accuracy is ±1 degree. At the same time, the system continuously collects the temperature change rate of the fire source and the area covered by the fire extinguishing agent through a sensor network. The temperature change rate sensor records temperature data every 200 milliseconds and calculates the temperature drop slope. The fire extinguishing agent coverage area is calculated in real time using an infrared thermal imager combined with an image analysis algorithm, with a coverage area accuracy of ±0.1 square meters. In actual operation, when the temperature drop rate reaches -15°C / second and the fire extinguishing agent coverage area reaches 8 square meters, the system assesses it as having a good fire extinguishing effect.

[0064] Based on the collected temperature change rate and extinguishing agent coverage area, the system calculates a fire extinguishing effectiveness evaluation index. This index takes into account the temperature drop rate (weighted 0.6) and extinguishing agent utilization efficiency (weighted 0.4). Extinguishing agent utilization efficiency is calculated as the ratio of coverage area to consumption. For example, if the temperature drop rate is -15°C / second, the coverage area is 8 square meters, and the extinguishing agent consumption is 20 liters / minute, the calculated fire extinguishing effectiveness evaluation index is 8.2 points (out of a maximum of 10).

[0065] The fire extinguishing effectiveness evaluation index was then input into a deep learning model for analysis. This model utilizes a five-layer neural network structure, consisting of an input layer (two nodes: temperature change rate and coverage efficiency), three hidden layers (containing 64, 32, and 16 neurons, respectively), and an output layer (two nodes: pressure adjustment and angle adjustment). The model was trained using a batch gradient descent algorithm with a learning rate set to 0.001 and a Reluctant Unit (ReLU) activation function. When the fire extinguishing effectiveness evaluation index score of 8.2 was input, the model output parameter adjustment recommendations: pressure adjustment -0.3 MPa, angle adjustment +5 degrees. Based on these recommendations, the system adjusted the spray pressure to 2.5 MPa and the spray angle to 40 degrees to improve fire extinguishing efficiency.

[0066] After the parameter adjustments, the system continued to monitor the firefighting effectiveness. Ten seconds after the parameter adjustments, the temperature drop rate increased to -22°C / second, the coverage area increased to 9.5 square meters, the fire extinguishing agent consumption remained unchanged, and the firefighting effectiveness evaluation index rose to 9.1 points. This demonstrates that the parameter adjustments effectively improved firefighting efficiency. The system stored this successful parameter combination (fire intensity index 7.4, injection pressure 2.5 MPa, injection angle 40 degrees) in the deep learning model's empirical database for rapid parameter matching in similar situations in the future.

[0067] To verify the system's adaptability, tests were conducted under various fire scenarios. For a small oil fire (Fire Intensity Index 5.2), the system automatically adjusted its parameters to 1.8 MPa pressure and a 25-degree angle, reducing fire extinguishing time by 32% compared to a fixed-parameter solution. For a large solid material fire (Fire Intensity Index 8.7), the system adjusted its parameters to 3.4 MPa pressure and a 45-degree angle, reducing extinguishing agent usage by 25%. This demonstrates the system's ability to intelligently adjust extinguishing strategies based on the characteristics of different fires.

[0068] It also features self-learning optimization. After each firefighting operation, actual results data is fed back into the deep learning model, and the model weights are updated through error backpropagation. After accumulating data from 50 actual combat scenarios, the model's prediction accuracy increased from an initial 82% to 94%, and the average parameter adjustment time was reduced from 1.2 seconds to 0.4 seconds, resulting in an average 38% improvement in firefighting efficiency compared to traditional fixed-parameter solutions.

[0069] This implementation method achieves high efficiency in the use of fire extinguishing agents and optimization of fire extinguishing effects through a closed-loop system of adaptive adjustment of fire development levels, real-time monitoring and deep learning feedback, providing a feasible technical path for intelligent firefighting technology.

[0070] A second aspect of an embodiment of the present invention provides a multi-sensor-based detection cabin fire monitoring and fire processing system, comprising: The first unit is configured to obtain real-time monitoring data from multiple sensors installed in the detection cabin, and use a deep learning model to perform real-time assessment of the fire status of the detection cabin based on the real-time monitoring data of the multiple sensors to obtain a fire assessment result; The second unit is configured to make a fire response decision based on the fire assessment result using a distributed edge computing method, distribute the fire assessment result to multiple edge computing nodes for parallel computing processing, each edge computing node generates a corresponding fire handling strategy based on a preset fire response rule, and votes on the fire handling strategies through a consensus protocol, selecting the fire handling strategy with the highest number of votes as the final fire response decision result; The third unit is used to control the fire-fighting system in the detection cabin to perform corresponding fire-fighting treatment operations according to the fire response decision result. The fire-fighting system includes a movable fire-fighting device, which can autonomously move to the optimal fire-fighting position according to the fire source location information, and adaptively adjust the injection pressure and injection angle of the fire-fighting agent according to the fire development level, monitor the execution effect of the fire-fighting treatment operation in real time, and feed back the monitoring results to the deep learning model to optimize the fire assessment accuracy of the deep learning model.

[0071] According to a third aspect of an embodiment of the present invention, an electronic device is provided, including: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0072] According to a fourth aspect of an embodiment of the present invention, a computer-readable storage medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the method described above is implemented.

[0073] The present invention may be a method, an apparatus, a system and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for executing various aspects of the present invention.

[0074] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A fire monitoring and fire treatment method for a detection cabin based on multiple sensors, characterized in that: include: Acquire real-time monitoring data from multiple sensors installed in the detection cabin, and use a deep learning model to perform real-time assessment of the fire status of the detection cabin based on the real-time monitoring data of the multiple sensors to obtain a fire assessment result; A distributed edge computing method is used to make fire response decisions based on the fire assessment results. The fire assessment results are distributed to multiple edge computing nodes for parallel computing processing. Each edge computing node generates a corresponding fire handling strategy based on preset fire response rules, and votes on the fire handling strategies through a consensus protocol. The fire handling strategy with the highest number of votes is selected as the final fire response decision result. According to the fire response decision result, the fire protection system in the detection cabin is controlled to perform corresponding fire protection treatment operations. The fire protection system includes a movable fire extinguishing device. The movable fire extinguishing device can autonomously move to the optimal fire extinguishing position according to the fire source location information, and adaptively adjust the injection pressure and injection angle of the fire extinguishing agent according to the fire development level. The execution effect of the fire protection treatment operation is monitored in real time, and the monitoring results are fed back to the deep learning model to optimize the fire assessment accuracy of the deep learning model.

2. The method according to claim 1, characterized in that Real-time monitoring data from multiple sensors installed in the detection cabin is obtained. Based on the real-time monitoring data from the multiple sensors, a deep learning model is used to perform a real-time assessment of the fire status of the detection cabin. The fire assessment results obtained include: Acquire real-time monitoring data from multiple sensors in the detection chamber, the multiple sensors including a temperature sensor, a smoke sensor, and a gas concentration sensor; A deep learning model optimized by transfer learning is used to build a multimodal data fusion network based on the correlation analysis of time series features of different types of sensor data to perform feature extraction and fire situation analysis on the real-time monitoring data. According to the results of the fire situation analysis, a fire assessment result including the probability of fire occurrence, fire type determination and fire development level is output.

3. The method according to claim 1, characterized in that Based on the fire assessment results, a distributed edge computing method is used to make fire response decisions. The fire assessment results are distributed to multiple edge computing nodes for parallel computing processing. Each edge computing node generates a corresponding fire handling strategy based on preset fire response rules, including: Extracting features from the fire assessment results according to fire type, fire severity, and fire source location, and dividing the fire response task into a plurality of prioritized subtasks based on the extracted features; The subtasks are matched with resources based on the task priority and the real-time load status of the edge computing nodes, and the subtasks are allocated to the corresponding edge computing nodes using a dynamic load balancing algorithm, so that multiple edge computing nodes can synchronously process subtasks of different priorities based on a distributed parallel computing architecture; After receiving the assigned subtask, each edge computing node generates a corresponding fire handling strategy based on the fire response rules corresponding to the subtask.

4. The method according to claim 3, characterized in that A dynamic load balancing algorithm is used to assign subtasks to corresponding edge computing nodes, enabling multiple edge computing nodes to synchronously process subtasks of different priorities based on a distributed parallel computing architecture. These subtasks include: Monitor the resource utilization of each edge computing node in real time, build a node health assessment model based on the resource utilization, obtain the availability score of each node, and match the availability score with the priority of the subtask to be assigned; A task allocation priority queue is established based on the order of the availability scores. A task scheduling algorithm that takes load balancing into consideration is used to preferentially allocate priority subtasks to nodes with availability scores higher than a preset score threshold. The load level of each node is ensured to be consistent by dynamically adjusting the task allocation ratio. Each edge computing node starts task processing in the order of assigned task priority and maintains computational synchronization between nodes through a timestamp-based message passing mechanism, so that subtasks of different priorities can be coordinated and processed in parallel under a distributed architecture.

5. The method according to claim 1, wherein According to the fire response decision result, the fire protection system in the detection cabin is controlled to perform corresponding fire protection processing operations, wherein the fire protection system includes a movable fire extinguishing device, and the movable fire extinguishing device can autonomously move to the optimal fire extinguishing position according to the fire source location information, including: Analyzing the fire source location information and fire development parameters according to the fire response decision result, and determining the optimal fire extinguishing position based on the fire source location information and the fire development parameters; Based on the optimal fire extinguishing position and the real-time obstacle distribution in the detection cabin, a movement trajectory is planned for the movable fire extinguishing device that avoids obstacles and has the shortest path, and a motion control instruction sequence including position, speed and direction is generated; The movable fire extinguishing device is controlled to execute the motion control instruction sequence, continuously update the surrounding environment information and optimize the motion trajectory in real time during the autonomous movement process, and carry out the fire extinguishing operation after reaching the optimal fire extinguishing position.

6. The method according to claim 1, characterized in that Adaptively adjust the injection pressure and injection angle of the fire extinguishing agent according to the fire development level, monitor the execution effect of the fire treatment operation in real time, and feed the monitoring results back to the deep learning model, including: Calculating the current fire intensity index based on the fire source temperature, flame height, and spread speed, and determining the initial injection pressure and injection angle of the fire extinguishing agent according to the fire intensity index; performing a fire extinguishing operation using the initial injection pressure and the injection angle as control parameters, while continuously collecting a fire source temperature change rate and a fire extinguishing agent coverage area through a sensor network, and calculating a fire extinguishing effect evaluation index based on the fire source temperature change rate and the fire extinguishing agent coverage area; The fire extinguishing effect evaluation index is input into a deep learning model, and the initial injection pressure and the injection angle are fine-tuned in real time according to the fire extinguishing effect evaluation index. The fire extinguishing effect is improved through parameter iterative optimization, and the optimized parameter combination is stored in the deep learning model.

7. A multi-sensor-based fire monitoring and fire treatment system for a detection cabin, used to implement the method according to any one of claims 1 to 6, characterized in that: include: The first unit is configured to obtain real-time monitoring data from multiple sensors installed in the detection cabin, and use a deep learning model to perform real-time assessment of the fire status of the detection cabin based on the real-time monitoring data of the multiple sensors to obtain a fire assessment result; The second unit is configured to make a fire response decision based on the fire assessment result using a distributed edge computing method, distribute the fire assessment result to multiple edge computing nodes for parallel computing processing, each edge computing node generates a corresponding fire handling strategy based on a preset fire response rule, and votes on the fire handling strategies through a consensus protocol, selecting the fire handling strategy with the highest number of votes as the final fire response decision result; The third unit is used to control the fire-fighting system in the detection cabin to perform corresponding fire-fighting treatment operations according to the fire response decision result. The fire-fighting system includes a movable fire-fighting device, which can autonomously move to the optimal fire-fighting position according to the fire source location information, and adaptively adjust the injection pressure and injection angle of the fire-fighting agent according to the fire development level, monitor the execution effect of the fire-fighting treatment operation in real time, and feed back the monitoring results to the deep learning model to optimize the fire assessment accuracy of the deep learning model.

8. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Images