6835 results about "Smoke Emission" patented technology

The invention provides a smoke and fire detection method and system based on deep reinforcement learning and electronic equipment, and relates to the technical field of fire fighting. The method comprises the following steps: acquiring thermal imaging data of a forest area through a thermal imager, obtaining temperature distribution conditions of different positions, and determining position coordinates, the number and specific temperature values of temperature abnormal points for an area with abnormal temperature rise; if the temperature value of the temperature abnormal point exceeds a preset temperature threshold value, judging that the temperature abnormal point is a suspected fire point, and collecting a visible light video image of the area through a visible light camera according to the position of the suspected fire point; smog detection processing is carried out on the collected visible light video image, whether smog exists in the image is judged by analyzing the color, texture, motion and other characteristics of the image, and the concentration, morphological characteristics and diffusion speed of the smog are obtained; the meteorological parameters such as the wind speed and the wind direction of a fire occurrence area are obtained through the meteorological sensor, the meteorological parameters are input into the fire spreading model, and the spreading direction and the spreading range of the fire under the current wind field condition are simulated.

The invention relates to the technical field of communication machine room safety protection, and discloses a communication machine room intelligent fire monitoring and fire extinguishing system based on edge computing, which comprises a fire parameter acquisition module, a temperature characteristic analysis module, a smoke dynamic detection module, a multi-source fusion judgment module, a regulation and control parameter generation module and the like. The fire parameter acquisition module acquires parameters such as temperature gradient, smoke concentration and equipment current, judges fire risk and triggers cooperative monitoring; the temperature characteristic analysis module extracts temperature distribution characteristics and quantitatively evaluates hot spot abnormity; the smoke dynamic detection module analyzes the smoke diffusion trend; the multi-source fusion judgment module is combined with multiple parameters to generate a fire extinguishing or optimizing signal; and the regulation and control parameter generation module is matched with the strategy to generate injection and ventilation parameters. The system also corrects a risk index through an equipment coupling analysis module, and forms closed-loop control by using a dynamic feedback execution module. According to the system, multi-source data fusion monitoring and intelligent fire extinguishing are realized, and the fire prevention and control accuracy of the communication machine room is improved.

The invention relates to the technical field of fire early-stage recognition, in particular to a multi-source information-based conveying belt fire early-stage recognition and early-warning system, which comprises a sudden change detection module, a synchronous analysis module, a spatial trend recognition module, a coupling fluctuation screening module and a probability evaluation module. According to the invention, through multi-source information linkage acquisition, collaborative monitoring of parameters such as along-line temperature, smoke, gas, images, load and heat source temperature difference, combined characteristic analysis among parameters and synchronous response trend determination are driven, and active revelation of early risk hidden dangers and dynamic discrimination of spatial distribution consistency and fluctuation continuity are realized. Potential abnormal focusing locking under a high-interference complex working condition is promoted, collaborative fluctuation between a load and a heat source temperature difference further eliminates environmental noise influence, risk weight dynamic adjustment strengthens classification sensitivity of probability identification, fire risk clustering division promotes accurate mastering of distribution of tiny initial hidden dangers, and the probability identification accuracy is improved. And the reliability of fire early warning in a coal mine area conveying belt scene is obviously improved.

The invention relates to the technical field of fire hazard monitoring, and discloses a fire hazard remote monitoring and multistage early warning mechanism integration system which integrates a data acquisition module, a heat gathering map generation module, an image fusion module, a data calculation module, a risk prediction module, an early warning notification module and the like. Through an infrared temperature measurement unit, a smoke concentration acquisition unit and an image acquisition unit, the system can acquire temperature, smoke and image information of a key area in real time. A heat gathering map generation and image fusion technology is utilized, a fusion image is systematically generated, a fire scene and temperature distribution thereof are visually displayed, and the fire source positioning efficiency is improved. Meanwhile, the system analyzes the gradient change of the temperature, the heat accumulation area and the smoke concentration through a data calculation module, constructs a risk prediction model in combination with a machine learning algorithm, and achieves the refined prediction of the fire risk. And according to a prediction result, the early warning notification module automatically sends a corresponding early warning message to ensure rapid transmission and timely response of the early warning message. According to the invention, the intelligent, precise and efficient level of fire monitoring is improved.

The invention discloses a fire smoke and fire source accurate positioning method based on image recognition, and the method comprises the following steps: S1, collecting multi-modal data, and carrying out the graying, denoising, histogram equalization and shielding correction preprocessing; s2, inputting the preprocessed data into a vortex sensing network to generate a preliminary detection result of flame and smoke; s3, dividing the scene into dynamic space blocks by using a dynamic space division algorithm and allocating priorities; s4, eliminating non-fire feature interference through an environment feature filter, and generating an optimization candidate region; s5, inputting the candidate area into an event-driven time sequence analysis module, and extracting a key frame to analyze flame and smoke diffusion and time trend; s6, verifying position consistency by adopting a flame and smoke joint detection method, and generating a three-dimensional space positioning result; and S7, generating a fire three-dimensional thermodynamic diagram according to three-dimensional positioning and triggering an alarm mechanism. According to the invention, accurate detection and positioning of fire smoke and a fire source are realized by using a vortex sensing network, an event-driven mechanism and the like.

The invention discloses an energy storage cabinet intelligent fire extinguishing system and method suitable for a distributed architecture. The energy storage cabinet intelligent fire extinguishing system suitable for the distributed architecture comprises an environment sensing flame smoke detection module, an environment temperature and key position monitoring module, a control module and a fire extinguishing execution module. When detection or / and monitoring data of the environment sensing flame smoke detection module or / and the environment temperature and key position monitoring module reaches a preset threshold value or is abnormal, the control module sends a control instruction to the fire extinguishing execution module; and the fire extinguishing execution module executes fire extinguishing action. According to the energy storage cabinet intelligent fire extinguishing system suitable for the distributed architecture, the safety and reliability of the energy storage cabinet can be improved, and resource waste and potential economic loss caused by false alarm are reduced.

The invention belongs to the technical field of fire sensing and detection, and provides a fire sensing method and related device for a double-layer cable tunnel structure, and the method comprises the steps: respectively collecting the temperature, smoke and image data of a double-layer cable tunnel, and carrying out the double-layer data linkage; respectively constructing a temperature distribution model, a smoke diffusion model and an image target detection model according to the collected temperature, smoke and image data; and performing temperature, smoke and image multi-modal data fusion based on the temperature distribution model, the smoke diffusion model and the image target detection model, and when the fused data meets a set fire threshold, determining that a fire occurs. Through the fire sensing and verification method aiming at the characteristics of the double-layer cable tunnel, scientific and reliable decision support is provided for improving tunnel fire safety management, the safety of personnel and equipment in the tunnel is ensured, and meanwhile, the standardization process of the fire sensing technology is promoted.

The invention discloses a distribution box fire early warning method and system based on multi-source information fusion, and particularly relates to the technical field of distribution box fire early warning. Temperature, smoke, acoustic vibration and current harmonic signals are synchronously acquired, time sequence alignment and amplitude normalization are performed, and a dynamic baseline is generated in real time; extracting multi-scale statistics and morphological characteristics according to an adaptive window, and calibrating an operation mode through unsupervised clustering; then retrieving a matched baseline signature in a historical feature library, and calculating a multi-modal standardization deviation of a current window; mapping the deviation index into a graph node, combining a covariance edge weight, a phase synchronization index and a smoke discrete index, obtaining a coupling risk coefficient through a fusion model, and outputting a comprehensive risk index; and finally, carrying out multi-scale rate and acceleration analysis on the comprehensive risk index, triggering three-level early warning of attention, warning and danger by adopting a dynamic threshold, and correcting the threshold by utilizing operation and maintenance feedback self-learning, thereby effectively solving the problems of early warning response lag and unknown early warning level.

Disclosed are a method, system, and device for removing smoke from laparoscope images based on a conditional diffusion model. The method includes: segmenting a video of a laparoscopic surgery according to the number of frames to form a data set; performing smoke rendering on the obtained laparoscope smokeless images, and synthesizing paired smoky images to obtain a synthetic data set containing the smokeless images and the smoky images; inputting the smokeless images into the conditional diffusion model for forward noise addition, and continuously adding noise until the smokeless images are completely noised; inputting the smoky images into a smoke sensing module to obtain smoke concentration and position information, then training a neural network, and continuously performing reverse denoising on the completely noised images using the trained neural network until clear smokeless images are outputted; and optimizing a smoke removal model through a multi-loss function fusion strategy.

The invention discloses a forest fire hidden danger identification method and device for a power transmission line based on a multi-modal large model, and the method comprises the steps: constructing a forest fire hidden danger identification model for a smog or flame-containing image outputted by a conventional target detection model, and enabling the model to guide a dialogue set through the combination with forest fire hidden danger identification, thereby achieving the recognition of the forest fire hidden danger. The dialogue context and the image visual features are fused, so that the deep fusion of the text semantics and the firework image features is realized; and finally, inputting the deeply fused features into a large language model, so that a forest fire hidden danger output dialogue of the power transmission line can be obtained by utilizing the deep semantic understanding capability of the large language model, and then a forest fire hidden danger recognition result of the power transmission line is obtained. Therefore, according to the invention, the judgment of whether the smoke and fire can damage the power transmission line is realized, the technology upgrade from smoke and fire existence detection to equipment threat research and judgment is completed, and on the basis, the problems of resource waste and untimely crisis response caused by reporting all smoke and fire information in the traditional technology can be avoided.

The invention relates to the technical field of data processing, in particular to a data acquisition method and system based on internet-of-things equipment, and the method comprises the steps: installing a smoke concentration sensor and a temperature sensor in a cargo area of a warehouse, and calculating the smoke concentration and temperature of the cargo area according to the deviation degrees and change rates of the smoke concentration and temperature at the current moment; determining a smoke concentration risk index and a temperature risk index, and determining a comprehensive risk index by combining the mutual information of the smoke concentration and the temperature; and determining a comprehensive value index according to the respective uncertainty of the historical smoke concentration sequence and the historical temperature sequence and the similarity between the historical smoke concentration sequence and the historical temperature sequence and the corresponding standard sequence, and adjusting the acquisition frequency of the sensor by using the comprehensive risk index and the comprehensive value index to realize a self-adaptive data acquisition strategy. According to the method, environmental changes can be responded in time, potential risks can be recognized in time, and generation of excessive redundant data can be avoided.

Cooking devices, cooking systems, and smoke assemblies for use therewith are provided. In one embodiment, a cooking system is provided and includes a housing and a smoke assembly. The housing can have a lower base portion and an upper lid portion that together define a hollow cooking chamber. The smoke assembly can be coupled to the upper lid portion, and can be in fluid communication with the hollow cooking chamber. The smoke assembly can generate smoke during cooking operations in order to impart smoke flavor onto a food product disposed within the hollow cooking chamber.

The invention discloses a fire smoke image early recognition method based on multi-modal fusion. The method comprises the following steps: S1, collecting multi-modal data and carrying out time alignment; s2, constructing a visible light branch, an infrared branch and a photoacoustic branch, generating multi-modal features, and storing the multi-modal features in a shared feature pool; s3, embedding a space-time tracking module in the shared feature pool, and generating global feature representation; s4, generating a dynamic background template by adopting a low-rank matrix decomposition algorithm, and separating and removing background noise; s5, multi-modal features are fused in the shared feature pool through a lightweight self-attention mechanism, and missing features are compensated; s6, enhancing an early smoke signal through a sparse feature enhancement algorithm, and performing fire state classification and smoke diffusion path prediction by applying a multi-task classification network; and S7, updating the shared feature pool and the space-time tracking module through a dynamic feedback mechanism. According to the invention, early-stage accurate identification of fire smoke is realized by using multi-modal fusion, space-time tracking and dynamic feedback methods.

The invention belongs to the technical field of forest fire early warning, and discloses a forest fire prevention method based on smoke recognition, which comprises the following steps: acquiring heterogeneous image data from different monitoring devices, performing format unification and quality enhancement through a data preprocessing module, eliminating the difference of data from different sources, and obtaining preprocessed image data; a multi-scale feature extraction network based on deep learning is adopted, key features such as the color, texture and form of smoke are automatically learned and extracted from the preprocessed image data, and feature representation is obtained; according to the method, corresponding early warning threshold values are set according to different grades of fire judgment results, when the fire grade judgment results meet early warning conditions, early warning signals are triggered automatically, and early warning of the fire is achieved in the modes of message pushing, sound-light alarm and the like.

The invention relates to the technical field of fire-fighting engineering, and discloses a real-time fireproof monitoring data transmission system for fire-fighting engineering construction, which comprises a fire parameter analysis module, an environmental data analysis module, a fire source positioning module, a model construction module and a safety database. The fire parameter analysis module generates a temperature change record table through gridding temperature measurement nodes, estimates the smoke diffusion rate, and evaluates the thermal stress and deformation risk of the building structure; the environment data analysis module associates the air velocity with the smoke diffusion rate and corrects the parameter weight of the evaluation model; the fire source positioning module generates a fire source three-dimensional positioning map by using multispectral imaging equipment; the model construction module constructs a dynamic fire assessment model according to multi-dimensional data and outputs a comprehensive threat level; the security database stores building material parameters, historical case characteristics, and dynamic threshold rules. According to the system, through multi-module cooperation and multi-data fusion, the accuracy and timeliness of fire-fighting monitoring are improved, and the system is suitable for safety management of fire-fighting engineering.

The invention discloses a cable tunnel fire risk feature identification system and identification method, and belongs to the technical field of cable tunnel safety monitoring. The whole stage of a fire is covered through multi-class cooperative detection, the target identification precision and scene adaptability are greatly improved, and the safety of the cable tunnel is improved. According to the whole system, a flame and smoke detection unit and a flame / smoke special detection unit are innovatively added to a detection engine module, an original heat source detection unit and an original human body detection unit are combined, a heat source-flame-smoke-human body four-category collaborative detection framework is formed, high-precision recognition is achieved on the basis of a YOLO model framework, and the detection efficiency is improved. Compared with the problems that a traditional system is high in single target detection omission ratio and cannot cover the whole stage of smoldering-initial open fire-violent combustion of a fire, the system has the advantages that the recognition rate of early flame and weak smoke is increased, the false alarm rate is greatly reduced, meanwhile, a heat source of operation and maintenance personnel is prevented from being misjudged as a fire hazard through human body detection, and the safety of the fire hazard is improved. And cable monitoring and personnel safety protection are both considered.

The invention relates to the technical field of fire alarm systems, in particular to a distributed intelligent smoke monitoring and alarm management system, which comprises a data acquisition and trend analysis module for analyzing a smoke concentration change trend and generating a preliminary early warning signal based on real-time data acquired by a smoke sensor; the method specifically comprises the steps that the smoke concentration value Cs (t) of a smoke sensor in the time sequence is obtained, the average value Cmean and the standard deviation Std of the smoke concentration values in the current period are calculated, the upper limit threshold value and the lower limit threshold value of the smoke concentration are determined according to the formulas that Threshold High = Cmean + K1 * Std and Threshold Low = Cmean-K2 * Std, K1 and K2 are adjustment coefficients and are used for weighing sensitivity, when Cs (t) exceeds Threshold High or is lower than Threshold Low, preliminary early warning is triggered, and when Cs (t) is larger than Threshold Low, the smoke concentration is not subjected to early warning. The system analyzes the smoke concentration trend in real time, filters interference, links environment variables to adjust threshold values, performs multi-stage filtering, performs self-learning optimization, performs intelligent hierarchical scheduling, and efficiently prevents and controls fire disasters.

The invention discloses a ship dangerous goods transportation safety control method and system, and relates to the field of marine transportation, and the method comprises the steps: obtaining dangerous goods parameter data of a dangerous goods monitoring sensor, the dangerous goods parameter data comprising gas concentration, temperature and smoke concentration; under the condition that the sensor parameter data exceeds a preset parameter threshold value, acquiring motion attitude data and inclination angle data of the current ship and pressure distribution data of the cargo on the bottom and the side surface of the cargo hold; calculating the center-of-gravity offset of the current ship based on the motion attitude data, the inclination angle data and the pressure distribution data; calculating a fixed adjustment parameter of the cargo fixing device based on the gravity center offset; and sending a device adjustment instruction containing the fixed adjustment parameter to the target client. By implementing the method, the efficiency of adjusting the cargo fixing device can be improved.

A device, and method for its use, which may be worn by a firefighter for detecting the source of a fire in a smoke-filled or noisy environment, the device having directional infrared sensors for detecting the direction of maximum or minimum intensity of infrared radiation in two or more directions relative to the wearer and reporting at least the direction of maximum or minimum intensity to the wearer by a tactile indicator contacting the skin of the wearer.

The embodiment of the invention relates to the technical field of fire spreading prediction, and particularly discloses a fire spreading trend AI prediction method and system based on a high-rise building. According to the embodiment of the invention, fire alarm information is received; acquiring smoke monitoring data of the target fire building, and performing smoke spreading prediction on the smoke monitoring data; selecting an auxiliary monitoring building; flame monitoring data transmitted by the auxiliary monitoring building are acquired, and flame spreading prediction is carried out; aI prediction of fire hazards and visual display of the fire spreading trend are carried out. Smoke monitoring data of a target fire building can be acquired, auxiliary monitoring analysis is performed, an auxiliary monitoring building is selected, flame monitoring data transmitted by the auxiliary monitoring building is acquired, and AI prediction of fire hazards and visual display of a fire spreading trend are performed by integrating a smoke prediction result and a flame prediction result. The defect of single data acquisition is overcome, and the accuracy of fire spreading trend prediction is effectively ensured in combination with flame monitoring of an auxiliary monitoring building.

The invention relates to an emergency evacuation path generation method and system based on STDE, and the method comprises the following steps: extracting a topological network structure of a target building, and collecting smoke detection data and crowd density; building a building topology layer; obtaining a room grid layer according to the building topology layer; constructing a smoke risk value; constructing a congestion risk value; performing dynamic risk fusion on the smoke risk value and the congestion risk value to obtain a total risk; constructing a cost function from each node to an edge of an adjacent node in the building topology layer, and obtaining an optimal path of the building topology layer through a space-time joint reverse search algorithm; performing path planning on each room in the optimal path of the building topology layer in the corresponding room grid layer to obtain the optimal path of the room grid layer; and integrating the optimal paths of the building topology layer and the room grid layer to obtain a final emergency evacuation path. Compared with the prior art, the accuracy of evacuation risk assessment is improved, and different evacuation strategy requirements can be met.

The smoke detection method based on multi-sensor fusion analysis provided by the invention comprises the following steps: when a smoke sensor detects that the smoke concentration in a detection range reaches a first threshold value, acquiring a temperature rising speed in the detection range detected by a temperature sensor, acquiring the environmental heat radiation uniformity in the detection range detected by the human body sensor; generating a corresponding correction coefficient to correct the smoke concentration and the first threshold value to obtain a corresponding corrected smoke concentration and a corrected threshold value; if the corrected smoke concentration reaches a corrected threshold value, a comprehensive risk value is calculated based on the corrected smoke concentration and the temperature rising speed; and if the comprehensive risk value reaches a risk threshold, determining that the fire is a real fire. According to the invention, through the fusion processing of the smoke sensor, the temperature sensor and the human body sensor, the defect that fire detection through smoke is easily interfered by the environment and misinformation is generated at present is overcome.

The invention relates to the field of intelligent early warning, and particularly discloses a composite home fire early warning system combined with smoke and gas monitoring, which is characterized in that firstly, multi-parameter acquisition is started by taking temperature abnormity as a trigger condition, and gas concentration, smoke diffusion and electrical load data are synchronously acquired; local feature extraction is carried out on the time sequence of each parameter; and then establishing an association relationship among the parameters through a cross-time-sequence dynamic message passing fusion mechanism. And finally, fusing the multi-dimensional time sequence characteristics, inputting the fused multi-dimensional time sequence characteristics into a risk grading model, and judging the current risk grade in combination with a historical fire evolution rule. According to the scheme, through time-space correlation analysis of multi-dimensional parameters, daily environment fluctuation (such as cooking fume) and real fire initial characteristics can be effectively distinguished, the method can adapt to risk characteristics of different house type structures, and upgrading from passive alarm to intelligent pre-judgment is achieved.

The invention provides a tear-resistant, wear-resistant, low-smoke and non-dripping sheath material as well as a preparation method and application thereof. The tear-resistant wear-resistant low-smoke non-dripping sheath material is prepared from the following components in parts by weight: 25 to 35 parts of metallocene linear low-density polyethylene, 10 to 15 parts of ethylene-vinyl acetate copolymer, 6 to 8 parts of compatilizer, 50 to 60 parts of aluminum hydroxide, 10 to 15 parts of magnesium hydroxide, 3 to 4 parts of microencapsulated red phosphorus, 5 to 8 parts of lithium oxide synergist, 4 to 6 parts of silicone master batch, 0.3 to 0.7 part of carbon black master batch and 1.5 to 2.5 parts of coupling agent. The tear-resistant wear-resistant low-smoke non-dripping sheath material provided by the invention has the characteristics of being halogen-free, flame-retardant, tear-resistant and wear-resistant, and a low-smoke halogen-free flame-retardant power cable prepared from the sheath material has the characteristics of high smoke light transmittance during combustion, no dripping and cracking resistance.

The invention discloses a graded detection method and system for an air-breathing smoke-sensing fire in a complex environment, and relates to the technical field of fire-fighting early warning, and the method comprises the steps: recognizing a ventilation opening of a monitoring region in the complex environment, and arranging air-breathing smoke-sensing equipment on the ventilation opening; state monitoring is conducted on the ventilation opening, and working mode parameters of the ventilation opening are obtained and comprise the ventilation starting and stopping state, the wind speed and the ventilation direction; and analyzing to obtain a self-adaptive strategy matched with the working mode parameters to control the sucking pump to drive the air-breathing smoke sensing equipment to collect smoke gas samples, performing multi-time sequence grading detection, and outputting a fire early warning signal. According to the invention, the technical problem that the existing fire detection technology is difficult to accurately collect smoke samples and give out early warning in time in places with complex ventilation structures and obvious air disturbance is solved, and the air-breathing smoke sensing equipment is arranged at the ventilation opening and the sampling strategy is dynamically adjusted; and the accuracy of smoke collection and the real-time performance of fire early warning are improved.

The invention relates to the field of fire cooperative detection, in particular to a fire cooperative detection method and system based on multi-modal fusion, and the method comprises the steps: simulating the smoke diffusion and temperature change process of a fire in a machine room through a fire simulation method, determining the distribution region of smoke concentration and temperature rise, and constructing a smoke detector network and a temperature detector network; smog concentration data and temperature data of the detection points are collected and transmitted to a data processing center for data preprocessing, a comprehensive fire feature matrix is generated through a data fusion method, and the fire area of the fire is positioned through a cross analysis method according to the comprehensive fire feature matrix; the problem that after a machine room has a fire, due to smoke diffusion, the visualization degree is low, and the fire area cannot be accurately judged is solved.

The invention discloses a high-reliability fire detection method and system, and relates to the technical field of fire alarm, and the detection method comprises the following steps: S1, setting a threshold value; s2, initializing a background value, updating the detection value at regular time, and updating the background value to initialize the background value; s3, identifying the aerosol type; s4, identifying the movement condition of the personnel; s5, identifying a high-humidity environment; s6, identifying a high-temperature environment; and S7, judging a fire alarm. The detection system comprises a threshold setting module, an initialization module, a smoke detection module, a radar system, a humidity detection module, a temperature detection module, a background value updating module and an alarm module. According to the invention, the radar system is used for judging and detecting the motion state of environmental personnel, detection is carried out in combination with factors such as smoke, temperature and humidity, and different levels of alarm thresholds are set for judgment, so that the identification capability of the fire detector on non-fire smoke is enhanced, and the false alarm probability under the non-fire smoke condition is reduced.

The invention discloses a hydropower station unit fire alarm auxiliary decision-making management system, which relates to the technical field of power equipment safety and comprises a fire data acquisition module, a fire data analysis module, a fire alarm module, a linkage control module and a system management module. And the fire data acquisition module is used for acquiring fire related parameter data in a hydropower station unit area in real time, wherein the fire related parameter data comprises smoke concentration, temperature change rate, flame characteristics and carbon monoxide concentration. By adopting multiple types of sensor units, a smoke sensor, a temperature sensor, a flame sensor, a carbon monoxide sensor and the like are covered for combined acquisition, compared with the condition that the types of sensors of an existing system are limited, multi-dimensional fire related data can be comprehensively and accurately acquired, deep analysis is performed on the acquired data by means of an intelligent algorithm, and the accuracy of the system is improved. The historical data and the real-time data are compared, the false alarm probability is reduced, and the alarm accuracy is improved.

The invention discloses a flame smoke detection method and system based on dynamic convolution and multi-scale feature fusion, and belongs to the technical field of computer vision target detection, and the method comprises the steps: 1, obtaining a fire data set, carrying out the preprocessing of the fire data set, and dividing the fire data set into a training set and a test set; 2, constructing an improved YOLOv5s model: replacing a C3 module with C3-ODConv, introducing a GCB module into a backbone network, and adopting BiFPN on a neck; 3, training the improved YOLOv5s model by using the training set, and testing the improved YOLOv5s model by using the test set; and step 4, performing fire data detection by using the improved YOLOv5s model after training and testing. According to the invention, the learning and identification capability of flame and smoke characteristics can be enhanced, and the detection precision of the method in a complex scene is improved.

The invention discloses an automatic suction removal method, device and equipment for welding smoke and a medium, and relates to the technical field of machine learning. Linear interpolation and capacity expansion are conducted on a temperature matrix of a welding area; inputting the interpolated and expanded temperature matrix into a decision tree model, outputting predicted position information of the smoke source, and calculating spatial position coordinates; converting the point cloud data of the welding area into a grey-scale map, fusing the grey-scale map and the interpolated and expanded temperature matrix data to obtain a temperature space distribution matrix, inputting the temperature space distribution matrix into a first convolutional neural network, and outputting obstacle avoidance point cloud data; calculating parameters of the obstacle avoidance area, inputting the parameters, the spatial position coordinates and the initial joint rotation angle of the actuator into a second convolutional neural network, and outputting a target joint rotation angle; therefore, the welding smoke in the welding area can be automatically sucked by the actuator, the sucking efficiency is improved, the smoke concentration in the welding working area is reduced, the conflict condition of the smoke sucking disc working area is avoided, and the interference to the welding work and the sucking leakage rate are reduced.