An AI drill analysis method and system based on fire simulation escape

Abstract

The application relates to the technical field of fire escape drill analysis, and provides an AI drill analysis method and system based on fire simulation escape, which comprises the following steps: collecting images in the simulation escape through collection time periods divided according to fire data, obtaining first escape images of a fire condition unchanged stage and second escape images of a change stage, respectively identifying escape posture differences of each person in the images, judging whether the correct posture is consciously adopted for escape based on the relationship between the escape posture differences of each person and the corresponding fire data differences, giving a drill analysis conclusion of the fire simulation escape, improving the efficiency of the fire escape drill analysis, and providing protection for actual fire escape.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of fire escape drill analysis, and in particular to an AI drill analysis method and system based on fire simulation escape. BACKGROUND

[0002] People may lack a deep enough understanding of the danger of fire, and the main purpose of fire drill escape is to help people master correct escape knowledge and skills through simulation of real fire scenes, so as to improve survival probability, reduce casualties and losses in real fires;

[0003] The main reason for casualties in fire scenes is suffocation from inhaling toxic gases, so there should be strict requirements for the escape posture of personnel during escape, but the existing technology does not regulate the escape posture of escapees in fire simulation escape, for example, refer to the invention CN111008782A, which uses multiple sensors to detect the time of escape drill, the situation of escapees leaning against the wall and the touch height, ignoring the specific escape posture, and cannot effectively regulate the escape drill; The actual situation of the fire may affect the escape state of the escapee, for example, refer to the invention CN115660906A, although it collects data in various fire drills and collects data on the escapee's lowering his head, bending over and covering his mouth and nose, it ignores the influence of these postures in the fire environment, and escape without the influence of fire data is meaningless, and lowering the head, bending over and covering the mouth and nose are not suitable escape methods when the fire is small, which will affect the escape speed, but when the fire is large, these requirements should also change accordingly, so the existing fire drill escape analysis cannot be combined with the fire situation for analysis, affecting the analysis efficiency and making it difficult to guarantee the survival probability of the escape personnel in actual fires. SUMMARY

[0004] The present application provides an AI drill analysis method based on fire simulation escape, which is used to solve the problem of low efficiency of fire escape drill analysis in the prior art.

[0005] The present application provides an AI drill analysis method based on fire simulation escape, which is used to solve the problem of low efficiency of fire escape drill analysis in the prior art.

[0006] Obtain fire data of a fire simulation scene, construct a first image collection time period according to the stable time of the fire simulation in the fire data, and construct a second collection time period according to the change time of the fire simulation;

[0007] Obtain corresponding first escape images and second escape images based on the first collection time period and the second collection time period, identify the human escape postures of different personnel in the images using a preset AI model, calculate the escape posture differences, obtain the fire data differences, and construct the corresponding relationship between the fire data and the escape postures of each personnel;

[0008] determine whether the change relationship of the fire data of each person corresponds to the change of the escape posture meets a preset requirement, and output an analysis conclusion of the drill according to the difference relationship.

[0009] Optionally, after the fire data difference is obtained, the method further includes:

[0010] a first escape route is formed according to the position of the person in the first escape image, a second escape route is formed according to the position of the person in the second escape image, and a difference between the first escape route and the second escape route is identified;

[0011] Before the analysis conclusion of the drill is output according to the difference relationship, the method further includes: calculating distances between a difference route part of the first escape route and the second escape route and a smoke and fire point position of the fire simulation, respectively, and determining that there is a problem in the selection of the escape route when a variance of the distances between the difference route part and the smoke and fire point position is greater than a preset threshold.

[0012] Optionally, after the variance of the distances between the difference route part and the smoke and fire point position is greater than the preset threshold, the method further includes:

[0013] When the variance of the distances between the difference route part and the smoke and fire point position is less than the preset threshold, average escape speeds at each place in the first escape route and the second escape route are calculated, respectively, it is determined whether the escape speed of the difference route part that meets the variance less than the preset threshold is greater than the escape speed of a non-difference route part, and if not, it is determined that there is a deficiency in the escape.

[0014] Optionally, after the escape posture difference is calculated, the method further includes: obtaining physiological data of the person based on the first collection time period and the second collection time period to obtain fluctuation of the physiological data;

[0015] After the analysis conclusion of the drill is output according to the difference relationship, the method further includes: correlating the fluctuation of the physiological data with an abnormal escape posture and an abnormal escape route, and constructing an escape behavior prediction model of different persons, and performing targeted escape route correction based on the escape behavior prediction model of each escape person.

[0016] The second aspect of the present application provides an AI drill analysis system based on fire simulation escape, including:

[0017] a time division module configured to obtain fire data of a fire simulation scene, construct a first collection time period of an image according to a stable time of the fire simulation in the fire data, and construct a second collection time period according to a change time of the fire simulation;

[0018] The data processing module is configured to acquire corresponding first escape images and second escape images based on the first acquisition time period and the second acquisition time period, identify human body escape postures of different persons in the images by using a preset AI model, calculate escape posture differences, acquire fire data differences, and construct a corresponding relationship between fire data and escape postures of each person.

[0019] The drill analysis module is configured to determine whether a change relationship between fire data changes and escape posture changes of each person meets a preset requirement, and output a drill analysis conclusion according to the difference relationship.

[0020] Optionally, in the data processing module, after the fire data differences are acquired, the method further includes:

[0021] The first escape route is formed according to the positions of the persons in the first escape images, the second escape route is formed according to the positions of the persons in the second escape images, and the differences between the first escape route and the second escape route are identified.

[0022] In the drill analysis module, before the drill analysis conclusion is output according to the difference relationship, the method further includes: calculating distances between difference route parts of the first escape route and the second escape route and a smoke and fire point position of the fire simulation, respectively, and determining that there is a problem in escape route selection when a variance of the distances between the difference route parts and the smoke and fire point position is greater than a preset threshold.

[0023] Optionally, in the drill analysis module, after the variance of the distances between the difference route parts and the smoke and fire point position is greater than the preset threshold, the method further includes:

[0024] When the variance of the distances between the difference route parts and the smoke and fire point position is less than the preset threshold, average escape speeds of each part in the first escape route and the second escape route are calculated, respectively, it is determined whether the escape speed of the difference route part whose variance is less than the preset threshold is greater than the escape speed of a non-difference route part, and if not, it is determined that there is a deficiency in escape.

[0025] Optionally, in the data processing module, after the escape posture differences are calculated, the method further includes: acquiring physiological data of the persons based on the first acquisition time period and the second acquisition time period, and obtaining fluctuation of the physiological data.

[0026] In the drill analysis module, after the drill analysis conclusion is output according to the difference relationship, the method further includes: correlating the fluctuation of the physiological data with abnormal escape postures and abnormal escape routes, and constructing escape behavior prediction models of different persons, and performing targeted escape route correction based on the escape behavior prediction models of each escape person.

[0027] The third aspect of the present application provides an AI drill analysis method and device based on fire simulation escape, the device includes a processor and a memory:

[0028] The memory is configured to store program code and transmit the program code to the processor.

[0029] The processor is configured to execute the method according to the instructions in the program code.

[0030] The fourth aspect of the present application provides a computer readable storage medium configured to store program code, the program code being configured to execute the method according to any one of the first aspect of the present application.

[0031] From the above technical solutions, the present application has the following advantages: by collecting images in the simulated escape according to the collected time period of the fire data division, the first escape image in the unchanged stage of the fire situation and the second escape image in the changed stage are obtained, and the escape posture differences of each person are identified respectively, whether the correct posture is consciously adopted in the escape is judged based on the relationship between the escape posture differences of each person and the corresponding fire data differences, the analysis conclusion of the escape drill is given, and the efficiency of the fire escape drill analysis is improved, and the actual fire escape is ensured. BRIEF DESCRIPTION OF DRAWINGS

[0032] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained without creative labor.

[0033] Figure 1 The flowchart of the method for AI drill analysis based on fire simulation escape.

[0034] Figure 2 The structure diagram of the system for AI drill analysis based on fire simulation escape. DETAILED DESCRIPTION

[0035] In order to make the purposes, features and advantages of the present application more obvious and easy to understand, the technical solutions in the embodiments of the present application will be described clearly and completely in combination with the drawings in the embodiments of the present application. Obviously, the following described embodiments are only some embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0036] The application provides an AI drill analysis method based on fire simulation escape, and aims to solve the problem of low efficiency of fire escape drill analysis in the prior art.

[0037] Please refer to Figure 1 , Figure 1 The first flowchart of the AI drill analysis method based on fire simulation escape provided by the embodiment of the application.

[0038] S100, obtain fire data of a fire simulation scene, construct a first image collection time period according to a stable time of fire simulation in the fire data, and construct a second collection time period according to a change time of the fire simulation;

[0039] It should be noted that the fire simulation is realized by simulating flames and smoke in the site, the smoke is simulated by setting a smoke machine, and the flame is simulated by setting a fake fire, the fire data includes flame simulation point positions and smoke simulation point positions; the smoke machine and the fake fire can be controlled according to instructions, the smoke density, smoke diffusion and fire size are controlled based on preset programs, and the specific time periods of smoke and flame control are also included in the fire data; the programs have corresponding times and specific smoke and flame change data, when the smoke and flame do not change with program adjustment, the time period is set as the stable time of the fire simulation to construct the first collection time period; and when the smoke machine and the fake fire change with program control, the time period is set as the change time of the fire simulation to construct the second collection time period.

[0040] S200, obtain corresponding first escape images and second escape images based on the first collection time period and the second collection time period, identify human body escape postures of different persons in the images by using a preset AI model, calculate escape posture differences, obtain fire data differences, and construct a corresponding relationship between the fire data and the escape postures of each person;

[0041] It should be noted that cameras will be set near the smoke machine and the fake fire in the fire simulation site, and image shooting will be carried out in the scene of fire simulation in the first and second image acquisition time periods to obtain the first and second escape images, the first escape image reflects the escape state of personnel when the simulated fire condition is stable, and the second escape image reflects the escape state of personnel when the simulated fire condition is unstable; in the image, the human body dynamic posture is first identified by AI model through computer vision and sensor technology, and the core technical route includes two ways of visual analysis based on image video and motion capture based on MEMS sensor. The basic module of the technology is skeleton key point detection, the three-dimensional coordinates of 20-24 key points of the human body are extracted through convolutional neural network, the skeleton model is constructed to realize action analysis, the human body escape posture in each escape image is obtained, and the specific contents include the bending, bowing and crawling degree of personnel, for example, the angle between the two connecting lines constructed by connecting the key points of hip, spine and chest according to the corresponding three-dimensional coordinates is the bending angle, the angle between the two connecting lines constructed by connecting the key points of head, neck and shoulder according to the corresponding three-dimensional coordinates is the neck bending angle, the human face recognition of the escape personnel can also be carried out in the obtained escape image, the personnel identity is determined, the human body escape posture in the escape process is associated with the corresponding personnel, the escape posture of the same escape personnel in different escape images is calculated after face recognition, because the escape posture of personnel will be different in different fire simulation states, the difference value of the bending angle and the neck bending angle can be obtained according to the posture difference, and the difference of the escape posture of each escape personnel in different stages of fire can be calculated;

[0042] The intelligent fire and smoke training device for firefighters can simulate flames and smoke in the field. The parameters during the operation of the device are recorded according to time. The intelligent heat and smoke training device is a device for simulating flame heating and smoke generation. According to the first collection time period and the second collection time period, the corresponding fire data, specifically the size of the false fire or the smoke concentration, can be obtained. The difference in fire data between the first escape image and the second escape image can be calculated. During the escape process of a fire drill, the flame and smoke are controlled to change the simulated fire situation. Therefore, the same escape personnel may experience different smoke concentrations and fire sizes during the escape process. However, the escape personnel generally change the degree of bending and the degree of neck bending while maintaining the original escape posture and moving quickly, considering the fire situation. The fire data difference value and the escape posture difference value of the same escape personnel are compared according to the corresponding time. The correlation change relationship is identified. The positive and negative values of the non-dimensionalized fire data difference value correspond to whether the fire situation corresponding to the first escape image and the second escape image is larger or smaller. That is, when the fire data difference value is positive, the fire size and smoke concentration in the second escape image are larger than those in the first escape image. The positive and negative values of the non-dimensionalized escape posture difference value correspond to whether the escape posture corresponding to the first escape image and the second escape image is more curled or more stretched. That is, when the escape posture difference value is positive, the escape personnel posture in the second escape image is smaller than the bending and neck bending angle in the first escape image. The two values are directly divided after non-dimensionalization to identify the positive correlation and negative correlation. When the division is negative, it is a negative correlation. When the fire situation becomes larger, the bending and neck bending angle of the escape personnel becomes smaller, avoiding inhaling smoke and reducing the area of heat radiation. Or when the fire situation becomes smaller, the bending and neck bending angle of the escape personnel becomes larger, improving the moving speed to quickly escape, reflecting that the escape posture of each escape personnel consciously adjusts when encountering changes in the fire situation. When the division is positive, it indicates that the judgment of the escape personnel has errors, which is not conducive to escape.

[0043] Multiple escape images can be collected in the same time period. After identifying the escape posture angle of the same escape personnel in the images, the average value is calculated to obtain the escape posture corresponding to the time period. In the fire simulation escape, there are multiple first collection time periods and multiple second collection time periods. The difference between each two can be calculated, and then the average value is taken as the final corresponding relationship between the fire data and the escape posture.

[0044] S300, determining whether the change relationship between the fire data of each personnel and the escape posture change meets the preset requirements, and outputting a drill analysis conclusion according to the difference relationship.

[0045] It should be noted that when the fire and smoke change in the fire, the escape posture should be changed accordingly, for example, when the smoke concentration increases, the lowest height of the air suspension decreases, and the escape personnel should advance in the escape path at a lower posture to avoid inhaling smoke through the mouth and nose; and when the fire increases, the personnel needs to bend down to block the flame radiation;

[0046] The correlation between the fire data of each personnel constructed in the foregoing step and the escape posture is positive when the value obtained by dividing the difference in the escape posture by the difference in the fire data is positive, that is, the bending angle and the lowering angle of the posture increase when the fire increases and the smoke concentration increases, and the correlation is negative when the value is negative, that is, the bending angle and the lowering angle of the posture decrease when the fire increases and the smoke concentration increases; in the fire drill process, the escape personnel may look up when hearing the explosion sound caused by the increase of the fire or the change in the smoke concentration, or may look around when not paying attention to the drill, therefore, only when the correlation is positive, the preset requirement is met, and the escape drill is considered to be compliant, and when the correlation is negative, it is indicated that the escape personnel does not adjust the posture correctly during the escape process according to the fire condition, in this case, an analysis conclusion that the escape personnel does not respond correctly to the fire can be given for the escape drill, and further, when the absolute value of the division result of the correlation in the foregoing step is less than a preset threshold, it is indicated that the escape personnel does not respond to the change in the fire, that is, the posture is not changed according to the change in the fire or the smoke, and the escape personnel needs to pay attention to this in the subsequent drill or actual fire, and an analysis conclusion for the drill is obtained for each personnel.

[0047] In this embodiment, the images in the simulated escape are collected by dividing the collection time period of the fire data, the first escape image in the stage of no change in the fire condition and the second escape image in the stage of change are obtained, and the difference in the escape posture of each personnel is identified, whether the correct posture is consciously used for escape is judged based on the correlation between the difference in the escape posture of each personnel and the corresponding fire data, an analysis conclusion for the escape drill is given, and the efficiency of the analysis of the fire escape drill is improved, which provides protection for the actual fire escape.

[0048] The foregoing is a detailed description of a first embodiment of an AI drill analysis method based on fire simulated escape provided by the present application, and the following is a detailed description of a second embodiment of an AI drill analysis method based on fire simulated escape provided by the present application.

[0049] In this embodiment, an AI drill analysis method based on fire simulated escape is further provided, and after the fire data difference is obtained in the foregoing step S200, the method further includes: constructing a first escape route according to the position of the personnel in the first escape image; constructing a second escape route according to the position of the personnel in the second escape image, and identifying the difference between the first escape route and the second escape route;

[0050] Before the step of outputting the analysis conclusion according to the difference relationship in the foregoing step S300, the method further includes: calculating distances between the difference route part of the first escape route and the second escape route and the smoke and fire point position of the fire simulation respectively, and determining that there is a problem in the selection of the escape route when a variance of the distances between the difference route part and the smoke and fire point position is greater than a preset threshold.

[0051] It should be noted that multiple images can be taken in the first collection time period and the second collection time period during the escape process, and the shooting angle can be a top view of the escape passage or parallel to the escape direction, etc. The position of the personnel in the escape passage can be identified in the image, and the positions of the multiple escape personnel in the multiple escape images can constitute an escape route. The route constituted in the first escape image is the first escape route, and the route constituted in the second escape image is the second escape route. Multiple personnel will take similar routes during the escape process, and the route will avoid the positions of the fire source and the smoke source. When the fire and the smoke density increase, the personnel need to further move away from the fire source, and the position of the route changes. By identifying the difference between the first escape route and the second escape route, the part of the escape route that may change due to the fire and smoke can be found first, and then multiple discrete points are selected in the difference part of the first escape route and the second escape route. The distances to the smoke and fire point position included in the fire data are calculated, and the variance of the distance values is calculated. When the variance of the distances between the difference route part and the smoke and fire point position is greater than the preset threshold, it indicates that the difference part is not around the smoke and fire point position or the personnel cannot correctly adjust the route according to the specific fire scene. Further, for the route difference part that meets the variance condition, the fire difference data when the route is taken can be identified to determine whether the difference degree of the route matches the difference degree of the fire, such as whether the degree of increase in the fire matches the degree of difference in the personnel escape route.

[0052] Further, after the step of determining that the variance of the distances between the difference route part and the smoke and fire point position is greater than the preset threshold in the foregoing step, the method further includes: when the variance of the distances between the difference route part and the smoke and fire point position is less than the preset threshold, calculating the average escape speed of each part of the first escape route and the second escape route respectively, determining whether the escape speed of the difference route part that meets the variance less than the preset threshold is greater than the escape speed of the non-difference route part, and determining that there is a deficiency in the escape if not.

[0053] It should be noted that the escape speed of the personnel on the escape route can be calculated based on the position change of the same personnel after face recognition in the escape image and the shooting time difference of the escape image, and there can be different escape speeds of different escape personnel, and then the average escape speed is calculated; after the part of the difference route part and the smoke point position distance in the foregoing step is identified to be less than a preset threshold, the average speed of the difference route part and the average speed of the non-difference route part can be compared, because when passing through the fire source or smoke source position in the fire, it is necessary to pass quickly to avoid being burned and other influences, therefore, when passing through the fire source or smoke source, the difference route part needs to use a faster speed than the non-difference route part of the normal escape passage, when the escape speed of the difference route part is not faster than the escape speed of the non-difference route part, it indicates that the escape does not take the strategy of quickly moving away, and the simulation escape is insufficient.

[0054] Further, after the escape posture difference is calculated, it further includes: obtaining physiological data of the personnel based on the first acquisition time period and the second acquisition time period, and obtaining the fluctuation of the physiological data;

[0055] After the simulation analysis conclusion is output according to the difference relationship, it further includes: correlating the fluctuation of the physiological data with the abnormal escape posture and the abnormal escape route, and constructing an escape behavior prediction model of different personnel, and performing targeted escape route correction based on the escape behavior prediction model of each escape personnel.

[0056] It should be noted that sensors can be provided on the escape personnel during the escape drill to obtain physiological data of the human body, such as heart rate data, to determine the real-time emotional state of the escape personnel, for example, when the smoke concentration of the fire changes, the heart rate of the escape personnel increases significantly, which can be considered as emotional tension of the escape personnel, and emotional tension will lead to escape failure during the escape process, therefore, it is necessary to identify the physiological conditions corresponding to the abnormal escape posture and the abnormal escape route to determine whether the reason is tension; in the case that the positions and passages of the escape personnel in some fixed places such as school seats or company workstations are known, the escape routes of each personnel can be pre-set, and in the foregoing step, it can be obtained that some personnel will appear abnormal escape posture and abnormal escape route due to emotional tension, therefore, further route planning is needed for these personnel in the escape assistance system, a neural network model for human behavior prediction can be used to simulate the escape process of the personnel, record the escape time and path selection, for example, a convolutional neural network is used to predict the posture and route of the escape personnel, after the physiological characteristics in the foregoing step and the corresponding data of the abnormal escape posture and the abnormal escape route are input into the prediction model, the AI model is further adjusted and optimized to make an emergency plan or an escape route, so as to avoid that some escape personnel affect the overall escape efficiency due to emotional fluctuations; the escape route can avoid areas where special situations are prone to occur according to the physiological reflection characteristics of each escape personnel, affect the escape probability, and optimize the escape scheme, the escape assistance AI system uses an improved A* algorithm and a dynamic obstacle avoidance technology, adjusts based on the degree of influence and the possibility of being influenced based on the actual psychological situation, provides a personalized escape path, and dynamically adjusts based on real-time environmental data such as thermal radiation intensity and CO concentration.

[0057] The above is a detailed description of an AI drill analysis method based on fire simulation escape provided by the first aspect of the present application, and the following is a detailed description of an embodiment of an AI drill analysis system based on fire simulation escape provided by the second aspect of the present application.

[0058] Please refer to Figure 2 , Figure 2 It is an AI drill analysis system structure diagram based on fire simulation escape. The embodiment provides an AI drill analysis system based on fire simulation escape, which comprises:

[0059] The time division module 10 is used for obtaining fire data of the fire simulation scene, constructing an image first collection time period according to the stable time of the fire simulation in the fire data, and constructing a second collection time period according to the change time of the fire simulation;

[0060] The data processing module 20 is configured to acquire the first escape image and the second escape image based on the first acquisition time period and the second acquisition time period, identify human body escape postures of different persons in the images by using a preset AI model, calculate an escape posture difference, acquire a fire data difference, and construct a corresponding relationship between the fire data and the escape postures of each person.

[0061] The drill analysis module 30 is configured to determine whether the change relationship between the fire data and the escape postures of each person meets a preset requirement, and output a drill analysis conclusion according to the difference relationship.

[0062] Further, the data processing module 20 further includes the following steps after acquiring the fire data difference:

[0063] The first escape route is formed according to the positions of the persons in the first escape image, the second escape route is formed according to the positions of the persons in the second escape image, and the difference between the first escape route and the second escape route is identified.

[0064] In the drill analysis module 30, before outputting the drill analysis conclusion according to the difference relationship, the following steps are further included: the distance between the difference route part of the first escape route and the second escape route and the smoke and fire point position of the fire simulation is calculated respectively, and when the variance of the distance between the difference route part and the smoke and fire point position is greater than a preset threshold, it is determined that there is a problem in the escape route selection.

[0065] Further, in the drill analysis module 30, after the variance of the distance between the difference route part and the smoke and fire point position is greater than the preset threshold, the following steps are further included:

[0066] When the variance of the distance between the difference route part and the smoke and fire point position is less than the preset threshold, the average escape speed of each part in the first escape route and the second escape route is calculated respectively, it is determined whether the escape speed of the difference route part meeting the condition that the variance is less than the preset threshold is greater than the escape speed of the non-difference route part, and if not, it is determined that there is a deficiency in the escape.

[0067] Further, in the data processing module 20, after calculating the escape posture difference, the following steps are further included: physiological data of the persons is acquired based on the first acquisition time period and the second acquisition time period, and the fluctuation of the physiological data is obtained.

[0068] In the drill analysis module 30, after outputting the drill analysis conclusion according to the difference relationship, the following steps are further included: the physiological data fluctuation is associated with the abnormal escape posture and the abnormal escape route, and an escape behavior prediction model of different persons is constructed, and the escape route is corrected based on the escape behavior prediction model of each escape person.

[0069] The third aspect of the application further provides an AI drill analysis method based on fire simulation escape, comprising a processor and a memory: the memory is used to store program code and transmit the program code to the processor; the processor is used to execute the AI drill analysis method based on fire simulation escape according to the instructions in the program code.

[0070] The fourth aspect of the application provides a computer readable storage medium, characterized in that the computer readable storage medium is used to store program code, and the program code is used to execute the AI drill analysis method based on fire simulation escape.

[0071] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described device and equipment can refer to the corresponding process in the foregoing method embodiments, which will not be repeated here.

[0072] In several embodiments provided in the application, it should be understood that the disclosed system, device and method can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the units is only a logical function division, and actual implementation can have another division manner, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some interface, device or unit, which can be electrical, mechanical or other forms.

[0073] The units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, that is, they can be located in one place, or can be distributed on a plurality of network units. According to actual needs, part or all of the units can be selected to achieve the purpose of the embodiment scheme.

[0074] In addition, each functional unit in each embodiment of the application can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of software functional unit.

[0075] The integrated unit, if implemented in the form of a software function unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application or the entire or part of the technical solutions that essentially contribute to the prior art can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in the various embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0076] The above-described embodiments are only used to illustrate the technical solutions of the present application, rather than limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacements for some technical features; and these modifications or replacements do not make the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. An AI drill analysis method based on fire simulation escape, characterized in that The method comprises the following steps: acquiring fire data of a fire simulation scene, constructing a first image acquisition time period according to a stable time of the fire simulation in the fire data, and constructing a second acquisition time period according to a change time of the fire simulation; based on the first acquisition time period and the second acquisition time period, acquiring corresponding first escape images and second escape images, identifying human body escape postures of different persons in the images by using a preset AI model, calculating escape posture differences, specifically, performing face recognition on escape persons in the escape images, associating the human body escape postures with the persons, and calculating the escape posture differences of the same escape person in different escape images, the difference calculation including a difference value of a waist and neck bending angle; acquiring fire data differences, specifically, calculating a difference value of fire intensity or smoke concentration between the first escape images and the second escape images, and constructing a corresponding relationship between the fire data and the escape postures of each person, specifically, comparing the fire data difference values and the escape posture difference values of the same escape person according to the time to which they belong, and identifying an associated change relationship therebetween; judging whether the change relationship between the fire data of each person and the escape posture conforms to a preset requirement, and outputting an analysis conclusion of the drill according to the difference relationship.

2. The AI drill analysis method based on fire simulation evacuation according to claim 1, wherein, After the fire data differences are acquired, the method further comprises the following steps: constructing a first escape route according to the positions of the persons in the first escape images, constructing a second escape route according to the positions of the persons in the second escape images, and identifying a difference between the first escape route and the second escape route; Before the analysis conclusion of the drill is output according to the difference relationship, the method further comprises the following steps: calculating distances between a difference route part of the first escape route and the second escape route and a smoke and fire point position of the fire simulation, respectively, and judging that there is a problem in the escape route selection when a variance of the distances between the difference route part and the smoke and fire point position is greater than a preset threshold value. 3.The AI drill analysis method based on fire simulation evacuation of claim 2, wherein, After the variance of the distances between the difference route part and the smoke and fire point position is greater than the preset threshold value, the method further comprises the following steps: when the variance of the distances between the difference route part and the smoke and fire point position is less than the preset threshold value, calculating average escape speeds of each part in the first escape route and the second escape route, respectively, judging whether the escape speed of the difference route part that satisfies the variance less than the preset threshold value is greater than the escape speed of a non-difference route part, and judging that there is a deficiency in the escape when the escape speed of the difference route part is not greater than the escape speed of the non-difference route part.

4. The AI drill analysis method based on fire simulation evacuation according to claim 3, wherein, After the escape posture differences are calculated, the method further comprises the following step: acquiring physiological data of the persons based on the first acquisition time period and the second acquisition time period, and obtaining fluctuation of the physiological data; After the analysis conclusion of the drill is output according to the difference relationship, the method further comprises the following steps: associating the fluctuation of the physiological data with abnormal escape postures and abnormal escape routes, and constructing escape behavior prediction models of different persons, and performing targeted escape route correction based on the escape behavior prediction models of the escape persons.

5. An AI drill analysis system based on fire simulation evacuation, characterized by, The method comprises the following steps: a time division module is configured to acquire fire data of a fire simulation scene, construct a first image acquisition time period according to a stable time of the fire simulation in the fire data, and construct a second acquisition time period according to a change time of the fire simulation; The data processing module is configured to obtain corresponding first escape images and second escape images based on the first acquisition time period and the second acquisition time period, identify human body escape postures of different persons in the images by using a preset AI model, calculate escape posture differences, specifically, perform face recognition on escape persons in the escape images, associate the human body escape postures with the persons, and calculate differences in escape postures of the same escape person in different escape images, wherein the difference calculation includes a difference in a waist and neck bending angle; obtain fire data differences, specifically, calculate a difference in fire intensity or smoke concentration between the first escape images and the second escape images, and construct a corresponding relationship between fire data and escape postures of each person, specifically, compare the fire data difference values and the escape posture difference values of the same escape person according to the time to which the values belong, and identify an associated change relationship therefrom. The drill analysis module is configured to determine whether a change relationship between fire data changes and escape posture changes of each person meets a preset requirement, and output a drill analysis conclusion according to the difference relationship.

6. The AI drill analysis system based on fire simulation evacuation of claim 5, wherein, In the data processing module, after obtaining the fire data differences, the following steps are further included: A first escape route is constructed according to the positions of the persons in the first escape images, a second escape route is constructed according to the positions of the persons in the second escape images, and a difference between the first escape route and the second escape route is identified. In the drill analysis module, before outputting the drill analysis conclusion according to the difference relationship, the following steps are further included: distances between difference route portions of the first escape route and the second escape route and a smoke and fire point position of the fire simulation are calculated respectively, and when a variance of the distances between the difference route portions and the smoke and fire point position is greater than a preset threshold, it is determined that there is a problem with escape route selection.

7. The AI drill analysis system based on fire simulation evacuation of claim 6, wherein, In the drill analysis module, after the variance of the distances between the difference route portions and the smoke and fire point position is greater than the preset threshold, the following steps are further included: When the variance of the distances between the difference route portions and the smoke and fire point position is less than the preset threshold, average escape speeds at different positions in the first escape route and the second escape route are calculated respectively, it is determined whether the escape speed of the difference route portion that meets the variance less than the preset threshold is greater than the escape speed of a non-difference route portion, and if not, it is determined that there is a deficiency in escape.

8. The AI drill analysis system based on fire simulation evacuation of claim 7, wherein, In the data processing module, after calculating the escape posture differences, the following steps are further included: physiological data of the persons are obtained based on the first acquisition time period and the second acquisition time period, and fluctuation of the physiological data is obtained. In the drill analysis module, after outputting the drill analysis conclusion according to the difference relationship, the following steps are further included: the physiological data fluctuation is associated with abnormal escape postures and abnormal escape routes, and escape behavior prediction models of different persons are constructed, and the escape behavior prediction models of the escape persons are used to make targeted escape route corrections. 9.A device for analyzing AI drill based on fire simulation evacuation, characterized by, The device includes a processor and a memory: The memory is configured to store program code and transmit the program code to the processor; The processor is configured to execute the instructions in the program code to perform the AI drill analysis method based on fire simulation escape of any one of claims 1-4.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium is configured to store program code for performing the AI analysis method for fire simulation escape based on the method according to any one of claims 1-4.

Images