Security and protection equipment and monitoring method

Abstract

The invention relates to the technical field of security and protection monitoring, in particular to security and protection equipment and a monitoring method, and the method comprises the steps: obtaining the historical data of a disaster, carrying out the preprocessing of the data, carrying out the simulation, obtaining an ideal escape path, generating a data set training model through training data and the corresponding ideal escape path data, and carrying out the simulation. The verification data are input into the training models to verify the response speed and decision accuracy of the models, the training models with qualified decision accuracy are arranged in the security and protection equipment, a plurality of escape paths are automatically generated based on the data collected by the camera according to the disaster situation, and an instruction is sent to the signal receiver based on the single escape path; and the signal receiver performs feedback based on the received instruction. The security and protection monitoring system provided by the invention autonomously judges and gives a coping scheme to provide timely and accurate escape guidance for personnel evacuation in the case of a disaster, so that serious casualties caused by untimely personnel evacuation are avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of security monitoring, and in particular to a security device and a monitoring method. Background Art

[0002] Disasters such as fires and earthquakes often occur suddenly, leaving people with extremely limited time to escape. Being able to quickly plan a reasonable escape route means that people can be guided to evacuate dangerous areas more quickly, reducing casualties. The accuracy of the escape route cannot be ignored either. An accurate escape route can ensure that people do not get lost during the evacuation process, avoid entering dangerous areas, and reach a safe area safely. If the escape route is not planned accurately, people may encounter obstacles during the evacuation process, or even mistakenly enter a fire source or other dangerous areas, increasing the risk of casualties.

[0003] Smart security equipment integrates advanced sensor technology, IoT communications, big data analysis and artificial intelligence algorithms. When disasters occur, such as fires and earthquakes, the system can respond quickly and monitor environmental changes in real time through various sensors deployed in buildings (such as smoke detectors, temperature sensors, earthquake monitors, etc.). Once a disaster signal is received, the system's built-in algorithm will immediately start to analyze and process the collected data. Taking into account multiple factors such as building structure, personnel distribution, and safe exit locations, the system can automatically plan multiple optimal escape routes in a very short time, and dynamically adjust according to the development of the disaster to avoid dangerous areas, ensuring that personnel can evacuate quickly along the shortest and safest path, providing strong protection for the evacuation of people in disaster situations.

[0004] Chinese Patent Publication No.: CN114584746B discloses a security monitoring system and a security monitoring method. The security monitoring system in the application includes: security monitoring equipment, a server and a mobile terminal. The camera device is set at a preset position in the park for security monitoring to shoot the monitoring area of ​​the preset position in the park to obtain video data; the DSP is used to obtain the video data from the camera device and cache the video data; the camera device sends the video data to the server via the WIFI module; the WIFI module is used to receive the wake-up request sent by the server to notify the main control module to turn on the main power circuit; the main control module is also used to turn off the main power circuit after detecting that the video data has been sent. This application embodiment solves the problem that the installation of current wired security monitoring equipment is inconvenient and has low installation efficiency.

[0005] It can be seen that the security monitoring system and security monitoring method have the following problems: although the invention can solve the problem of inconvenient installation and low installation efficiency of wired security monitoring equipment in the prior art, the monitoring system in the invention cannot independently judge and give practical response plans when abnormal situations occur, resulting in serious casualties due to untimely evacuation of personnel. Summary of the invention

[0006] To this end, the present invention provides a security device and a monitoring method to overcome the problem in the prior art that the monitoring system is unable to independently judge and provide a response plan when facing a disaster, thereby causing serious casualties due to untimely evacuation of personnel.

[0007] To achieve the above object, on the one hand, the present invention provides a security equipment monitoring method, comprising: Acquire historical data of several disaster situations based on a network platform, pre-process the historical data of any disaster situation to obtain several groups of data to be simulated, and split any group of data to be simulated into training data and verification data at a first preset ratio, wherein the historical data includes the type of disaster situation, the diffusion rate, the building information of the occurrence location, the environmental information, and the information of the trapped persons; Performing escape simulation analysis based on the data to be simulated to determine a number of ideal escape paths, combining the data to be simulated and the ideal escape paths corresponding to the training data into a number of data sets, and inputting the data sets into a neural network model to train the model; Input the verification data into the training model to output a number of actual escape paths, determine the path planning time, determine the eligibility of the response speed of the training model based on the path planning time, so as to increase the removal ratio of duplicate data, or, for a single group of the data to be simulated, determine whether the single decision accuracy of the training model is qualified based on the path overlap; Counting the percentage of qualified single decision accuracy, determining whether the decision accuracy of the training model is qualified based on the percentage, and determining the transformation method of the training data according to the variance of path overlap based on the training model with unqualified decision accuracy; A training model with qualified decision accuracy is set in the security equipment, and several escape routes are generated based on the data collected by the camera according to the disaster situation. Instructions are sent to the signal receiver based on the single escape path. The signal receiver provides feedback based on the received instructions to guide the crowd to evacuate along the escape path.

[0008] Furthermore, the process of preprocessing the historical data includes: Cleaning the historical data to eliminate noise, remove missing values ​​and outliers to obtain cleaned data; Normalizing the cleaned data to extract feature data; Eliminating duplicate data in the characteristic data at a second preset ratio to obtain data to be simulated; Cluster analysis is performed based on the data to be simulated, and the same type of data to be simulated is grouped into one group of data to be simulated to obtain several groups of data to be simulated.

[0009] Further, the process of determining whether the response speed of the training model is qualified based on the path planning time includes: Comparing the path planning time with a preset planning time; Based on the comparison result that the path planning time is greater than the preset planning time, it is determined that the response speed of the training model is unqualified.

[0010] Further, the process of increasing the removal ratio of duplicate data based on the condition that the response speed is unqualified includes: Subtracting the path planning time from the preset planning time to obtain a time difference; Comparing the time difference with a preset time difference; Determining to use a first ratio adjustment coefficient to adjust the second preset ratio based on a comparison result that the time difference is greater than the preset time difference; Based on the comparison result that the time difference is less than or equal to the preset time difference, it is determined to use the second ratio adjustment coefficient to adjust the second preset ratio.

[0011] Further, the process of determining the path overlap based on the comparison between the ideal escape path and the actual escape path includes: The ideal escape path and the actual escape path that are completely consistent are recorded as coincident paths; Counting the number of overlaps of the overlapped paths; The ratio of the number of overlaps to the number of actual escape paths is determined as the path overlap degree.

[0012] Furthermore, the process of determining whether the accuracy of a single decision is qualified based on the path overlap includes: Comparing the path overlap with a preset overlap; Based on the comparison result that the path overlap is less than the preset overlap, it is determined that the single decision accuracy is unqualified.

[0013] Furthermore, the process of determining whether the decision accuracy of the training model is qualified based on the quantity proportion includes: Comparing the quantity ratio with a preset quantity ratio; Based on the comparison result that the quantity proportion is less than the preset quantity proportion, it is determined that the decision accuracy of the training model is unqualified.

[0014] Furthermore, the process of determining the transformation mode of the training data according to the path overlap variance based on the training model with unqualified decision accuracy includes: Calculating the path overlap variance based on the path overlap based on the training model with unqualified accuracy; Comparing the path overlap variance with a preset variance; Determining to increase the amount of the training data based on a comparison result that the path overlap variance is greater than the preset variance; Based on the comparison result that the path overlap variance is less than or equal to the preset variance, it is determined that the abnormal data in the historical data is corrected.

[0015] Furthermore, a training model with qualified decision accuracy is set in the security equipment, and several escape routes are generated based on the data collected by the camera according to the disaster situation. The escape routes are updated in real time based on changes in the disaster situation. Different instructions are sent to signal receivers for different escape paths. After receiving the instructions, the signal receivers provide different feedback to guide the crowd to evacuate along a single escape path.

[0016] On the other hand, the present invention also provides a security device, comprising: Data collection components, including cameras, temperature sensors, and smoke sensors, to obtain information from the disaster site; a signal receiver for receiving instructions to indicate an escape route; An escape control center is connected to the camera and the signal receiver respectively, and is used for planning an escape route in real time based on the data collected by the camera, and sending instructions to the signal receiver on a single escape route.

[0017] Compared with the prior art, the beneficial effect of the present invention lies in that, in order to address the situation in which casualties occur due to failure to respond in time when a disaster occurs, the present invention obtains data on the occurrence of a disaster on a network platform, pre-processes these data, and then simulates them based on the data to accurately reproduce the environment and conditions of the disaster site, including building structure, topography, transportation network, and personnel distribution, so as to determine the ideal escape path with the highest efficiency in evacuating people, and inputs these data and the corresponding ideal escape path into a neural network model to train the model, and adjusts the model training data based on the response speed and decision accuracy of the model to improve the accuracy of the training model, so that the trained model can quickly generate an escape path based on the on-site situation when a disaster occurs, and provide timely and accurate escape guidance for the evacuation of people.

[0018] Furthermore, the present invention improves the overall quality of historical data by preprocessing the historical data, so that the model can learn more accurate information during the training process, improve the performance stability of the model on different data sets, reduce the model performance fluctuations caused by data differences, improve the decision-making accuracy of the training model, reduce the complexity of the model, and reduce the risk of overfitting, so that the trained model can quickly generate an escape route according to the on-site conditions when a disaster occurs, and provide timely and accurate escape guidance for personnel evacuation.

[0019] Furthermore, the present invention determines whether the response speed of the model is qualified through the path planning time. When facing a disaster, the time left for people to escape is extremely limited. Determining the escape route in the shortest time can maximize personal safety. Therefore, the response speed of the model is extremely important. The response speed of the model is judged by the path planning time, and the removal ratio of duplicate data is increased for the case where the response speed is unqualified, thereby reducing the data processing volume of the model and improving the response speed of the model. It also provides an example for the subsequent model to obtain data, while ensuring the accuracy of the model. Reduce the data processing volume, so that the trained model can quickly generate an escape route according to the on-site situation when a disaster occurs, and provide timely and accurate escape guidance for personnel evacuation.

[0020] Furthermore, the present invention determines the decision accuracy of the training model through the path overlap. An accurate escape route can ensure that people do not get lost during the evacuation process and avoid entering dangerous areas, so that they can safely reach a safe area, which is of great significance to ensuring personal safety. The escape path planning is qualified by comparing the simulation results with the output results of the training model, and the data of the training model is transformed for unqualified cases to improve the decision accuracy of the training model, so that the trained model can quickly generate an escape path according to the on-site conditions when a disaster occurs, providing timely and accurate escape guidance for personnel evacuation.

[0021] Furthermore, the present invention installs the trained model in the security equipment. When facing a disaster, the escape route is quickly planned according to the data collected by the data acquisition component and the escape route is updated in real time according to the development of the disaster. At the same time, different instructions are sent to the signal receiving device on each escape path, so that the signal receiving devices on different escape paths make different feedback, which can enable people to evacuate along the path with consistent feedback, providing timely and accurate escape guidance for the evacuation of people. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 This is a structural block diagram of a security device according to an embodiment of the present invention; Figure 2 This is a flow chart of a security equipment monitoring method according to an embodiment of the present invention; Figure 3A flow chart of determining whether the response speed of a training model is qualified according to an embodiment of the present invention; Figure 4 The present invention is a flowchart for determining whether the decision accuracy of a training model is qualified. DETAILED DESCRIPTION

[0023] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0024] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0025] It should be noted that, in the description of the present invention, terms such as "up", "down", "left", "right", "inside" and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the drawings. This is merely for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it cannot be understood as a limitation on the present invention.

[0026] In addition, it should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0027] See also Figure 1-Figure 4 As shown, Figure 1 This is a structural block diagram of a security device according to an embodiment of the present invention; Figure 2 This is a flow chart of a security equipment monitoring method according to an embodiment of the present invention; Figure 3 A flow chart of determining whether the response speed of a training model is qualified according to an embodiment of the present invention; Figure 4 The present invention is a flowchart for determining whether the decision accuracy of a training model is qualified.

[0028] An embodiment of the present invention provides a security equipment monitoring method, comprising: Step S1, obtaining historical data of a number of disasters based on a network platform, preprocessing the historical data of any disaster to obtain a number of groups of data to be simulated, and splitting any group of the data to be simulated into training data and verification data in a first preset ratio, wherein the historical data includes the type of disaster, the diffusion rate, the building information of the occurrence location, the environmental information, and the information of the trapped persons; Step S2, performing escape simulation analysis based on the data to be simulated to determine a number of ideal escape paths, combining the data to be simulated and the ideal escape paths corresponding to the training data into a number of data sets, and inputting the data sets into a neural network model to train the model; Step S3, inputting the verification data into the training model to output a number of actual escape paths, determining the path planning time, and determining the eligibility of the response speed of the training model based on the path planning time to increase the removal ratio of duplicate data, or, for a single group of the data to be simulated, determining whether the single decision accuracy of the training model is qualified based on the path overlap; Step S4, counting the percentage of qualified single decision accuracy, so as to determine whether the decision accuracy of the training model is qualified based on the percentage of the number, and determining the transformation method of the training data according to the variance of the path overlap based on the training model with unqualified decision accuracy; Step S5, setting the training model with qualified decision accuracy in the security equipment, generating several escape routes based on the data collected by the camera according to the disaster situation, sending instructions to the signal receiver based on the single escape path, and the signal receiver giving feedback based on the received instructions to guide the crowd to evacuate along the escape route.

[0029] Specifically, disasters include earthquakes, typhoons, floods, and major fires, without limitation.

[0030] Specifically, the first preset ratio is training set / validation set = 7 / 3.

[0031] Specifically, historical data on disasters are obtained, and cluster analysis is performed after preprocessing the historical data. Disasters with similar conditions are classified into one category, such as fire, flood, and earthquake disasters. For a specific type of disaster, the disaster can be further classified according to its severity. For example, fires can be divided into Class A, Class B, etc. according to standards. Modeling is carried out based on the classified data to simulate the disaster scene. An ideal escape route is planned according to the on-site conditions so that trapped people can be evacuated from the dangerous area as safely as possible. Based on the results of cluster analysis, a data set is generated for each type of disaster together with the corresponding ideal escape route data. Finally, several data sets are generated according to the number of classifications. The data set is divided into a training set and a validation set. The training set is input into the neural network model for model training, and the validation set is input into the trained model for model performance verification.

[0032] Specifically, the neural network model is, for example, a convolutional network model or a recurrent network model, which is not specifically limited.

[0033] Specifically, a training model with qualified decision accuracy is set in the escape control center, and a signal receiver is set in, for example, a safety indicator light, without specific limitation. When a disaster is determined based on the information collected by the data acquisition component, the building is divided into areas based on the acquired building information including the floor plan and spatial structure, an escape route is planned for any area, and instructions are sent to the signal receiver on each escape route. For example, a color instruction can be sent to the signal receiver installed in the safety indicator light, or each escape route can be numbered. Different escape routes are marked with different colors or numbered with different numbers, and the same escape route uses the same color or the same number. The safety indicator light that is not on the planned escape route is not lit, so as to guide the crowd to evacuate along the optimal safe route in a chaotic situation, and also help to carry out rescue in a timely and accurate manner.

[0034] Specifically, the process of preprocessing the historical data includes: Cleaning the historical data to eliminate noise, remove missing values ​​and outliers to obtain cleaned data; Normalizing the cleaned data to extract feature data; Eliminating duplicate data in the characteristic data at a second preset ratio to obtain data to be simulated; Cluster analysis is performed based on the data to be simulated, and the data to be simulated of the same type are grouped into one group of data to be simulated to obtain several groups of data to be simulated.

[0035] Specifically, the characteristic data includes but is not limited to the type of disaster, disaster level, number of trapped people, geographical location and disaster frequency.

[0036] Specifically, the value range of the second preset ratio is set to [10%, 30%], and 20% is preferred in the embodiment of the present invention. Too much duplicate data will lead to overfitting of the model, making the model perform well on the training data but poorly on the new data. Excessive duplicate data needs to be removed to ensure the accuracy of the model.

[0037] Specifically, the process of determining whether the response speed of the training model is qualified based on the path planning time includes: Comparing the path planning time with a preset planning time; Determining that the response speed of the training model is unqualified based on a comparison result that the path planning time is greater than a preset planning time; Based on the comparison result that the path planning time is less than or equal to the preset planning time, it is determined that the response speed of the training model is qualified.

[0038] Specifically, the value range of the preset planning time is set to [3min, 20min], and 5min is preferred in the embodiment of the present invention.

[0039] Specifically, the process of increasing the removal ratio of duplicate data based on the condition that the response speed is unqualified includes: Subtracting the path planning time from the preset planning time to obtain a time difference; Comparing the time difference with a preset time difference; Determining to use a first ratio adjustment coefficient to adjust the second preset ratio based on a comparison result that the time difference is greater than the preset time difference; Based on the comparison result that the time difference is less than or equal to the preset time difference, it is determined to use the second ratio adjustment coefficient to adjust the second preset ratio.

[0040] Specifically, the value range of the preset time difference is set to [1min, 5min], and 3min is preferred in the embodiment of the present invention; the value range of the first proportional adjustment coefficient is set to [1.3, 1.5], and 1.4 is preferred in the embodiment of the present invention; the value range of the second proportional adjustment coefficient is set to [1.01, 1.2], and 1.15 is preferred in the embodiment of the present invention.

[0041] Specifically, the process of adjusting the second preset ratio by the proportional adjustment coefficient is: updating the product of the proportional adjustment coefficient and the second preset ratio to a new second preset ratio. For example, if the first proportional adjustment coefficient is used, 1.4×20%=28%, that is, the repeated data in the characteristic data is eliminated at the second preset ratio of 28% to obtain the data to be simulated.

[0042] Specifically, the process of determining the path overlap based on the comparison between the ideal escape path and the actual escape path includes: The ideal escape path and the actual escape path that are completely consistent are recorded as coincident paths; Counting the number of overlaps of the overlapped paths; The ratio of the number of overlaps to the number of actual escape paths is determined as the path overlap degree.

[0043] Specifically, the process of determining whether the accuracy of a single decision is qualified based on the path overlap includes: Comparing the path overlap with a preset overlap; Determining that the single decision accuracy is unqualified based on the comparison result that the path overlap is less than the preset overlap; The accuracy of a single decision is determined to be qualified based on a comparison result that the path overlap is greater than or equal to the preset overlap.

[0044] Specifically, the preset overlap range is set to [85%, 98%], and 95% is preferred in the embodiment of the present invention.

[0045] Specifically, the process of determining whether the decision accuracy of the training model is qualified based on the quantity proportion includes: Comparing the quantity ratio with a preset quantity ratio; Determining that the decision accuracy of the training model is unqualified based on a comparison result that the quantity ratio is less than the preset quantity ratio; Based on the comparison result that the quantity proportion is greater than or equal to the preset quantity proportion, it is determined that the decision accuracy of the training model is qualified.

[0046] Specifically, the value range of the preset quantity ratio is set to [70%, 95%], and 90% is preferred in the embodiment of the present invention.

[0047] Specifically, the process of determining the transformation mode of the training data according to the path overlap variance based on the training model with unqualified decision accuracy includes: Calculating the path overlap variance based on the path overlap based on the training model with unqualified accuracy; Comparing the path overlap variance with a preset variance; Determining to increase the amount of the training data based on a comparison result that the path overlap variance is greater than the preset variance; Based on the comparison result that the path overlap variance is less than or equal to the preset variance, it is determined that the abnormal data in the historical data is corrected.

[0048] Specifically, the value range of the preset variance is set to [0.5, 2], and 0.8 is preferred in the embodiment of the present invention.

[0049] Specifically, if the variance of path overlap is greater than the preset variance, it means that the model has poor decision-making accuracy when facing new data due to the lack of certain data details in the training data, so it is necessary to supplement the data details of the training data. For example, for building information, the building simulation drawings of each view should be supplemented, and the obstacles and other details should be refined to increase the data volume; if the variance of path overlap is less than the preset variance, it means that the model has insufficient decision-making accuracy when facing new data due to the lack of certain data, and the abnormal data removed when preprocessing the historical data should be corrected based on the existing technology.

[0050] Specifically, the abnormal data may be, for example, the temperature at the fire scene is 20° C., which is not specifically limited.

[0051] Specifically, a training model with qualified decision-making accuracy is set in the security equipment, and several escape routes are generated based on the data collected by the camera according to the disaster situation. The escape routes are updated in real time based on changes in the disaster situation. Different instructions are sent to signal receivers for different escape paths. After receiving the instructions, the signal receivers provide different feedback to guide the crowd to evacuate along a single escape route.

[0052] Specifically, the level and scope of the disaster change in real time. For example, when a fire occurs, the scope of the fire may expand over time, and the originally safe escape route may become unsafe. Therefore, it is necessary to judge the disaster situation in real time and update the escape route according to the actual situation to provide a more accurate and safe escape route for the evacuated people.

[0053] The embodiment of the present invention further provides a security device, including: Data collection components, including cameras, temperature sensors, and smoke sensors, to obtain information from the disaster site; a signal receiver for receiving instructions to indicate an escape route; An escape control center is connected to the camera and the signal receiver respectively, and is used for planning an escape route in real time based on the data collected by the camera, and sending instructions to the signal receiver on a single escape route.

[0054] Specifically, the camera is an industrial high-definition camera, which is not specifically limited; the temperature sensor is, for example, a digital temperature sensor, which is not specifically limited and only needs to meet the monitoring requirements; the smoke sensor is, for example, an MQ-2 smoke sensor, which is not specifically limited and only needs to meet the monitoring requirements.

[0055] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

Claims

1. A security equipment monitoring method, characterized in that: include: Acquire historical data of several disaster situations based on a network platform, pre-process the historical data of any disaster situation to obtain several groups of data to be simulated, and split any group of data to be simulated into training data and verification data at a first preset ratio, wherein the historical data includes the type of disaster situation, the diffusion rate, the building information of the occurrence location, the environmental information, and the information of the trapped persons; Performing escape simulation analysis based on the data to be simulated to determine a number of ideal escape paths, combining the data to be simulated and the ideal escape paths corresponding to the training data into a number of data sets, and inputting the data sets into a neural network model to train the model; Input the verification data into the training model to output a number of actual escape paths, determine the path planning time, determine the eligibility of the response speed of the training model based on the path planning time, so as to increase the removal ratio of duplicate data, or, for a single group of the data to be simulated, determine whether the single decision accuracy of the training model is qualified based on the path overlap; Counting the percentage of qualified single decision accuracy, determining whether the decision accuracy of the training model is qualified based on the percentage, and determining the transformation method of the training data according to the variance of path overlap based on the training model with unqualified decision accuracy; A training model with qualified decision accuracy is set in the security equipment, and several escape routes are generated based on the data collected by the camera according to the disaster situation. Instructions are sent to the signal receiver based on the single escape path. The signal receiver provides feedback based on the received instructions to guide the crowd to evacuate along the escape path.

2. The security equipment monitoring method according to claim 1, characterized in that: The process of preprocessing the historical data includes: Cleaning the historical data to eliminate noise, remove missing values ​​and outliers to obtain cleaned data; Normalizing the cleaned data to extract feature data; Eliminating duplicate data in the characteristic data at a second preset ratio to obtain data to be simulated; Cluster analysis is performed based on the data to be simulated, and the same type of data to be simulated is grouped into one group of data to be simulated to obtain several groups of data to be simulated.

3. The security equipment monitoring method according to claim 2, characterized in that: The process of determining whether the response speed of the training model is qualified based on the path planning time includes: Comparing the path planning time with a preset planning time; Based on the comparison result that the path planning time is greater than the preset planning time, it is determined that the response speed of the training model is unqualified.

4. The security equipment monitoring method according to claim 3, characterized in that: The process of increasing the removal ratio of duplicate data based on the condition that the response speed is unqualified includes: Subtracting the path planning time from the preset planning time to obtain a time difference; Comparing the time difference with a preset time difference; Determining to use a first ratio adjustment coefficient to adjust the second preset ratio based on a comparison result that the time difference is greater than the preset time difference; Based on the comparison result that the time difference is less than or equal to the preset time difference, it is determined to use the second ratio adjustment coefficient to adjust the second preset ratio.

5. The security equipment monitoring method according to claim 4, characterized in that: The process of determining the path overlap based on the comparison between the ideal escape path and the actual escape path includes: The ideal escape path and the actual escape path that are completely consistent are recorded as coincident paths; Counting the number of overlaps of the overlapped paths; The ratio of the number of overlaps to the number of actual escape paths is determined as the path overlap degree.

6. The security equipment monitoring method according to claim 5, characterized in that: The process of determining whether the accuracy of a single decision is qualified based on the path overlap includes: Comparing the path overlap with a preset overlap; Based on the comparison result that the path overlap is less than the preset overlap, it is determined that the single decision accuracy is unqualified.

7. The security equipment monitoring method according to claim 6, characterized in that: The process of determining whether the decision accuracy of the training model is qualified based on the quantity proportion includes: Comparing the quantity ratio with a preset quantity ratio; Based on the comparison result that the quantity proportion is less than the preset quantity proportion, it is determined that the decision accuracy of the training model is unqualified.

8. The security equipment monitoring method according to claim 7, characterized in that: The process of determining the transformation method of the training data according to the path overlap variance based on the training model with unqualified decision accuracy includes: Calculating the path overlap variance based on the path overlap based on the training model with unqualified accuracy; Comparing the path overlap variance with a preset variance; Determining to increase the amount of the training data based on a comparison result that the path overlap variance is greater than the preset variance; Based on the comparison result that the path overlap variance is less than or equal to the preset variance, it is determined that the abnormal data in the historical data is corrected.

9. The security equipment monitoring method according to claim 8, characterized in that: A training model with qualified decision-making accuracy is set in the security equipment. Several escape routes are generated based on the data collected by the camera according to the disaster situation, and the escape routes are updated in real time based on changes in the disaster situation. Different instructions are sent to signal receivers for different escape paths. After receiving the instructions, the signal receivers provide different feedback to guide the crowd to evacuate along a single escape route.

10. A security device using the security device monitoring method according to any one of claims 1 to 9, characterized in that: include: Data collection components, including cameras, temperature sensors, and smoke sensors, to obtain information from the disaster site; a signal receiver for receiving instructions to indicate an escape route; An escape control center is connected to the camera and the signal receiver respectively, and is used for planning an escape route in real time based on the data collected by the camera, and sending instructions to the signal receiver on a single escape route.

Images