Artificial intelligence comprehensive fire extinguishing system of electric vehicle charging station

By using artificial intelligence technology in electric vehicle charging stations combined with environmental and visual sensing modules, it can monitor and intelligently judge fire risks in real time, and use robotic arms and intelligent vehicle moving modules to automatically extinguish fires, solving the fire monitoring and fire extinguishing systems in the existing technology. The problem of low accuracy improves the efficiency and safety of fire handling.

CN118987535AInactive Publication Date: 2024-11-22BEIJING UNIVERSAL FUTURE TECH CO LTD
View PDF 0 Cites 3 Cited by

Patent Information

Application Number
CN202411100997.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-12
Publication Date
2024-11-22
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The accuracy of existing electric vehicle charging station fire monitoring and fire extinguishing systems is low, resulting in inaccurate fire identification and low fire extinguishing efficiency, which may cause property damage and casualties.

Method used

Using artificial intelligence technology combined with environmental sensing module and visual sensing module, the data analysis and processing module monitors the temperature, smoke, voltage and charging status of the charging station in real time, intelligently judges fire risks, and uses robotic arm fire extinguishing module and intelligent vehicle moving module. Automatic fire extinguishing.

Benefits of technology

It improves the accuracy of fire monitoring and the intelligence level of fire extinguishing systems, significantly reduces property losses caused by fires, and ensures the rapid and effective treatment of fires.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118987535A_ABST
    Figure CN118987535A_ABST
Patent Text Reader

Abstract

The invention discloses an artificial intelligence comprehensive fire extinguishing system for an electric vehicle charging station, belongs to the technical field of intelligent fire extinguishing, and can monitor the temperature, smoke, voltage and charging conditions of the charging station in real time and intelligently judge fire risks by adopting an artificial intelligence technology, so that the intelligent level of the fire extinguishing system is improved, and the fire monitoring accuracy is improved. And the mechanical arm fire extinguishing module and the intelligent vehicle moving module are arranged, so that property loss caused by fire disasters can be effectively reduced.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent fire extinguishing, and in particular relates to an artificial intelligence integrated fire extinguishing system for an electric vehicle charging station. Background Art

[0002] As the concept of environmental protection has become more popular, electric vehicles have gradually become the preferred means of transportation for people due to their green, environmentally friendly, energy-saving and emission-reduction characteristics. As an important place for charging electric vehicles, the safety of electric vehicle charging stations is of vital importance. However, since heat may be generated during the charging process, if the heat cannot be effectively controlled, it may lead to fire accidents. In the prior art, the method of over-limit sensing is often used to judge fire accidents. Although it can play a certain early warning role, it sometimes leads to inaccurate identification, and it takes a certain amount of time from the occurrence to the alarm and finally to the fire extinguishing, causing a large amount of property losses, and more seriously, it may cause casualties. Summary of the invention

[0003] The present invention provides an artificial intelligence integrated fire extinguishing system for an electric vehicle charging station, which is used to solve the technical problems that the prior art uses a sensor overlimit method to judge fire accidents, resulting in low accuracy and low fire extinguishing efficiency.

[0004] An artificial intelligence integrated fire extinguishing system for electric vehicle charging stations, comprising: an environmental sensing module, a visual sensing module, a data analysis and processing module, a mechanical arm fire extinguishing module, and an intelligent vehicle moving module;

[0005] The environmental sensing module is used to collect operating environment data sensed by environmental sensors deployed in the electric vehicle charging station, and transmit the operating environment data to the data analysis and processing module;

[0006] The visual sensing module is used to collect video data sensed by the visual sensor deployed in the electric vehicle charging station and transmit the video data to the data analysis and processing module;

[0007] The data analysis and processing module is used to analyze the operating environment data and the video data using artificial intelligence to determine the parking space where the fire occurred and the severity of the fire;

[0008] The data analysis and processing module is also used to control the mechanical arm fire extinguishing module and the intelligent car moving module to extinguish the fire according to the parking space where the fire occurs and the severity of the fire;

[0009] The mechanical arm fire extinguishing module is used to accept the control of the data analysis and processing module to extinguish the electric vehicle on the parking space where the fire occurs or move the two-wheeled electric vehicle on the parking space where the fire occurs to the fire extinguishing room;

[0010] The intelligent vehicle moving module is used to accept the control of the data analysis and processing module to move the four-wheel electric vehicle on the parking space where the fire occurs to the fire extinguishing house.

[0011] Furthermore, collecting operating environment data sensed by environmental sensors deployed in the electric vehicle charging station and transmitting the operating environment data to the data analysis and processing module includes:

[0012] Based on the preset sampling frequency, when the data sampling time arrives, the smoke data collected by the smoke sensor deployed in the electric vehicle charging station, the temperature data collected by the temperature sensor, the voltage data on each parking space, and the charging status of each parking space are collected;

[0013] The smoke data, temperature data, voltage data and charging status are used together as operating environment data in the electric vehicle charging station, and the operating environment data is transmitted to a data analysis and processing module.

[0014] Furthermore, collecting video data sensed by a visual sensor deployed in an electric vehicle charging station and transmitting the video data to a data analysis and processing module includes:

[0015] Collect video streams sensed by visual sensors deployed in electric vehicle charging stations;

[0016] Based on the preset sampling frequency, when the data sampling moment arrives, the video stream is subjected to frame extraction processing to obtain video data;

[0017] The video data is transmitted to a data analysis and processing module.

[0018] Furthermore, artificial intelligence is used to analyze the operating environment data and the video data to determine the parking space where the fire occurred and the severity of the fire, including:

[0019] A data matrix is ​​constructed with the operating environment data, and a preset first artificial intelligence model is used to identify the data matrix to determine a fire warning situation; wherein the fire warning situation is the presence of a fire or the absence of a fire;

[0020] When the fire warning situation indicates that a fire exists, a preset second artificial intelligence model is used to analyze the video data to determine the parking space where the fire occurred and the severity of the fire.

[0021] Furthermore, a data matrix is ​​constructed with the operating environment data, and a preset first artificial intelligence model is used to identify the data matrix to determine the fire warning situation, including:

[0022] Based on the current data sampling time, the smoke data, temperature data, voltage data and charging status in the environmental data corresponding to the previous K data sampling times are taken out to obtain a smoke data sequence, a temperature data sequence, a voltage data sequence and a charging status sequence;

[0023] The extracted smoke data sequence, temperature data sequence, voltage data sequence and charging state sequence are used to construct a data matrix;

[0024] The data matrix is ​​used as input of a preset first artificial intelligence model to obtain a fire warning situation.

[0025] Furthermore, when the fire warning condition indicates that a fire exists, a preset second artificial intelligence model is used to analyze the video data to determine the parking space where the fire occurred and the severity of the fire, including:

[0026] When the fire warning condition is that there is a fire, determining the corresponding data sampling time as the target data sampling time;

[0027] Based on the target data sampling time, the target position is determined according to the smoke data sequence, the temperature data sequence, the voltage data sequence and the charging state sequence;

[0028] Taking the target data sampling time as the starting time, a preset second artificial intelligence model is used to identify the video data at the following M data sampling times, and determine the parking space number prediction frame and the fire prediction frame where the fire occurs near the target location;

[0029] The parking space where the fire occurred and the severity of the fire are determined based on the parking space number prediction frame and the fire prediction frame where the fire occurred near the target location.

[0030] Further, based on the target data sampling time, the target position is determined according to the smoke data sequence, the temperature data sequence, the voltage data sequence and the charging state sequence, including:

[0031] Based on the target data sampling time, determining a target smoke sensor with abnormal smoke data in the smoke data sequence; wherein the abnormal smoke data indicates that the smoke concentration exceeds a preset smoke threshold;

[0032] Based on the target data sampling time, determine the target temperature sensor with abnormal temperature data in the temperature data sequence; wherein the abnormal temperature data indicates that the temperature concentration exceeds a preset temperature threshold;

[0033] Based on the target data sampling time, a target voltage sensor with abnormal voltage data is determined in the voltage data sequence; wherein the abnormal voltage data indicates that when the charging state is charging, the voltage data is outside a preset voltage range;

[0034] The points where the target smoke sensor, the target temperature sensor and the target voltage sensor are located are taken as target positions;

[0035] Among them, environmental sensors are set on different parking spaces in the electric vehicle charging station, and each parking space is equipped with a smoke sensor, a temperature sensor and a voltage sensor. The charging status is recorded by the background system of the electric vehicle charging station.

[0036] Further, the parking space where the fire occurred and the severity of the fire are determined according to the parking space number prediction frame and the fire prediction frame near the target location, including:

[0037] According to the fire prediction box, the size of the fire prediction box is determined, and the severity of the fire is determined according to the size of the fire prediction box; wherein the fire prediction box is set to M sizes, and the larger the fire prediction box is, the more serious the fire is, so the range corresponding to the severity of the fire is set to 1-M, corresponding to the M fire prediction boxes arranged in ascending order of size;

[0038] According to the parking space number prediction frame, the parking space number in the parking space number prediction frame is identified by using text recognition technology, and the nearest parking space number and the second nearest parking space number located on the same side of the fire prediction frame are determined;

[0039] Subtract the closest parking space number from the second closest parking space number to get the difference in parking space numbers, and subtract the difference in parking space numbers from the closest parking space number to get the parking space where the fire occurred.

[0040] Furthermore, according to the parking space where the fire occurs and the severity of the fire, the robotic arm fire extinguishing module and the intelligent car moving module are controlled to extinguish the fire, including:

[0041] Determine the vehicle type of the electric vehicle in which the fire occurred according to the parking space in which the fire occurred; wherein the vehicle type is a two-wheeled electric vehicle or a four-wheeled electric vehicle;

[0042] Determine whether the severity of the fire exceeds the preset level. If so, control the robotic arm fire extinguishing module or the intelligent vehicle moving module to extinguish the fire according to the vehicle type. Otherwise, control the robotic arm fire extinguishing module to use a spray fire extinguishing device to extinguish the fire.

[0043] Furthermore, according to the vehicle type, the mechanical arm fire extinguishing module or the intelligent vehicle moving module is controlled to extinguish the fire, including:

[0044] When the vehicle type is a two-wheeled electric vehicle, the mechanical arm fire extinguishing module is controlled to move the two-wheeled electric vehicle on the parking space where the fire occurs to the fire extinguishing room;

[0045] When the vehicle type is a four-wheel electric vehicle, the intelligent vehicle moving module is controlled to move the four-wheel electric vehicle on the parking space where the fire occurs to the fire extinguishing house.

[0046] The present invention provides an artificial intelligence integrated fire extinguishing system for an electric vehicle charging station. By adopting artificial intelligence technology, it can monitor the temperature, smoke, voltage and charging status of the charging station in real time, intelligently judge the fire risk, improve the intelligence level of the fire extinguishing system, and improve the accuracy of fire monitoring. In addition, by setting a mechanical arm fire extinguishing module and an intelligent car moving module, it can effectively reduce the property loss caused by the fire. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0048] Figure 1 A schematic diagram of an artificial intelligence integrated fire extinguishing system for an electric vehicle charging station provided in an embodiment of the present invention.

[0049] Figure 2 A schematic diagram of an environmental sensor arrangement provided by an embodiment of the present invention.

[0050] Figure 3 A schematic diagram of another arrangement of environmental sensors provided in an embodiment of the present invention.

[0051] The above drawings have shown clear embodiments of the present invention, which will be described in more detail below. These drawings and text descriptions are not intended to limit the scope of the present invention in any way, but to illustrate the concept of the present invention for those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0052] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Instead, they are merely examples of devices and methods consistent with some aspects of the present invention as detailed in the appended claims.

[0053] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0054] like Figure 1 As shown, an embodiment of the present invention provides an artificial intelligence integrated fire extinguishing system for an electric vehicle charging station, comprising: an environmental sensing module, a visual sensing module, a data analysis and processing module, a mechanical arm fire extinguishing module, and an intelligent vehicle moving module;

[0055] The environmental sensing module is used to collect operating environment data sensed by environmental sensors deployed in the electric vehicle charging station, and transmit the operating environment data to the data analysis and processing module;

[0056] For example, environmental sensors can be installed on each parking space to monitor the parking space one-to-one, which is more helpful to improve the accuracy of fire monitoring. By collecting operating environment data, fire situations can be effectively analyzed and early warnings can be issued.

[0057] The visual sensing module is used to collect video data sensed by the visual sensor deployed in the electric vehicle charging station and transmit the video data to the data analysis and processing module;

[0058] After a fire occurs or a possible fire is detected, the actual situation may be a fire or not (for example, a sensor is damaged, or a fire occurs in other parking spaces, resulting in smoke detection, and the location is in hot weather, resulting in a misjudgment). Therefore, in order to further detect the fire situation and realize intelligent fire extinguishing, it is necessary to collect video data for analysis.

[0059] The data analysis and processing module is used to analyze the operating environment data and the video data using artificial intelligence to determine the parking space where the fire occurred and the severity of the fire;

[0060] By analyzing the operating environment data, we can preliminarily determine the location where the fire may have occurred, and then analyze the video data to determine the parking space where the fire actually occurred and the severity of the fire. This linkage recognition method can greatly improve the accuracy of fire recognition.

[0061] The data analysis and processing module is also used to control the mechanical arm fire extinguishing module and the intelligent car moving module to extinguish the fire according to the parking space where the fire occurs and the severity of the fire;

[0062] The mechanical arm fire extinguishing module is used to accept the control of the data analysis and processing module to extinguish the electric vehicle on the parking space where the fire occurs or move the two-wheeled electric vehicle on the parking space where the fire occurs to the fire extinguishing room;

[0063] The embodiment of the present invention can automatically determine the location and size of the fire through various sensors, and automatically take fire extinguishing actions according to the size of the fire. If the fire is small, the fire extinguishing device of the mechanical arm is used to directly extinguish the fire. If the fire is large, the two-wheeled electric vehicle is directly grabbed by the mechanical arm and moved to the fire extinguishing room to extinguish the fire, so as to ensure that the fire is extinguished in the shortest time. Where there is a fire, the mechanical arm automatically slides to the top of the fire location through the slide rail. Since the parking lot where the two-wheeled electric vehicle is located is not large and the distance between the ceiling and the ground is not high, the fire extinguishing structure on the mechanical arm can directly extinguish the fire (when the fire is small and can be controlled).

[0064] The intelligent vehicle moving module is used to accept the control of the data analysis and processing module to move the four-wheel electric vehicle on the parking space where the fire occurs to the fire extinguishing house.

[0065] A fire extinguishing house is set up at the designated location of the charging station, and artificial intelligence fire extinguishing facilities are installed in the fire extinguishing house. When the fire extinguishing device on the charging station's robotic arm fire extinguishing module cannot extinguish the open fire ignited by the car battery, an artificial intelligence car moving robot (i.e., intelligent car moving module) is needed to move the burning vehicle to the fire extinguishing house for fire extinguishing. There is a special fire extinguishing pool in the fire extinguishing house, which contains special fire extinguishing agent. The car moving robot moves the burning vehicle to the fire extinguishing pool in the fire extinguishing house for flooding fire extinguishing. There are several fire extinguishing robots on the walls of the fire extinguishing house, which can carry out fire extinguishing work at the same time, so as to ensure that other vehicles are not implicated by the burning vehicle, and to solve the pain point of uncontrollable vehicle battery fire. All equipment and real-time dynamics of the charging station are connected to the general monitoring center, so as to understand the fire situation at any time and make relevant actions and deployments in time.

[0066] The fire-fighting house can be set up as a semi-enclosed space without a roof (to facilitate the robotic arm fire-fighting module to grab the two-wheeled electric vehicle into the fire-fighting house), with a vehicle entry position on one side (to facilitate the intelligent vehicle moving module to move the four-wheeled electric vehicle into the fire-fighting house), and the three walls are made of fire-resistant materials.

[0067] The present invention provides an artificial intelligence integrated fire extinguishing system for an electric vehicle charging station. By adopting artificial intelligence technology, it can monitor the temperature, smoke, voltage and charging status of the charging station in real time, intelligently judge the fire risk, improve the intelligence level of the fire extinguishing system, and improve the accuracy of fire monitoring. In addition, by setting a mechanical arm fire extinguishing module and an intelligent car moving module, it can effectively reduce the property loss caused by the fire.

[0068] In an embodiment of the present invention, collecting operating environment data sensed by environmental sensors deployed in an electric vehicle charging station and transmitting the operating environment data to a data analysis and processing module includes:

[0069] Based on the preset sampling frequency, when the data sampling time arrives, the smoke data collected by the smoke sensor deployed in the electric vehicle charging station, the temperature data collected by the temperature sensor, the voltage data on each parking space, and the charging status of each parking space are collected;

[0070] The smoke data, temperature data, voltage data and charging status are used together as operating environment data in the electric vehicle charging station, and the operating environment data is transmitted to a data analysis and processing module.

[0071] like Figure 2As shown, in order to ensure the collection of operating environment data, the same sensors can be set at multiple points to achieve comprehensive detection of the operating environment of the charging station.

[0072] like Figure 3 As shown, smoke sensors and temperature sensors may also be set at different points to collect voltage data and charging status at each parking space. When abnormality judgment is made, the parking space with abnormal voltage data within the area surrounded by the abnormal smoke sensors and abnormal temperature sensors will be regarded as an abnormal position.

[0073] For example, in Figure 3 In the figure, the four sensors at the bottom all detected abnormal smoke and temperature. Figure 3 The following area can be determined as an abnormal area. When the voltage data in parking spaces No. 4, No. 5, and No. 6 in the abnormal area are abnormal, the corresponding parking spaces are determined to be abnormal positions.

[0074] In an embodiment of the present invention, collecting video data sensed by a visual sensor deployed in an electric vehicle charging station and transmitting the video data to a data analysis and processing module includes:

[0075] Collect video streams sensed by visual sensors deployed in electric vehicle charging stations;

[0076] Based on the preset sampling frequency, when the data sampling moment arrives, the video stream is subjected to frame extraction processing to obtain video data;

[0077] The video data is transmitted to a data analysis and processing module.

[0078] There are not too many requirements for visual sensors. It is sufficient to have multiple visual sensors fully cover the charging station and be able to capture the parking space number of each parking space for subsequent identification.

[0079] In an embodiment of the present invention, artificial intelligence is used to analyze the operating environment data and the video data to determine the parking space where the fire occurred and the severity of the fire, including:

[0080] A data matrix is ​​constructed with the operating environment data, and a preset first artificial intelligence model is used to identify the data matrix to determine a fire warning situation; wherein the fire warning situation is the presence of a fire or the absence of a fire;

[0081] Optionally, the embodiment of the present invention preferably sets an environmental sensor on each parking space to achieve high-precision detection. Therefore, the operating environment data of each parking space can be constructed into a data matrix, and then the constructed data matrix is ​​identified using a preset first artificial intelligence model to determine the fire warning situation of each parking space.

[0082] When the fire warning situation indicates that a fire exists, a preset second artificial intelligence model is used to analyze the video data to determine the parking space where the fire occurred and the severity of the fire.

[0083] When it is determined that there is a fire, it proves that there is an abnormal situation in reality. In order to avoid vehicle losses caused by misidentification, the video data is secondary analyzed through a preset second artificial intelligence model to determine whether a fire has occurred and the severity, and intelligent fire extinguishing is carried out based on this.

[0084] In an embodiment of the present invention, a data matrix is ​​constructed using the operating environment data, and a preset first artificial intelligence model is used to identify the data matrix to determine the fire warning situation, including:

[0085] Based on the current data sampling time, the smoke data, temperature data, voltage data and charging status in the environmental data corresponding to the previous K data sampling times are taken out to obtain a smoke data sequence, a temperature data sequence, a voltage data sequence and a charging status sequence;

[0086] The extracted smoke data sequence, temperature data sequence, voltage data sequence and charging state sequence are used to construct a data matrix;

[0087] The data matrix is ​​used as input of a preset first artificial intelligence model to obtain a fire warning situation.

[0088] In the embodiment of the present invention, when the fire warning condition is that there is a fire, a preset second artificial intelligence model is used to analyze the video data to determine the parking space where the fire occurred and the severity of the fire, including:

[0089] When the fire warning condition is that there is a fire, determining the corresponding data sampling time as the target data sampling time;

[0090] Based on the target data sampling time, the target position is determined according to the smoke data sequence, the temperature data sequence, the voltage data sequence and the charging state sequence;

[0091] Taking the target data sampling time as the starting time, a preset second artificial intelligence model is used to identify the video data at the following M data sampling times, and determine the parking space number prediction frame and the fire prediction frame where the fire occurs near the target location;

[0092] That is, after the target location is determined, the video data containing the target location is analyzed, so that the parking space number prediction frame and the fire prediction frame where the fire occurs near the target location can be determined.

[0093] The parking space where the fire occurred and the severity of the fire are determined based on the parking space number prediction frame and the fire prediction frame where the fire occurred near the target location.

[0094] Optionally, in the embodiment of the present invention, the first artificial intelligence model is set as a convolutional neural network, and the second artificial intelligence model is set as a YOLO neural network. After the initial analysis of the operating environment by the convolutional neural network, a fire warning can be achieved. Then, the parking space and the fire location are identified by the YOLO neural network, so that it can be accurately determined whether a fire has occurred.

[0095] Both the first artificial intelligence model and the second artificial intelligence model need to be pre-trained before they can perform normal data recognition functions. Therefore, an embodiment of the present invention provides a pre-training method to obtain a preset artificial intelligence model, which may include:

[0096] A1. Initialize the hyperparameters of the artificial intelligence model to be trained, obtain the hyperparameter encoding, and repeatedly obtain multiple hyperparameter encodings to complete the hyperparameter initialization. The hyperparameters may be weight parameters.

[0097] A2. Obtain the fitness value encoded by each hyperparameter; wherein the fitness value is set to the negative of the loss function value.

[0098] A3. For each hyperparameter encoding, first arrange it in order from small to large according to the fitness value, and then use the optimal position reference to perform a search as follows:

[0099]

[0100] Among them, V i t represents the speed corresponding to the i-th hyperparameter encoding during the t-th training process, V i t+1 represents the speed corresponding to the i-th hyperparameter encoding during the t+1-th training process, α represents the speed control coefficient, β represents the own position learning factor, R represents the constant factor, i=1,2,…,I, I represents the total number of hyperparameter encodings; represents the optimal hyperparameter encoding;

[0101]

[0102] in, represents the i-th hyperparameter encoding during the t-th training process, Indicates the number of

[0103] By searching the hyperparameter encoding once, the worse encoding can learn to update the optimal position information, and the better encoding can reduce its dependence on the optimal position, which can enhance the global search. As the algorithm proceeds, t gradually increases, which can make the worse encoding approach the optimal encoding faster, thereby improving the local search ability of the algorithm.

[0104] A4. For the hyperparameter encoding after the first search, the entire population information is integrated to further balance the global search and the local search, and the second search is performed as follows:

[0105]

[0106] Where c represents the adaptive coefficient, and c max represents the maximum value of the adaptive coefficient, c min represents the minimum value of the adaptive coefficient, T represents the preset maximum number of training times, UB represents the upper bound vector, LB represents the lower bound vector, Representation Encoding With encoding The degree of information interaction between represents the hyperparameter encoding after the jth search, represents the hyperparameter encoding after the mth search, Indicates the number of D jm Representation Encoding With encoding The Euclidean distance between represents the optimal hyperparameter encoding, and d represents the d-th dimension parameter.

[0107] The function of the degree of information interaction between codes is:

[0108]

[0109] Among them, e represents the natural production constant, λ represents the interaction factor, and x represents D jm .

[0110] By performing a secondary search on the hyperparameter encoding, the algorithm can perform a smaller search in the later stage, which is more conducive to exploring more space. The algorithm has a certain degree of jumpiness in the early stage, and a large-scale rapid transfer is more conducive to the algorithm's global search.

[0111] A5. For the hyperparameter encoding after the second search, obtain the reverse learning solution of the hyperparameter encoding, and retain the encoding with greater fitness between the reverse learning solution and the hyperparameter encoding after the second search, to obtain the hyperparameter encoding after the third search.

[0112] Compared with the previous two searches, this reverse learning can effectively avoid falling into the local optimum without deteriorating the training effect, thus improving the convergence ability of the algorithm.

[0113] A6. Determine whether the number of training times is greater than or equal to the maximum number of training times. If so, re-obtain the optimal hyperparameter encoding based on the hyperparameter encoding after three searches, and use the optimal hyperparameter encoding as the final parameter of the artificial intelligence model to be trained to complete pre-training. Otherwise, return to step A2.

[0114] In an embodiment of the present invention, based on the target data sampling time, determining the target position according to the smoke data sequence, the temperature data sequence, the voltage data sequence and the charging state sequence includes:

[0115] Based on the target data sampling time, determining a target smoke sensor with abnormal smoke data in the smoke data sequence; wherein the abnormal smoke data indicates that the smoke concentration exceeds a preset smoke threshold;

[0116] Based on the target data sampling time, determine the target temperature sensor with abnormal temperature data in the temperature data sequence; wherein the abnormal temperature data indicates that the temperature concentration exceeds a preset temperature threshold;

[0117] Based on the target data sampling time, a target voltage sensor with abnormal voltage data is determined in the voltage data sequence; wherein the abnormal voltage data indicates that when the charging state is charging, the voltage data is outside a preset voltage range;

[0118] The points where the target smoke sensor, target temperature sensor and target voltage sensor are located are taken as target positions (that is, when a fire occurs, the smoke sensor, temperature sensor and voltage sensor on the same parking space will change, so these abnormal points can be determined as target positions); the embodiment of the present invention performs a secondary screening on all the existing abnormal points, which can greatly improve the efficiency of fire identification, and also reduce the data processing pressure brought by video analysis, so that the background operation pressure is relatively small.

[0119] Among them, environmental sensors are set on different parking spaces in the electric vehicle charging station, and each parking space is equipped with a smoke sensor, a temperature sensor and a voltage sensor. The charging status is recorded by the background system of the electric vehicle charging station.

[0120] When a sensor data on a parking space shows an abnormality, it is very likely not a fire, so linkage identification is required, and the temporal connection of the data is analyzed to initially determine the location where the fire may have occurred.

[0121] After the user places an order, the charging status is set to "charging" during the charging time of the order; and after the charging time of the order ends or the user actively ends it, the charging status is set to "not charging". A parking space that is being charged is more favorable than a parking space that is not being charged, and fire may occur. Therefore, the embodiment of the present invention introduces linkage identification of the charging status to avoid misidentification.

[0122] In an embodiment of the present invention, determining the parking space where the fire occurred and the severity of the fire according to the parking space number prediction frame and the fire prediction frame near the target position includes:

[0123] According to the fire prediction box, the size of the fire prediction box is determined, and the severity of the fire is determined according to the size of the fire prediction box; wherein the fire prediction box is set to M sizes, and the larger the fire prediction box is, the more serious the fire is, so the range corresponding to the severity of the fire is set to 1-M, corresponding to the M fire prediction boxes arranged in ascending order of size;

[0124] After training, the YOLO neural network can identify parking space numbers and fire areas, and the YOLO neural network can set prediction frames of various sizes to anchor the target. In the actual situation of the fire, if the position of the visual sensor does not change, the larger the fire, the larger the prediction frame required. Therefore, the serial number corresponding to the size of the fire prediction frame can be used as the severity of the fire. The larger the serial number, the larger the fire.

[0125] According to the parking space number prediction frame, the parking space number in the parking space number prediction frame is identified by using text recognition technology, and the nearest parking space number and the second nearest parking space number located on the same side of the fire prediction frame are determined;

[0126] Subtract the closest parking space number from the second closest parking space number to get the difference in parking space numbers, and subtract the difference in parking space numbers from the closest parking space number to get the parking space where the fire occurred.

[0127] Since the fire situation analyzed through the operating environment data may be misidentified or a single data abnormality may indicate the existence of a fire, after a fire occurs, the fire location can be further determined through video data analysis, thereby facilitating intelligent fire extinguishing.

[0128] In an embodiment of the present invention, according to the parking space where the fire occurs and the severity of the fire, the mechanical arm fire extinguishing module and the intelligent car moving module are controlled to extinguish the fire, including:

[0129] Determine the vehicle type of the electric vehicle in which the fire occurred according to the parking space in which the fire occurred; wherein the vehicle type is a two-wheeled electric vehicle or a four-wheeled electric vehicle;

[0130] Determine whether the severity of the fire exceeds the preset level. If so, control the robotic arm fire extinguishing module or the intelligent vehicle moving module to extinguish the fire according to the vehicle type. Otherwise, control the robotic arm fire extinguishing module to use a spray fire extinguishing device to extinguish the fire.

[0131] In an embodiment of the present invention, controlling a mechanical arm fire extinguishing module or an intelligent vehicle moving module to extinguish a fire according to the vehicle type includes:

[0132] When the vehicle type is a two-wheeled electric vehicle, the mechanical arm fire extinguishing module is controlled to move the two-wheeled electric vehicle on the parking space where the fire occurs to the fire extinguishing room;

[0133] When the vehicle type is a four-wheel electric vehicle, the intelligent vehicle moving module is controlled to move the four-wheel electric vehicle on the parking space where the fire occurs to the fire extinguishing house.

[0134] Those skilled in the art will readily appreciate other embodiments of the present invention after considering the specification and practicing the invention disclosed herein. The present invention is intended to cover any variations, uses or adaptations of the present invention that follow the general principles of the present invention and include common knowledge or customary techniques in the art that are not disclosed by the present invention. It should be understood that the present invention is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from the scope thereof. The scope of the present invention is limited only by the appended claims.

Claims

1. An artificial intelligence integrated fire extinguishing system for electric vehicle charging stations, characterized in that: include: Environmental sensing module, visual sensing module, data analysis and processing module, robotic arm fire extinguishing module and intelligent vehicle moving module; The environmental sensing module is used to collect operating environment data sensed by environmental sensors deployed in the electric vehicle charging station, and transmit the operating environment data to the data analysis and processing module; The visual sensing module is used to collect video data sensed by the visual sensor deployed in the electric vehicle charging station and transmit the video data to the data analysis and processing module; The data analysis and processing module is used to analyze the operating environment data and the video data using artificial intelligence to determine the parking space where the fire occurred and the severity of the fire; The data analysis and processing module is also used to control the mechanical arm fire extinguishing module and the intelligent car moving module to extinguish the fire according to the parking space where the fire occurs and the severity of the fire; The mechanical arm fire extinguishing module is used to accept the control of the data analysis and processing module to extinguish the electric vehicle on the parking space where the fire occurs or move the two-wheeled electric vehicle on the parking space where the fire occurs to the fire extinguishing room; The intelligent vehicle moving module is used to accept the control of the data analysis and processing module to move the four-wheel electric vehicle on the parking space where the fire occurs to the fire extinguishing house.

2. The electric vehicle charging station artificial intelligence integrated fire extinguishing system according to claim 1 is characterized in that: Collecting operating environment data sensed by environmental sensors deployed in electric vehicle charging stations, and transmitting the operating environment data to a data analysis and processing module, including: Based on the preset sampling frequency, when the data sampling time arrives, the smoke data collected by the smoke sensor deployed in the electric vehicle charging station, the temperature data collected by the temperature sensor, the voltage data on each parking space, and the charging status of each parking space are collected; The smoke data, temperature data, voltage data and charging status are used together as operating environment data in the electric vehicle charging station, and the operating environment data is transmitted to a data analysis and processing module.

3. The electric vehicle charging station artificial intelligence integrated fire extinguishing system according to claim 2 is characterized in that: Collecting video data sensed by a visual sensor deployed in an electric vehicle charging station and transmitting the video data to a data analysis and processing module, including: Collect video streams sensed by visual sensors deployed in electric vehicle charging stations; Based on the preset sampling frequency, when the data sampling moment arrives, the video stream is subjected to frame extraction processing to obtain video data; The video data is transmitted to a data analysis and processing module.

4. The electric vehicle charging station artificial intelligence integrated fire extinguishing system according to claim 3 is characterized in that: Artificial intelligence is used to analyze the operating environment data and video data to determine the parking space where the fire occurred and the severity of the fire, including: A data matrix is ​​constructed with the operating environment data, and a preset first artificial intelligence model is used to identify the data matrix to determine a fire warning situation; wherein the fire warning situation is the presence of a fire or the absence of a fire; When the fire warning situation indicates that a fire exists, a preset second artificial intelligence model is used to analyze the video data to determine the parking space where the fire occurred and the severity of the fire.

5. The electric vehicle charging station artificial intelligence integrated fire extinguishing system according to claim 4 is characterized in that: Constructing a data matrix with the operating environment data, and using a preset first artificial intelligence model to identify the data matrix to determine the fire warning situation, including: Based on the current data sampling time, the smoke data, temperature data, voltage data and charging status in the environmental data corresponding to the previous K data sampling times are taken out to obtain a smoke data sequence, a temperature data sequence, a voltage data sequence and a charging status sequence; The extracted smoke data sequence, temperature data sequence, voltage data sequence and charging state sequence are used to construct a data matrix; The data matrix is ​​used as input of a preset first artificial intelligence model to obtain a fire warning situation.

6. The electric vehicle charging station artificial intelligence integrated fire extinguishing system according to claim 5 is characterized in that: When the fire warning condition indicates that there is a fire, the video data is analyzed using a preset second artificial intelligence model to determine the parking space where the fire occurred and the severity of the fire, including: When the fire warning condition is that there is a fire, determining the corresponding data sampling time as the target data sampling time; Based on the target data sampling time, the target position is determined according to the smoke data sequence, the temperature data sequence, the voltage data sequence and the charging state sequence; Taking the target data sampling time as the starting time, a preset second artificial intelligence model is used to identify the video data at the following M data sampling times, and determine the parking space number prediction frame and the fire prediction frame where the fire occurs near the target location; The parking space where the fire occurred and the severity of the fire are determined based on the parking space number prediction frame and the fire prediction frame where the fire occurred near the target location.

7. The electric vehicle charging station artificial intelligence integrated fire extinguishing system according to claim 6 is characterized in that: Based on the target data sampling time, the target position is determined according to the smoke data sequence, the temperature data sequence, the voltage data sequence and the charging state sequence, including: Based on the target data sampling time, determining a target smoke sensor with abnormal smoke data in the smoke data sequence; wherein the abnormal smoke data indicates that the smoke concentration exceeds a preset smoke threshold; Based on the target data sampling time, determine the target temperature sensor with abnormal temperature data in the temperature data sequence; wherein the abnormal temperature data indicates that the temperature concentration exceeds a preset temperature threshold; Based on the target data sampling time, a target voltage sensor with abnormal voltage data is determined in the voltage data sequence; wherein the abnormal voltage data indicates that when the charging state is charging, the voltage data is outside a preset voltage range; The points where the target smoke sensor, the target temperature sensor and the target voltage sensor are located are taken as target positions; Among them, environmental sensors are set on different parking spaces in the electric vehicle charging station, and each parking space is equipped with a smoke sensor, a temperature sensor and a voltage sensor. The charging status is recorded by the background system of the electric vehicle charging station.

8. The electric vehicle charging station artificial intelligence integrated fire extinguishing system according to claim 7 is characterized in that: The parking space where the fire occurred and the severity of the fire are determined according to the parking space number prediction frame and the fire prediction frame near the target location, including: According to the fire prediction box, the size of the fire prediction box is determined, and the severity of the fire is determined according to the size of the fire prediction box; wherein the fire prediction box is set to M sizes, and the larger the fire prediction box is, the more serious the fire is, so the range corresponding to the severity of the fire is set to 1-M, corresponding to the M fire prediction boxes arranged in ascending order of size; According to the parking space number prediction frame, the parking space number in the parking space number prediction frame is identified by using text recognition technology, and the nearest parking space number and the second nearest parking space number located on the same side of the fire prediction frame are determined; Subtract the closest parking space number from the second closest parking space number to get the difference in parking space numbers, and subtract the difference in parking space numbers from the closest parking space number to get the parking space where the fire occurred.

9. The electric vehicle charging station artificial intelligence integrated fire extinguishing system according to claim 8 is characterized in that: According to the parking space where the fire occurred and the severity of the fire, the robot arm fire extinguishing module and the intelligent car moving module are controlled to extinguish the fire, including: Determine the vehicle type of the electric vehicle in which the fire occurred according to the parking space in which the fire occurred; wherein the vehicle type is a two-wheeled electric vehicle or a four-wheeled electric vehicle; Determine whether the severity of the fire exceeds the preset level. If so, control the robotic arm fire extinguishing module or the intelligent vehicle moving module to extinguish the fire according to the vehicle type. Otherwise, control the robotic arm fire extinguishing module to use a spray fire extinguishing device to extinguish the fire.

10. The electric vehicle charging station artificial intelligence integrated fire extinguishing system according to claim 9 is characterized in that: Control the robotic arm fire extinguishing module or the intelligent vehicle moving module to extinguish the fire according to the vehicle type, including: When the vehicle type is a two-wheeled electric vehicle, the mechanical arm fire extinguishing module is controlled to move the two-wheeled electric vehicle on the parking space where the fire occurs to the fire extinguishing room; When the vehicle type is a four-wheel electric vehicle, the intelligent vehicle moving module is controlled to move the four-wheel electric vehicle on the parking space where the fire occurs to the fire extinguishing house.

Citation Information

Cited By

  • AI-based linkage response and control system

    CN119296296A

  • Charging station safety control method, system and device and readable storage medium

    CN119795982A

  • Safety control methods, systems, equipment, and readable storage media for charging stations

    CN119795982B