Method and system for detecting abnormal position of automobile charging gun and storage medium

Abstract

The invention discloses an automobile charging gun abnormal position detection method and system and a storage medium, and the method comprises the steps: collecting the current video data of an automobile charging gun, and extracting a plurality of current state images from the current video data based on a preset period; identifying the plurality of current state images based on a preset identification model formed by training pre-collected sample data, and judging whether the automobile charging gun is in an abnormal position or not; and if the automobile charging gun is located at the abnormal position, obtaining a charging gun identifier and first position information corresponding to the automobile charging gun, and outputting abnormal prompt information based on the charging gun identifier and the first position information. According to the method, the image reflecting the real-time state of the automobile charging gun is extracted from the collected video data of the automobile charging gun, the abnormal position is recognized by the preset recognition model, and the abnormal prompt information is output, so that the automobile charging gun at the abnormal position is found in time, and the automobile charging gun is prevented from being damaged.

Description

Technical Field

[0001] The present invention relates to the technical field of electric vehicles, and in particular to a method, system and storage medium for detecting an abnormal position of a vehicle charging gun. Background Art

[0002] With the advancement of technology, the promotion of electric vehicles is getting higher and higher. Charging piles are charging devices that provide electric energy for electric vehicles. The wide distribution of charging piles is an important factor affecting the promotion of electric vehicles. In recent years, although the distribution of electric vehicle charging piles has increased, there are situations where the use of other users is affected by the improper use of some users. For example, for the charging gun in the charging pile, some users may not hang the charging gun back into the charging pile after use, but place it on the ground or hang it on other objects, so that the charging gun is in an abnormal position, and then it may be crushed by the car or damaged by collision. Therefore, how to promptly detect that the car charging gun is in an abnormal position to avoid damage to the car charging gun is a technical problem that needs to be solved urgently. Summary of the invention

[0003] The main purpose of the present invention is to provide a method, system and storage medium for detecting the abnormal position of a vehicle charging gun, aiming to solve the technical problem of how to promptly detect that the vehicle charging gun is in an abnormal position to avoid damage to the vehicle charging gun.

[0004] To achieve the above object, the present invention provides a method for detecting an abnormal position of a vehicle charging gun, the method comprising:

[0005] Collecting current video data of the vehicle charging gun, and extracting a plurality of current state images from the current video data based on a preset period;

[0006] Recognize the multiple current state images based on a preset recognition model to determine whether the vehicle charging gun is in an abnormal position, wherein the preset recognition model is trained based on pre-collected sample data;

[0007] If the vehicle charging gun is in an abnormal position, a charging gun identifier and first position information corresponding to the vehicle charging gun are obtained, and abnormal prompt information is output based on the charging gun identifier and the first position information.

[0008] Preferably, the step of outputting abnormal prompt information based on the charging gun identifier and the first position information includes:

[0009] Determine a generation time point of the current state image, and obtain video data within a preset time length before the generation time point;

[0010] Determine, according to the video data, a car charging gun user corresponding to the current state image, and search for a first type of account corresponding to the car charging gun user;

[0011] Determine whether there is a historical abnormality mark for the first category of accounts; if there is the historical abnormality mark, obtain the number of historical abnormalities corresponding to the historical abnormality mark, and update the number of historical abnormalities;

[0012] Compare the historical number of abnormalities with each preset abnormality number interval to determine a target abnormality number interval corresponding to the historical number of abnormalities, and determine target restriction information corresponding to the target abnormality number interval according to the correspondence between each preset abnormality number interval and restriction information;

[0013] The target restriction information is added to the first type of account, where the target restriction information includes at least one of information on limiting the number of uses and information on increasing the deposit amount.

[0014] Preferably, the step of outputting abnormal prompt information based on the charging gun identifier and the first position information includes:

[0015] Determine whether there is a charging user in a neighboring area within a preset range of the first location information, and if there is the charging user, send a help message to the second type account corresponding to the charging user;

[0016] When it is detected that the state of the car charging gun changes based on the help information, reward information is added to the second type of account corresponding to the charging user, and the reward information at least includes any one of increasing the number of uses and reducing the deposit amount.

[0017] Preferably, if the charging user exists, the step of sending a help message to the second type of account corresponding to the charging user includes:

[0018] receiving feedback information sent based on the help information, and determining second location information corresponding to the feedback information;

[0019] Generate path information based on the first location information and the second location information, and generate confirmation information based on the user name of the charging user;

[0020] The path information and the confirmation information are sent to the second type of account of the charging user, so as to guide the charging user based on the path information and the confirmation information.

[0021] Preferably, the step of generating path information based on the first location information and the second location information comprises:

[0022] Determine whether there are multiple path information, and if there are multiple path information, obtain the path distances corresponding to each of the path information;

[0023] Comparing the path distances, and determining the target path information having the shortest distance among the plurality of path information;

[0024] A charging user corresponding to the target path information is searched from the charging users, and a step of generating confirmation information based on the user name of the charging user is performed.

[0025] Preferably, the step of identifying the current state image based on a preset recognition model and determining whether the vehicle charging gun is in an abnormal position includes:

[0026] Recognize the current state image based on a preset recognition model, obtain current state information of the vehicle charging gun, and determine the information type of the current state information;

[0027] If the information type is a ground type, it is determined that the vehicle charging gun is in an abnormal position;

[0028] If the information type is a hanging gun type or a charging type, it is determined that the vehicle charging gun is in a non-abnormal position, and the status information of the vehicle charging gun is updated based on the non-abnormal position.

[0029] Preferably, the step of collecting current video data of the vehicle charging gun includes:

[0030] Acquire sample data corresponding to the vehicle charging gun, the sample data including a gun hanging image sample, a charging image sample, and a ground placement image sample;

[0031] The preset initial model is trained based on the sample data, and when the training reaches a preset end condition, the preset initial model is generated as a preset recognition model.

[0032] Preferably, the step of training a preset initial model based on the sample data and generating the preset initial model into a preset recognition model when the training reaches a preset end condition comprises:

[0033] Training the preset initial model based on the training samples in the sample data, and adjusting the model parameters of the preset initial model based on the parameter adjustment samples in the sample data after the training reaches a preset number of times;

[0034] Training the adjusted preset initial model based on the training samples, and after the training reaches a preset number of times, testing the preset initial model based on the test samples in the sample data to obtain a test result;

[0035] Based on the test result, determining whether the training of the preset initial model has reached a preset end condition, and if the preset end condition has been reached, generating the preset initial model as a preset recognition model;

[0036] If the preset end condition is not reached, the step of adjusting the model parameters of the preset initial model based on the parameter adjustment sample in the sample data is executed until the training of the preset initial model reaches the preset end condition.

[0037] Furthermore, to achieve the above-mentioned object, the present invention also provides a detection system for abnormal position of a vehicle charging gun, the detection system for abnormal position of a vehicle charging gun comprising: a memory, a processor, a communication bus and a control program stored in the memory:

[0038] The communication bus is used to realize the connection and communication between the processor and the memory;

[0039] The processor is used to execute the control program to implement the steps of the method for detecting the abnormal position of the vehicle charging gun as described above.

[0040] Furthermore, to achieve the above-mentioned purpose, the present invention also provides a storage medium, on which a control program is stored, and when the control program is executed by a processor, the steps of the method for detecting the abnormal position of a vehicle charging gun as described above are implemented.

[0041] The detection method, system and storage medium of the abnormal position of the automobile charging gun of the present invention, after collecting the current video data of the automobile charging gun, extract multiple current state images from the current video data according to a preset period; then recognize the multiple current state images through a preset recognition model formed by training the pre-collected sample data, and judge whether the automobile charging gun is in the ground state through recognition; if it is determined that the automobile charging gun is in the ground state, it means that the automobile charging gun is in an abnormal position; then obtain the charging gun identification and first position information corresponding to the automobile charging gun, and output abnormal prompt information based on the charging gun identification and the first position information. In this way, by collecting the current real-time video data of the automobile charging gun, an image reflecting the real-time state of the automobile charging gun is extracted therefrom, and the preset recognition model is used to identify and determine whether the automobile charging gun is in an abnormal position. Once it is detected that it is in an abnormal position, the charging gun identification and the first position information of the automobile charging gun in the abnormal position are formed as abnormal prompt information output, so as to timely find the automobile charging gun in the abnormal position and avoid damage to the automobile charging gun. In addition, the preset recognition model is formed by training a large amount of pre-collected car charging gun video data, which can make the detection of abnormal positions of car charging guns more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 This is a flow chart of a first embodiment of a method for detecting an abnormal position of a vehicle charging gun according to the present invention;

[0043] Figure 2 This is a flow chart of a second embodiment of a method for detecting an abnormal position of a vehicle charging gun according to the present invention;

[0044] Figure 3 This is a flow chart of a third embodiment of a method for detecting an abnormal position of a vehicle charging gun according to the present invention;

[0045] Figure 4 This is a schematic diagram of the structure of the hardware operating environment involved in an embodiment of the system for detecting abnormal position of a vehicle charging gun of the present invention.

[0046] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0047] 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.

[0048] The present invention provides a method for detecting abnormal position of a vehicle charging gun, please refer to Figure 1 , Figure 1 This is a flow chart of a first embodiment of a method for detecting an abnormal position of a vehicle charging gun according to the present invention.

[0049] The embodiment of the present invention provides an embodiment of a method for detecting an abnormal position of a vehicle charging gun. It should be noted that although the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than that here. Specifically, the tunnel anchor safety performance evaluation in this embodiment includes:

[0050] Step S10, collecting current video data of the vehicle charging gun, and extracting a plurality of current state images from the current video data based on a preset period.

[0051] The car charging gun of this embodiment is a charging gun in a charging pile for charging electric vehicles. In order to ensure that the charging pile can charge the electric vehicle normally, a control center for managing and controlling the charging pile is provided. The detection method of the abnormal position of the car charging gun of this embodiment is applied to the control center. Specifically, a monitoring device for monitoring the charging pile is provided near each charging pile, such as a monitoring camera for collecting video for monitoring. This type of monitoring camera can be directly connected to the control center for communication to directly transmit the collected video data to the control center for processing, or it can be connected to the control center for communication through the charging pile, and the collected video data is first transmitted to the charging pile, and then transmitted to the control center by the charging pile according to a certain time period or in real time.

[0052] Furthermore, the video data collected by the surveillance camera can be a video of the environment where the charging pile is located, or a video of a pre-set specific location, such as the state of the charging gun being in a hanging gun state when not charging, the side charging state when charging, and the abnormal state of being placed on the ground. The control center receives the transmitted video data of this type and processes it as the current video data of the car charging gun. Specifically, an extraction time is pre-set, and according to the preset cycle formed by the extraction time, the current video data such as the environment video where the charging pile is located or the video of the specific location is extracted. In addition, in order to detect the abnormality of the position of the charging gun, the video image extraction of the environment where the charging pile is located is to extract the image related to the charging gun from it. The extracted image is used as the current state image to reflect the current state of the charging gun.

[0053] Step S20, identifying the multiple current state images based on a preset recognition model to determine whether the vehicle charging gun is in an abnormal position, wherein the preset recognition model is trained based on pre-collected sample data.

[0054] Furthermore, a preset recognition model is pre-set, and the preset recognition model is formed by training with pre-collected sample data of the car charging gun. The preset recognition model is used to recognize multiple current state images extracted to determine whether the car charging gun is in an abnormal position. Among them, the charging gun is in the hanging gun position and the charging position is a normal position, while the ground position is an abnormal position. Therefore, whether the car charging gun is in an abnormal position can be determined by judging whether the car charging gun is in the ground position through the recognition of the preset recognition model. Specifically, the step of recognizing the current state image based on the preset recognition model and judging whether the car charging gun is in an abnormal position includes:

[0055] Step S21, identifying the current state image based on a preset recognition model, obtaining current state information of the vehicle charging gun, and determining the information type of the current state information;

[0056] Step S22, if the information type is a ground type, determining that the vehicle charging gun is in an abnormal position;

[0057] Step S23: if the information type is a hanging gun type or a charging type, it is determined that the vehicle charging gun is in a non-abnormal position, and the status information of the vehicle charging gun is updated based on the non-abnormal position.

[0058] Furthermore, a plurality of current state images are identified by a preset recognition model to obtain the current state information of the vehicle charging gun, and determine the information type of the current state information. The current state information can be represented by numbers, for example, 0, 1, and 2 are used to represent the gun hanging position, the charging position, and the ground position, respectively. If the preset recognition model identifies that the vehicle charging gun is in the gun hanging position, the generated current state information is 0; if the recognition determines that the vehicle charging gun is in the charging position, the generated current state information is 1; if the recognition determines that the vehicle charging gun is in the ground position, the generated current state information is 2. The information type of the generated current state information can then be determined based on the specific value of the generated current state information.

[0059] Further, the information type is divided into ground type, hanging gun type and charging type. If it is determined that the information type of the current status information is the ground type, it means that the car charging gun may be damaged when placed on the ground, and thus the car charging gun is determined to be in an abnormal position. On the contrary, if it is determined that the information type is the hanging gun type or the charging type, it means that the car charging gun has no risk of damage, and thus it is determined to be in a non-abnormal position. At this time, the status information of the car charging gun is updated according to the change between the non-abnormal position and the position at the previous moment. That is, it is determined whether the charging gun state reflected by the position at the previous moment and the charging gun state reflected by the non-abnormal position have changed. If there is a change, the status information of the car charging gun is updated, and if there is no change, the status information of the car charging gun is not updated. For example, the non-abnormal position of the car charging gun is the charging position, and the position of the charging gun at the previous moment is the hanging gun position, which means that a user has used the charging gun to charge the electric car, and the state of the charging gun has changed, so the status information of the car charging gun needs to be updated. If the position of the charging gun at the previous moment is also the charging position, it means that the state of the charging gun has not changed, so there is no need to update the status information of the car charging gun.

[0060] Step S30: If the vehicle charging gun is in an abnormal position, a charging gun identifier and first position information corresponding to the vehicle charging gun are obtained, and abnormal prompt information is output based on the charging gun identifier and the first position information.

[0061] Understandably, if the car charging gun is determined to be in an abnormal position, it means that the car charging gun is at risk of damage. At this time, in order to avoid damage to the car charging gun, the charging gun identification and the first position information corresponding to the car charging gun are obtained. Among them, the charging gun identification is an identification that reflects the uniqueness of the car charging gun, and the first position information is the positioning information that reflects the location of the car charging gun. The charging gun identification and the first position information are then generated as abnormal prompt information output, which can be output to a display device connected to the control center, or it can be output to a terminal held by a person who operates or maintains the charging gun or after-sales service, so that the operation and maintenance or after-sales personnel can place the charging gun in the abnormal position to the hanging gun position in time after checking the abnormal prompt information to avoid damage to the car charging gun.

[0062] The detection method of abnormal position of automobile charging gun implemented in this embodiment, after collecting the current video data of automobile charging gun, extracts multiple current state images from the current video data according to a preset period; then recognizes multiple current state images through a preset recognition model formed by training the sample data of automobile charging gun collected in advance, and judges whether the automobile charging gun is in the ground state through recognition; if it is determined that the automobile charging gun is in the ground state, it means that the automobile charging gun is in an abnormal position; then obtains the charging gun identification and first position information corresponding to the automobile charging gun, and outputs abnormal prompt information based on the charging gun identification and first position information. In this way, by collecting the current real-time video data of the automobile charging gun, extracting the image reflecting the real-time state of the automobile charging gun, and judging whether the automobile charging gun is in an abnormal position by the preset recognition model, once it is detected to be in an abnormal position, the charging gun identification and the first position information of the automobile charging gun in the abnormal position are formed as abnormal prompt information output, so as to timely find the automobile charging gun in the abnormal position and avoid damage to the automobile charging gun. In addition, the preset recognition model is formed by training a large amount of pre-collected car charging gun video data, which can make the detection of abnormal positions of car charging guns more accurate.

[0063] For further information, please refer to Figure 2 Based on the first embodiment of the method for detecting the abnormal position of a vehicle charging gun of the present invention, a second embodiment of the method for detecting the abnormal position of a vehicle charging gun of the present invention is proposed.

[0064] The difference between the second embodiment of the method for detecting an abnormal position of a vehicle charging gun and the first embodiment of the method for detecting an abnormal position of a vehicle charging gun is that after the step of outputting abnormal prompt information based on the charging gun identifier and the first position information, the following steps are included:

[0065] Step S40, determining a generation time point of the current state image, and acquiring video data within a preset time period before the generation time point;

[0066] Step S50, determining the car charging gun user corresponding to the current state image according to the video data, and searching for a first type of account corresponding to the car charging gun user;

[0067] Step S60, determining whether there is a historical abnormality mark for the first type of account, and if there is the historical abnormality mark, obtaining the historical abnormality count corresponding to the historical abnormality mark, and updating the historical abnormality count;

[0068] Step S70, comparing the historical abnormal number with each preset abnormal number interval, determining a target abnormal number interval corresponding to the historical abnormal number, and determining target restriction information corresponding to the target abnormal number interval according to the correspondence between each preset abnormal number interval and restriction information;

[0069] Step S80: adding the target restriction information to the first type of account, wherein the target restriction information includes at least one of information on limiting the number of uses and information on increasing the deposit amount.

[0070] After outputting the abnormal prompt information, this embodiment determines the generation time point of the current state image, and obtains the video data of the preset time period before the generation time point. The preset duration can be pre-set according to actual needs, for example, set to 5 minutes. If the determined generation time point is 11:30 am, the video data between 11:25 and 11:30 is obtained. Then the video data is identified to determine the portrait therein. The portrait is the user who uses the car charging gun to charge and places the car charging gun on the ground. It can be determined as the car charging gun user corresponding to the current state image. Then the user account is searched through face recognition of the portrait as the first type of account corresponding to the car charging gun user.

[0071] Understandably, this may not be the first time that the user has placed the car charging gun on the ground, and each time the operation of placing it on the ground is detected, it will be recorded. After determining the first type of account, it is judged whether the first type of account carries a historical abnormality mark. If it carries a historical abnormality mark, it means that the user has placed the car charging gun on the ground after charging. Therefore, the number of historical abnormalities corresponding to the historical abnormality mark is obtained, and the number of historical abnormalities indicates the number of times the user has placed the car charging gun on the ground, and the number of historical abnormalities is updated based on the current placement behavior. The update method is to accumulate the value 1 based on the number of historical abnormalities. The updated number of historical abnormalities is stored for subsequent search and use.

[0072] Furthermore, in order to punish users who place the car charging gun on the ground, the correspondence between the abnormal number interval and the restriction information is preset. The larger the value corresponding to the preset abnormal number interval, the more restrictions corresponding to the restriction information. Compare the updated historical abnormal number with each preset abnormal number interval, find the preset abnormal number interval into which the updated historical abnormal number falls, and use the found preset abnormal number interval as the target abnormal number interval. Then, according to the correspondence between the preset abnormal number interval and the restriction information, find the restriction information corresponding to the target abnormal number interval as the target restriction information, and add the target restriction information to the first category account. Among them, the added target restriction information includes but is not limited to limiting the number of uses of the user corresponding to the first category account, increasing the deposit amount, and increasing the usage fee. Moreover, when the historical abnormal number exceeds the upper limit of each preset abnormal number interval, indicating that the user has placed the charging gun on the ground for a large number of times, the user can be directly restricted from using it within a certain period of time. In this way, the user's behavior of placing the car charging gun on the ground is regulated to avoid damage to the car charging gun.

[0073] Furthermore, considering that it is difficult for operation and maintenance or after-sales personnel to handle the abnormal prompt information in time after it is output, a mechanism is set up to ask other users to handle the car charging gun placed on the ground. Specifically, the step of outputting the abnormal prompt information based on the charging gun identifier and the first location information includes:

[0074] Step S90, determining whether there is a charging user in a neighboring area within a preset range from the first location information, and if there is the charging user, sending a help message to the second type account corresponding to the charging user;

[0075] Furthermore, a preset range is pre-set, and the preset range preferably exists in the form of a radius, so as to form a circular proximity area with the position of the car charging gun placed on the ground as the center and the preset range as the radius. The proximity area is identified to determine whether there is a charging user therein. For example, if it is determined that the first position information of the car charging gun is position A and the preset range is 100 meters, the proximity area formed is a 100-meter circular area with position A as the center. If it is determined that there is a charging user in the proximity area, a help message is sent to the second type of account of the charging user, so that the charging user can put the charging gun placed on the ground back to the hanging gun position after checking the help message. Specifically, if the charging user exists, the step of sending a help message to the second type of account corresponding to the charging user includes:

[0076] Step a1, receiving feedback information sent based on the help information, and determining second location information corresponding to the feedback information;

[0077] Step a2, generating path information based on the first location information and the second location information, and generating confirmation information based on the user name of the charging user;

[0078] Step a3: sending the path information and the confirmation information to the second type of account of the charging user, so as to guide the charging user based on the path information and the confirmation information.

[0079] Furthermore, after viewing the help-seeking information, if the charging user has the intention to respond to the help-seeking information, he / she will send feedback information through his / her second type of account based on the help-seeking information. After receiving the feedback information, the control center determines the second location information corresponding to the feedback information, that is, the location information of the charging user. When sending feedback information, the charging user may add the location information of the charging user to the feedback information and send it together. Alternatively, after receiving the feedback information, the control center identifies that if the location information is not carried, it will send a request to the charging user to obtain the location information, and the charging user will send the location information of the charging user to the control center based on the request.

[0080] Furthermore, the control center uses the obtained location information as the second location information corresponding to the feedback information, and generates path information based on the first location information and the second location information. The path information represents the path of the charging user from his location to the location of the car charging gun. In addition, the feedback information carries the user name of the charging user, which reflects the uniqueness of the charging user. The user name is generated as confirmation information and sent to the second type of account of the charging user together with the path information. The charging user is guided through the path information, and the charging user is confirmed through the confirmation information.

[0081] Understandably, there may be multiple charging users. After receiving the help message, multiple charging users may send feedback information, so that the control center receives multiple feedback information, and then can determine multiple second location information corresponding to the feedback information, and generate multiple path information. At this time, it is necessary to determine relatively close path information from multiple path information, so as to select charging users who are closer to the car charging gun placed on the ground, and realize rapid processing of the car charging gun. Specifically, the step of generating path information based on the first location information and the second location information includes:

[0082] Step a4, determining whether there are multiple path information, and if there are multiple path information, obtaining the path distances corresponding to the respective path information;

[0083] Step a5, comparing the path distances to determine the target path information having the shortest distance among the plurality of path information;

[0084] Step a6, searching for a charging user corresponding to the target path information from the charging users, and executing a step of generating confirmation information based on the user name of the charging user.

[0085] Furthermore, after the path information is generated based on the first position information and the second position information, it is determined whether the number of generated path information is multiple. If there are multiple, it means that multiple charging users have provided feedback on the help information. At this time, the path distance of each path information is obtained, and the path distances are compared to determine the shortest path distance. Then, the path information from which the shortest path distance comes is searched as the target path information. The second position information for producing the target path information and the charging user corresponding to the second position information are searched. The location of the charging user is closest to the location of the car charging gun placed on the ground, so the user name of the charging user is obtained to generate confirmation information, and then the confirmation information and the target path information are sent to the charging user to confirm and guide the charging user.

[0086] Step S100, when it is detected that the state of the car charging gun changes based on the help information, reward information is added to the second type of account corresponding to the charging user, and the reward information at least includes any one of increasing the number of uses and reducing the deposit amount.

[0087] Understandably, after the charging user processes the car charging gun on the ground based on the help request, the position status of the car charging gun will change. The surveillance camera continuously shoots video data related to the car charging gun and transmits it to the control center. The control center identifies the state of the car charging gun through such video data, and compares the identified state with the state on the ground in the previous period to determine whether the position status of the car charging gun has changed. If it has changed, it means that the charging user has picked up the car charging gun from the ground and placed it in the hanging gun position. The state of the car charging gun has changed based on the help request and is already in a non-abnormal state. In order to encourage charging users to give feedback on the help request, after detecting that the state of the car charging gun has changed based on the help request, the account corresponding to the charging user is used as a second-class account, and reward information is added to the second-class account, such as reducing the deposit, increasing the number of uses, and giving away the number of charging times. This type of reward is used to encourage more charging users to quickly process the car charging guns placed on the ground, thereby reducing the operation and maintenance costs of the car charging guns placed on the ground.

[0088] On the one hand, this embodiment restricts users who place the car charging gun on the ground to avoid the situation where the charging gun is placed on the ground multiple times in the future. On the other hand, it finds charging users in the vicinity of the charging gun placed on the ground and sends a message for help to the charging users, so as to find the nearest charging user to quickly handle the car charging gun placed on the ground, place the car charging gun in the hanging gun position, avoid damage to the car charging gun, and reduce operation and maintenance costs.

[0089] For further information, please refer to Figure 3 Based on the first and second embodiments of the method for detecting the abnormal position of a vehicle charging gun of the present invention, a third embodiment of the method for detecting the abnormal position of a vehicle charging gun of the present invention is proposed.

[0090] The third embodiment of the method for detecting an abnormal position of a vehicle charging gun is different from the first and second embodiments of the method for detecting an abnormal position of a vehicle charging gun in that the step of collecting current video data of the vehicle charging gun includes:

[0091] Step S110, obtaining sample data corresponding to the vehicle charging gun, the sample data including a gun hanging image sample, a charging image sample and a ground placement image sample;

[0092] Step S120, training a preset initial model based on the sample data, and generating the preset initial model as a preset recognition model when the training reaches a preset end condition.

[0093] In this embodiment, before collecting the video data of the car charging gun, a preset recognition model for image recognition is formed through training. Specifically, the image data of the charging guns in a specific position in multiple charging piles are obtained as sample data corresponding to the car charging gun. In addition, such image data can be obtained through a monitoring camera, and the specific position at least includes the hanging gun position, the charging position and the ground position. The monitoring camera is positioned and adjusted, for example, the monitoring camera is installed at the edge position of the top of the charging pile box, and then the camera is calibrated to capture the video data of the charging gun in the hanging gun position, the charging position and the ground position, so that the camera can accurately capture the video data of the charging gun in the hanging gun position, the charging position and the ground position. The camera is controlled to collect videos of the hanging gun position, the charging position and the ground position of the charging gun under different lighting conditions, different background colors and different weather conditions. The hanging gun image, the charging image and the ground placement image are extracted from the collected video, and the corresponding labels are added to each type of image to form sample data containing hanging gun image samples, charging image samples and ground placement image samples.

[0094] Furthermore, the sample data is transmitted to a preset initial model, and the preset initial model is trained. In addition, a preset end condition indicating the completion of the training is preset, and it is determined whether the training reaches the preset end condition. If the preset end condition is reached, the preset initial model is generated as a preset recognition model that can identify the charging gun's hanging position, charging position, and ground position. Specifically, the step of training the preset initial model based on the sample data and generating the preset initial model as a preset recognition model when the training reaches the preset end condition includes:

[0095] Step S121, training the preset initial model based on the training samples in the sample data, and after the training reaches a preset number of times, adjusting the model parameters of the preset initial model based on the parameter adjustment samples in the sample data;

[0096] Step S122, training the adjusted preset initial model based on the training sample, and after the training reaches a preset number of times, testing the preset initial model based on the test sample in the sample data to obtain a test result;

[0097] Step S123, based on the test result, determining whether the training of the preset initial model has reached a preset end condition, and if so, generating the preset initial model as a preset recognition model;

[0098] Step S124, if the preset end condition is not reached, the step of adjusting the model parameters of the preset initial model based on the parameter adjustment sample in the sample data is executed until the training of the preset initial model reaches the preset end condition.

[0099] Furthermore, the sample data is randomly divided into training samples, parameter adjustment samples and test samples according to a certain ratio, such as 4:4:2, or 5:3:2, etc. The preset initial model is first trained by the training samples, and after the number of training reaches the preset number of times, the model parameters of the preset initial model are adjusted by the parameter adjustment samples. Among them, the preset number of times can be set in advance according to the needs. Thereafter, the preset initial model after parameter adjustment is continuously trained with the training samples, and after the number of training reaches the preset number of times again, the preset initial model is tested by the test samples to obtain the test results. The test results are compared with the preset end conditions to determine whether the training of the preset initial model reaches the preset end conditions. Among them, the test result can be a loss function value, and the corresponding preset end condition is a numerical value. The loss function value generated by the test is compared with the numerical value to determine whether the loss function value is less than the numerical value. If it is less than the numerical value, it means that the preset initial model has achieved the required effect after training, and it can be determined that the training of the preset recognition model has reached the preset end condition, so the preset initial model can be used as the preset recognition model. On the contrary, if it is determined that the training of the preset initial model has not yet reached the preset end condition, the model parameters of the preset initial model will continue to be adjusted through the parameter adjustment sample, and after the adjustment, the training will continue to be performed according to the training sample until the training of the preset initial model reaches the preset end condition after testing, and the preset initial model will be generated as the preset recognition model.

[0100] This embodiment divides the collected gun hanging sample images, charging sample images and ground placement image samples into training samples, parameter adjustment samples and test samples, trains the preset initial model, and obtains a preset recognition model that can identify the gun hanging position, charging position and ground position of the charging gun. A large amount of sample data, as well as the division of training samples, parameter adjustment samples and test samples, enables the preset recognition model to accurately identify the charging gun position status image, thereby making the detection of abnormal position of the vehicle charging gun more accurate.

[0101] In addition, the embodiment of the present invention also provides a detection system for abnormal position of a vehicle charging gun. Figure 4 , Figure 4 It is a structural schematic diagram of the equipment hardware operating environment involved in the embodiment of the detection system for abnormal position of a vehicle charging gun of the present invention.

[0102] like Figure 4As shown, the detection system for the abnormal position of the car charging gun may include: a processor 1001, such as a CPU, a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display), an input unit such as a keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface). The memory 1005 may be a high-speed RAM memory, or it may be a stable memory (non-volatile memory), such as a disk memory. The memory 1005 may optionally be a storage device independent of the aforementioned processor 1001.

[0103] Those skilled in the art will understand that Figure 4 The hardware structure of the system for detecting abnormal position of a vehicle charging gun shown in the figure does not constitute a limitation of the system for detecting abnormal position of a vehicle charging gun, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.

[0104] like Figure 4 As shown, the memory 1005 as a storage medium may include an operating system, a network communication module, a user interface module and a control program. The operating system is a program for managing and controlling the detection system and software resources of the abnormal position of the vehicle charging gun, and supports the operation of the network communication module, the user interface module, the control program and other programs or software; the network communication module is used to manage and control the network interface 1004; the user interface module is used to manage and control the user interface 1003.

[0105] exist Figure 4 In the hardware structure of the detection system for abnormal position of a vehicle charging gun shown in the figure, the network interface 1004 is mainly used to connect to the background server and communicate data with the background server; the user interface 1003 is mainly used to connect to the client (user end) and communicate data with the client; the processor 1001 can call the control program stored in the memory 1005 and perform the following operations:

[0106] Collecting current video data of the vehicle charging gun, and extracting a plurality of current state images from the current video data based on a preset period;

[0107] Recognize the multiple current state images based on a preset recognition model to determine whether the vehicle charging gun is in an abnormal position, wherein the preset recognition model is trained based on pre-collected sample data;

[0108] If the vehicle charging gun is in an abnormal position, a charging gun identifier and first position information corresponding to the vehicle charging gun are obtained, and abnormal prompt information is output based on the charging gun identifier and the first position information.

[0109] Further, after the step of outputting abnormal prompt information based on the charging gun identifier and the first position information, the step includes:

[0110] Determine a generation time point of the current state image, and obtain video data within a preset time length before the generation time point;

[0111] Determine, according to the video data, a car charging gun user corresponding to the current state image, and search for a first type of account corresponding to the car charging gun user;

[0112] Determine whether there is a historical abnormality mark for the first category of accounts; if there is the historical abnormality mark, obtain the number of historical abnormalities corresponding to the historical abnormality mark, and update the number of historical abnormalities;

[0113] Compare the historical number of abnormalities with each preset abnormality number interval to determine a target abnormality number interval corresponding to the historical number of abnormalities, and determine target restriction information corresponding to the target abnormality number interval according to the correspondence between each preset abnormality number interval and restriction information;

[0114] The target restriction information is added to the first type of account, where the target restriction information includes at least one of information on limiting the number of uses and information on increasing the deposit amount.

[0115] Further, after the step of outputting abnormal prompt information based on the charging gun identifier and the first position information, the step includes:

[0116] Determine whether there is a charging user in a neighboring area within a preset range of the first location information, and if there is the charging user, send a help message to the second type account corresponding to the charging user;

[0117] When it is detected that the state of the car charging gun changes based on the help information, reward information is added to the second type of account corresponding to the charging user, and the reward information at least includes any one of increasing the number of uses and reducing the deposit amount.

[0118] Furthermore, if the charging user exists, the step of sending a help message to the second type of account corresponding to the charging user includes:

[0119] receiving feedback information sent based on the help information, and determining second location information corresponding to the feedback information;

[0120] Generate path information based on the first location information and the second location information, and generate confirmation information based on the user name of the charging user;

[0121] The path information and the confirmation information are sent to the second type of account of the charging user, so as to guide the charging user based on the path information and the confirmation information.

[0122] Further, the step of generating path information based on the first location information and the second location information includes:

[0123] Determine whether there are multiple path information, and if there are multiple path information, obtain the path distances corresponding to each of the path information;

[0124] Comparing the path distances, and determining the target path information having the shortest distance among the plurality of path information;

[0125] A charging user corresponding to the target path information is searched from the charging users, and a step of generating confirmation information based on the user name of the charging user is performed.

[0126] Furthermore, the step of identifying the current state image based on a preset recognition model to determine whether the vehicle charging gun is in an abnormal position includes:

[0127] Recognize the current state image based on a preset recognition model, obtain current state information of the vehicle charging gun, and determine the information type of the current state information;

[0128] If the information type is a ground type, it is determined that the vehicle charging gun is in an abnormal position;

[0129] If the information type is a hanging gun type or a charging type, it is determined that the vehicle charging gun is in a non-abnormal position, and the status information of the vehicle charging gun is updated based on the non-abnormal position.

[0130] Furthermore, the step of collecting current video data of the vehicle charging gun includes:

[0131] Acquire sample data corresponding to the vehicle charging gun, the sample data including a gun hanging image sample, a charging image sample, and a ground placement image sample;

[0132] The preset initial model is trained based on the sample data, and when the training reaches a preset end condition, the preset initial model is generated as a preset recognition model.

[0133] Furthermore, the step of training a preset initial model based on the sample data and generating the preset initial model into a preset recognition model when the training reaches a preset end condition includes:

[0134] Training the preset initial model based on the training samples in the sample data, and adjusting the model parameters of the preset initial model based on the parameter adjustment samples in the sample data after the training reaches a preset number of times;

[0135] Training the adjusted preset initial model based on the training samples, and after the training reaches a preset number of times, testing the preset initial model based on the test samples in the sample data to obtain a test result;

[0136] Based on the test result, determining whether the training of the preset initial model has reached a preset end condition, and if the preset end condition has been reached, generating the preset initial model as a preset recognition model;

[0137] If the preset end condition is not reached, the step of adjusting the model parameters of the preset initial model based on the parameter adjustment sample in the sample data is executed until the training of the preset initial model reaches the preset end condition.

[0138] The specific implementation of the abnormal position detection system of the vehicle charging gun of the present invention is basically the same as the various embodiments of the abnormal position detection method of the vehicle charging gun described above, and will not be repeated here.

[0139] The embodiment of the present invention further provides a storage medium having a control program stored thereon, and when the control program is executed by a processor, the steps of the above-mentioned method for detecting an abnormal position of a vehicle charging gun are implemented.

[0140] The storage medium of the present invention may be a computer-readable storage medium, and its specific implementation is substantially the same as the above-mentioned embodiments of the method for detecting an abnormal position of a vehicle charging gun, and will not be described in detail herein.

[0141] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the enlightenment of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims. All equivalent structures or equivalent process changes made using the contents of the specification and drawings of the present invention, or directly or indirectly used in other related technical fields, are protected by the present invention.

Claims

1. A method for detecting an abnormal position of a vehicle charging gun, characterized in that: The method for detecting the abnormal position of the vehicle charging gun includes: Collecting current video data of the vehicle charging gun, and extracting a plurality of current state images from the current video data based on a preset period; Recognize the multiple current state images based on a preset recognition model to determine whether the vehicle charging gun is in an abnormal position, wherein the preset recognition model is trained based on pre-collected sample data; If the vehicle charging gun is in an abnormal position, a charging gun identifier and first position information corresponding to the vehicle charging gun are obtained, and abnormal prompt information is output based on the charging gun identifier and the first position information.

2. The method for detecting abnormal position of a vehicle charging gun according to claim 1, characterized in that: The step of outputting abnormal prompt information based on the charging gun identifier and the first position information includes: Determine a generation time point of the current state image, and obtain video data within a preset time length before the generation time point; Determine, according to the video data, a car charging gun user corresponding to the current state image, and search for a first type of account corresponding to the car charging gun user; Determine whether there is a historical abnormality mark for the first category of accounts; if there is the historical abnormality mark, obtain the number of historical abnormalities corresponding to the historical abnormality mark, and update the number of historical abnormalities; Compare the historical number of abnormalities with each preset abnormality number interval to determine a target abnormality number interval corresponding to the historical number of abnormalities, and determine target restriction information corresponding to the target abnormality number interval according to the correspondence between each preset abnormality number interval and restriction information; The target restriction information is added to the first type of account, where the target restriction information includes at least one of information on limiting the number of uses and information on increasing the deposit amount.

3. The method for detecting abnormal position of a vehicle charging gun according to claim 1, characterized in that: The step of outputting abnormal prompt information based on the charging gun identifier and the first position information includes: Determine whether there is a charging user in a neighboring area within a preset range of the first location information, and if there is the charging user, send a help message to the second type account corresponding to the charging user; When it is detected that the state of the car charging gun changes based on the help information, reward information is added to the second type of account corresponding to the charging user, and the reward information at least includes any one of increasing the number of uses and reducing the deposit amount.

4. The method for detecting abnormal position of a vehicle charging gun according to claim 3, characterized in that: If the charging user exists, the step of sending a help message to the second type of account corresponding to the charging user includes: receiving feedback information sent based on the help information, and determining second location information corresponding to the feedback information; Generate path information based on the first location information and the second location information, and generate confirmation information based on the user name of the charging user; The path information and the confirmation information are sent to the second type of account of the charging user, so as to guide the charging user based on the path information and the confirmation information.

5. The method for detecting abnormal position of a vehicle charging gun according to claim 4, characterized in that: The step of generating path information based on the first location information and the second location information comprises: Determine whether there are multiple path information, and if there are multiple path information, obtain the path distances corresponding to each of the path information; Comparing the path distances, and determining the target path information having the shortest distance among the plurality of path information; A charging user corresponding to the target path information is searched from the charging users, and a step of generating confirmation information based on the user name of the charging user is performed.

6. The method for detecting an abnormal position of a vehicle charging gun according to any one of claims 1 to 5, characterized in that: The step of identifying the current state image based on a preset recognition model to determine whether the vehicle charging gun is in an abnormal position includes: Recognize the current state image based on a preset recognition model, obtain current state information of the vehicle charging gun, and determine the information type of the current state information; If the information type is a ground type, it is determined that the vehicle charging gun is in an abnormal position; If the information type is a hanging gun type or a charging type, it is determined that the vehicle charging gun is in a non-abnormal position, and the status information of the vehicle charging gun is updated based on the non-abnormal position.

7. The method for detecting an abnormal position of a vehicle charging gun according to any one of claims 1 to 5, characterized in that: The step of collecting current video data of the vehicle charging gun includes: Acquire sample data corresponding to the vehicle charging gun, the sample data including a gun hanging image sample, a charging image sample, and a ground placement image sample; The preset initial model is trained based on the sample data, and when the training reaches a preset end condition, the preset initial model is generated as a preset recognition model.

8. The method for detecting abnormal position of a vehicle charging gun according to claim 7, characterized in that: The step of training the preset initial model based on the sample data and generating the preset initial model into a preset recognition model when the training reaches a preset end condition comprises: Training the preset initial model based on the training samples in the sample data, and adjusting the model parameters of the preset initial model based on the parameter adjustment samples in the sample data after the training reaches a preset number of times; Training the adjusted preset initial model based on the training samples, and after the training reaches a preset number of times, testing the preset initial model based on the test samples in the sample data to obtain a test result; Based on the test result, determining whether the training of the preset initial model has reached a preset end condition, and if the preset end condition has been reached, generating the preset initial model as a preset recognition model; If the preset end condition is not reached, the step of adjusting the model parameters of the preset initial model based on the parameter adjustment sample in the sample data is executed until the training of the preset initial model reaches the preset end condition.

9. A detection system for abnormal position of a vehicle charging gun, characterized in that: The abnormal position detection system of the vehicle charging gun includes: a memory, a processor, a communication bus and a control program stored in the memory: The communication bus is used to realize the connection and communication between the processor and the memory; The processor is used to execute the control program to implement the steps of the method for detecting an abnormal position of a vehicle charging gun according to any one of claims 1 to 8.

10. A storage medium, characterized in that: The storage medium stores a control program, and when the control program is executed by the processor, the steps of the method for detecting the abnormal position of the vehicle charging gun according to any one of claims 1 to 8 are implemented.

Images