Charging pile occupation judgment method and device and medium

Abstract

The invention relates to the field of new energy automobile charging, in particular to a charging pile occupation judgment method and device and a medium. The method comprises the steps of firstly collecting a vehicle image, then carrying out vehicle recognition, carrying out license plate positioning and character recognition on a vehicle with a license plate by adopting a deep learning technology, and then judging whether the vehicle is a new energy vehicle or not according to a license plate rule. For a vehicle without a license plate, the characteristics of a vehicle model, a vehicle logo, a vehicle window shape, a vehicle lamp design, a vehicle body color, an air inlet grille and the like are analyzed through an image recognition technology, and matching is performed in combination with a large database, so that whether the vehicle is a new energy vehicle or not is judged. And the control module matches the vehicle identification result with the charging pile state information, and if the matching result shows that the vehicle is a non-charging vehicle and the charging pile is in an idle state, it is judged that the vehicle is occupied. The charging pile can intelligently identify and distinguish the charging vehicle and the non-charging vehicle, and can remind the illegal space occupying behavior and carry out billing, so as to optimize the use efficiency of the charging pile and the user experience.

Description

Technical Field

[0001] The present invention relates to the field of new energy vehicle charging, and in particular to a method, a device and a medium for determining whether a charging pile is occupied by a non-charging vehicle. Background Art

[0002] With the popularity of new energy vehicles, charging piles have become an important part of urban infrastructure. However, the problem of charging piles being occupied by non-charging vehicles is becoming increasingly serious, which not only reduces the efficiency of charging piles, but also brings inconvenience to car owners who need to charge.

[0003] In the prior art, the problem of charging pile occupancy is usually solved by manual inspection or simple ground lock system. These methods have the following problems: (1) Manual inspections require a lot of human resources and cannot achieve 24-hour uninterrupted monitoring.

[0004] (2) The installation and maintenance of the ground lock system requires additional costs, and these costs will increase further as the number of charging piles increases.

[0005] (3) The data value of charging piles (such as charging times, charging amount, user behavior, etc.) has not been fully utilized. Summary of the invention

[0006] In view of the defects of the prior art, the present invention provides a charging pile occupancy judgment method, device and medium, which realizes the automatic monitoring and management of the charging pile occupancy by integrating camera monitoring, vehicle identification, intelligent communication and intelligent matching algorithm. The system can intelligently identify and distinguish charging vehicles from non-charging vehicles, remind illegal occupation behaviors and implement billing, so as to optimize the efficiency of charging pile use and user experience.

[0007] In order to solve the technical problem, the technical solution adopted by the present invention is: a charging pile occupancy judgment method, comprising the following steps: S01, the camera collects the image of the vehicle driving into the parking space where the charging pile is located, and transmits the vehicle image to the vehicle identification module; S02, the vehicle identification module analyzes the received vehicle image in real time, and extracts the license plate number and other vehicle features, and obtains the vehicle identification result based on the license plate number and other vehicle features. The vehicle identification result is the type of vehicle driving into the parking space where the charging pile is located, and the vehicle type includes new energy vehicles and non-new energy vehicles; S03, uploading the status information of the charging pile to the control module, the status information including idle state, fault state, charging, charging completed, and maintenance state; S04. The control module makes an occupancy judgment based on the vehicle identification result obtained by the vehicle identification module and the charging pile status information, and matches the vehicle identification result with the charging pile status information. If the vehicle type is not a new energy vehicle, the charging status information is not considered and it is directly judged as a charging pile occupancy. If the vehicle type is a new energy vehicle, the charging pile status is idle, faulty, under maintenance, or exceeds the preset charging completion time, it is judged as a charging pile occupancy. If the charging pile status is charging or within the preset charging completion time, it is judged as a charging pile not occupancy.

[0008] Furthermore, the vehicle recognition module analyzes the vehicle image in the following process: S211, license plate positioning, using a deep learning model to locate the license plate. If the license plate is located, the deep learning model outputs the confidence and bounding box coordinates of the license plate area, where the license plate positioning function is: , in fdetect is a function used to detect the location of the license plate in the vehicle image. is the vehicle image, is the bounding box combination, are the parameters of the deep learning model, including weights, biases, and optimizer parameters; S212, character segmentation, using a projection-based method to perform character segmentation on the located license plate area, the character segmentation function is: , in fsegment Is a function used to perform character segmentation on the located license plate area. is the set of characters after segmentation; S213, character recognition, use the trained convolutional neural network to recognize each segmented character, the character recognition function is: , in frecognize is a function used to identify the character corresponding to each character image. is the set of recognized characters, are the parameters of the convolutional neural network, including weights, biases, pooling layer parameters, and normalization parameters; S214, judging the new energy license plate, judging whether the recognized characters are new energy license plates according to the format and character features of the license plate number, and the new energy license plate judging formula is: , in fjudge It is a function used to determine whether the license plate is a new energy license plate based on the format and character characteristics of the license plate number. is the license plate type tag, Equal to 1 means a new energy license plate, Equal to 0 means it is not a new energy vehicle license plate.

[0009] Furthermore, the process of the license plate recognition module analyzing the vehicle image is as follows: S221, license plate location, use the deep learning model to locate the license plate. If the license plate is not located, collect other features of the vehicle, including vehicle model, logo, window shape, headlight design, body color, and air intake grille. The vehicle feature extraction formula is: , in fextract is a function used to extract key features from vehicle images. These features help identify and distinguish different vehicles. I is the vehicle image and F is the set of extracted vehicle features. S222, feature matching, matching the extracted vehicle features with the feature templates in the database to determine whether the vehicle is a new energy vehicle. The feature matching formula is: , in fmatch is a function used to match the extracted vehicle features with the feature templates in the database and calculate the matching score. M is the matching score and T is the feature template set in the database. S223. Determine whether the vehicle is a new energy vehicle based on the matching score. The new energy vehicle model determination formula is: , in fclassify It is a function used to determine whether a vehicle is a new energy vehicle based on the matching score and other vehicle features. V is the vehicle type label. A value of 1 indicates a new energy vehicle license plate, and a value of 0 indicates a non-new energy vehicle license plate.

[0010] Furthermore, multiple cameras with different acquisition angles are deployed around the charging pile, and the images collected by the multiple cameras are fused and transmitted to the vehicle recognition module.

[0011] Furthermore, an adaptive learning rate algorithm is used during the training of the vehicle recognition module to accelerate the convergence of the model. The formula of the adaptive learning rate algorithm is: , in , are the model parameters at time steps t and t+1, respectively. is the adaptive learning rate, is the loss function.

[0012] Furthermore, distance sensors are deployed around the charging piles. When the distance sensor senses that a vehicle is approaching a preset area and the distance between the vehicle and the distance sensor is less than a preset value, the camera is started to capture images.

[0013] Furthermore, the result of step S05 is displayed, and a warning message is sent to the owner of the occupied vehicle through a display module, and the display module includes a display screen on the charging pile and a mobile application.

[0014] Furthermore, it also includes step S06 billing. When the occupancy behavior is determined, the occupancy fee billing system is started. When the occupancy time exceeds the preset time, the occupancy fee billing system is started through the billing module. The billing module calculates the occupancy fee based on the preset rate and occupancy time by multiplying the two.

[0015] The present invention also discloses a charging pile occupancy judgment device, comprising a processor and a memory storing program instructions, wherein the processor is configured to execute the charging pile occupancy judgment method as described above when running the program instructions.

[0016] A storage medium stores program instructions, and when the program instructions are run, the charging pile occupancy judgment method as described above is executed.

[0017] The beneficial effects of the present invention include: (1) The utilization efficiency of charging piles has been improved, and the occupancy of non-charging vehicles has been reduced.

[0018] (2) It realizes automatic and intelligent judgment of charging pile occupancy, reducing the need for manual inspections.

[0019] (3) Through warning and billing mechanisms, the occupation of non-charging vehicles is effectively prevented.

[0020] (4) Improved user experience and provided more convenient and efficient charging services for new energy vehicle owners. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 is a flow chart of the method described in Example 1; Figure 2 This is a functional block diagram of the device described in Example 3; Figure 3 This is a functional block diagram of the device described in Example 4. DETAILED DESCRIPTION

[0022] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.

[0023] Example 1 This embodiment discloses a method for determining the occupancy of a charging pile. Figure 1As shown, the following steps are included: S01. Charging pile monitoring is achieved through a monitoring module. The monitoring module uses a camera and a distance sensing sensor to achieve comprehensive monitoring of the charging pile parking space. When the distance sensing sensor senses that a vehicle is approaching a preset area, and the distance between the vehicle and the distance sensor is less than or equal to the preset distance, the camera is started to monitor the vehicle. When the vehicle enters the corresponding parking space, the camera automatically captures the vehicle image. The camera is fixedly installed above or around the charging pile to ensure comprehensive monitoring of the charging pile parking space, and the camera module has night vision function and waterproof and dustproof design to ensure stable operation under severe weather conditions. The monitoring module also has a self-detection function, which can promptly detect and report potential camera and distance sensor failures.

[0024] S02, vehicle identification, is achieved through the vehicle identification module, which analyzes the vehicle images captured by the monitoring module in real time and accurately extracts the license plate number and other vehicle features. The module integrates advanced image stabilization technology and image enhancement algorithms to adapt to different lighting conditions and weather conditions. In addition, the module will support multiple trigger mechanisms, such as motion detection or preset time intervals, to achieve more flexible data collection.

[0025] This part can have two situations: with a license plate number and without a license plate number. This embodiment describes the situation with a license plate number. For the license plate recognition algorithm, deep learning technology will be used to achieve fast and accurate license plate positioning and character recognition. Through training with a large amount of license plate image data, the model can recognize license plate numbers in different lighting, angles and backgrounds. Then, according to the license plate rules, it is determined whether it is a new energy vehicle. The process of vehicle recognition is: S211, license plate positioning, using a deep learning model to locate the license plate. If the license plate is located, the deep learning model outputs the confidence and bounding box coordinates of the license plate area, where the license plate positioning function is: , in fdetect is a function used to detect the location of the license plate in the vehicle image. is the vehicle image, is the bounding box combination, These are the parameters of the deep learning model, including weights, biases, and optimizer parameters.

[0026] In this embodiment, the license plate positioning function fdetect Use a deep learning model to identify the license plate area in the image and output the confidence and bounding box coordinates of each identified area. Use Python and OpenCV library to implement the specific process as follows: 1. Import libraries: Import OpenCV for image processing, NumPy for array operations, and TensorFlow for loading and using deep learning models.

[0027] 2. Load model: Load a pre-trained deep learning model that can recognize the license plate area in the image.

[0028] 3. Define the license plate detection function: fdetect: The main function, which calls parse_model_output and filter_boxe to read the image, preprocess the image, use the model to make predictions, and parse the output.

[0029] parse_model_output: parses the output of the model, extracting confidence and bounding box coordinates.

[0030] filter_boxes: Filter bounding boxes based on a confidence threshold, keeping only those with higher confidence.

[0031] 4. Use the function: Call the fdetect function with the actual image path and model parameters and print the detected bounding box.

[0032] S212, character segmentation, using a projection-based method to perform character segmentation on the located license plate area, the character segmentation function is: , in fsegment Is a function used to perform character segmentation on the located license plate area. Is the separated character set.

[0033] Character segmentation is a key step in the license plate recognition process. It further subdivides the located license plate area into individual characters. It is implemented through Python code. The specific process is as follows: 1. Import libraries: Import OpenCV and NumPy libraries for image processing and array operations.

[0034] 2. Define the character segmentation function: fsegment: The main function, which calls extract_plate_from_box, project_and_split, find_peaks, traverses each bounding box, extracts the license plate area, and uses the projection method for character segmentation.

[0035] extract_plate_from_box: Extracts the license plate region image from a bounding box.

[0036] project_and_split: Use the horizontal projection method to perform character segmentation on the license plate area.

[0037] find_peaks: Finds peaks in a signal, used to determine character boundaries.

[0038] 3. Use the function: Call the fsegment function with the bounding box set and the original image, and display the segmented characters.

[0039] S213, character recognition, use the trained convolutional neural network to recognize each segmented character, the character recognition function is: , in frecognize is a function used to identify the character corresponding to each character image. is the set of recognized characters, These are the parameters of the convolutional neural network, including weights, biases, pooling layer parameters, and normalization parameters.

[0040] In this embodiment, a trained convolutional neural network (CNN) is used to recognize each segmented character. The specific process is as follows: 1. Import libraries: Import NumPy for array operations and TensorFlow for loading and using deep learning models.

[0041] 2. Load model: Load a pre-trained character recognition model that can recognize single characters.

[0042] 3. Define the character recognition function: frecognize: The main function, which calls preprocess_char, char2gray, parse_prediction, traverses each character image, performs preprocessing, uses the model to make predictions, and parses the output.

[0043] preprocess_char: Preprocess character images.

[0044] char2gray: Converts a character image to a grayscale image.

[0045] parse_prediction: parses the model output to get the most likely character.

[0046] 4. Use the function: Call the frecognize function with the character set and model parameters, and print the recognized character set.

[0047] S214, judging the new energy license plate, judging whether the recognized characters are new energy license plates according to the format and character features of the license plate number, and the new energy license plate judging formula is: , in fjudge It is a function used to determine whether the license plate is a new energy license plate based on the format and character characteristics of the license plate number. is the license plate type tag, Equal to 1 means a new energy license plate, Equal to 0 means it is not a new energy vehicle license plate.

[0048] New energy license plates have specific formats and rules, such as the first digit is a Chinese character, the second digit is a letter (new energy license plates are not allowed to use I and O), the third and subsequent digits are letters or numbers, and the first letter of the new energy license plate indicates the type of new energy vehicle, such as pure electric vehicles (D) and plug-in hybrid vehicles (F). This embodiment defines a regular expression regex for matching the format of new energy license plates. This regular expression takes into account the characteristics of new energy license plates, such as the first digit is a Chinese character, the second digit is a specific letter (D or F), and the subsequent combination of numbers and letters. The fjudge function accepts a license plate number string C as input and uses a regular expression to determine whether the license plate conforms to the format of a new energy license plate. If the license plate number conforms to the format of a new energy license plate, the function returns 1, indicating a new energy license plate; otherwise, it returns 0, indicating a non-new energy license plate.

[0049] In this embodiment, multiple cameras with different acquisition angles are deployed around the charging pile, and the images collected by the multiple cameras are fused and transmitted to the vehicle recognition module. The formula for multi-camera data fusion is: , in is a function that integrates data from different cameras to improve the accuracy and robustness of recognition. I is the fused data. It is data from different cameras.

[0050] Multi-camera data fusion is a complex process that involves integrating data from multiple perspectives to obtain a more comprehensive and accurate view of the vehicle. The specific process is: 1. Import libraries: OpenCV for image processing and NumPy for array operations.

[0051] 2. Define the multi-camera data fusion function: ffuse: This function accepts multiple images as input and performs data fusion.

[0052] First, check whether there are enough images to fuse.

[0053] Then, all images are resized to the same size, here we simply choose the size of the first image as the target size.

[0054] Using a simple average fusion method, the pixel values ​​of all images are added together and then divided by the number of images to get the fused image.

[0055] 3. Use functions: read images from different cameras, call the ffuse function, and display the fused image.

[0056] This embodiment is a simplified version, and more complex data fusion algorithms may be required in actual applications, such as feature-based fusion, decision-based fusion, etc. In addition, the selection of fusion methods may need to be adjusted according to specific application scenarios and requirements to obtain the best fusion effect.

[0057] In this embodiment, an adaptive learning rate algorithm is used during the vehicle recognition module training process to accelerate the convergence of the model and improve the recognition accuracy. The adaptive learning rate algorithm formula is: , in , are the model parameters at time steps t and t+1, respectively. is the adaptive learning rate, is the loss function.

[0058] S03. Upload the status information of the charging pile to the control module, the status information including idle state, fault state, charging, charging completed, and maintenance state.

[0059] S04. The control module makes an occupancy judgment based on the vehicle identification result obtained by the vehicle identification module and the charging pile status information, and matches the vehicle identification result with the charging pile status information. If the vehicle type is not a new energy vehicle, the charging status information is not considered and it is directly judged as a charging pile occupancy. If the vehicle type is a new energy vehicle, the charging pile status is idle, faulty, under maintenance, or exceeds the preset charging completion time, it is judged as a charging pile occupancy. If the charging pile status is charging or within the preset charging completion time, it is judged as a charging pile not occupancy.

[0060] The specific judgment logic is: S05, result display, used to display the occupancy judgment result, and send a warning message to the owner of the occupied vehicle through a display module, the display module includes a display screen on the charging pile and a mobile application.

[0061] S06. Billing. When the seat-occupancy behavior is determined, the seat-occupancy fee billing system is started. When the seat-occupancy time exceeds the preset time, the seat-occupancy fee billing system is started through the billing module. The billing module calculates the seat-occupancy fee based on the preset rate and the seat-occupancy time by multiplying the two.

[0062] Example 2 This embodiment discloses a method for judging the occupancy of a charging pile, which is described for the case where the vehicle identification result is no license plate. For vehicles without license plates, the algorithm will use other features of the vehicle, such as vehicle model, vehicle logo, window shape, headlight design, body color, air intake grille, etc. These features can be analyzed through image recognition technology and matched with a large database to determine whether it is a new energy vehicle. The judgment process is as follows: S221, license plate location, use the deep learning model to locate the license plate. If the license plate is not located, extract features from the vehicle image, including vehicle model, logo, window shape, headlight design, body color, and air intake grille. The vehicle feature extraction formula is: , in fextract is a function used to extract key features from vehicle images. These features can help identify and distinguish different vehicles. I is the vehicle image and F is the set of extracted vehicle features.

[0063] Vehicle feature extraction is a complex process that usually involves computer vision and machine learning techniques. We use Python and OpenCV libraries to implement the vehicle feature extraction function. This function will extract features such as vehicle model, logo, window shape, headlight design, body color, and grille from the input image. The implementation process is: 1. Import libraries: Import OpenCV for image processing, NumPy for array operations, and scikit-learn’s KMeans clustering algorithm for color extraction.

[0064] 2. Define vehicle feature extraction function: fextract: The main function, which calls other functions such as extract_body_type_and_logo, extract_window_and_headlight, extract_car_color, extract_grille, etc. to extract different vehicle features.

[0065] extract_body_type_and_logo: This function is used to extract the body type and logo. The actual implementation requires complex image analysis.

[0066] extract_window_and_headlight: This function is used to extract the window shape and headlight design. The actual implementation requires complex image analysis.

[0067] extract_car_color: Extracts the main color from the image using the KMeans clustering algorithm.

[0068] extract_grille: This function is used to extract the grille. The actual implementation requires complex image analysis.

[0069] 3. Use the function: read the image, call the fextract function, and print the extracted features.

[0070] S222, feature matching, matching the extracted vehicle features with the feature templates in the database to determine whether the vehicle is a new energy vehicle. The feature matching formula is: , in fmatch It is a function used to match the extracted vehicle features with the feature templates in the database and calculate the matching score. M is the matching score and T is the set of feature templates in the database.

[0071] In order to implement the feature matching function, it is necessary to compare the extracted vehicle feature F with the feature template set T in the database. The matching score M can be calculated based on the similarity between the features. This embodiment implements feature matching through Python code, and the specific process is: 1. Import libraries: Import scikit-learn’s cosine similarity function for calculating the similarity between features, and NumPy for array operations.

[0072] 2. Define the feature matching function: fmatch: This is the main function, which iterates over each feature template, calculates the cosine similarity between the extracted vehicle features and the template features, and adds the similarity scores to get the match score.

[0073] Cosine similarity is used to calculate the similarity between two vectors, and its value ranges from -1 to 1, and the closer the value is to 1, the more similar it is.

[0074] 3. Use the function: Call the fmatch function using the extracted vehicle feature set and the feature template set in the database, and print the matching score.

[0075] S223. Determine whether the vehicle is a new energy vehicle based on the matching score. The new energy vehicle model determination formula is: , in fclassify It is a function used to determine whether a vehicle is a new energy vehicle based on the matching score and other vehicle features. V is the vehicle type label. A value of 1 indicates a new energy vehicle license plate, and a value of 0 indicates a non-new energy vehicle license plate.

[0076] In order to implement the new energy vehicle model judgment function, we will judge whether the vehicle is a new energy vehicle based on the matching score M. If the matching score exceeds a certain set threshold, the vehicle is considered to be a new energy vehicle. The implementation process is: 1. Define the new energy vehicle model judgment function: fclassify: This function accepts the matching score match_score as input and uses a preset threshold threshold to determine whether the vehicle is a new energy vehicle.

[0077] The threshold is a value between 0 and 1 that determines the matching score. If the matching score is greater than or equal to the threshold, the function returns 1, indicating that the vehicle is a new energy vehicle; otherwise, it returns 0, indicating that the vehicle is not a new energy vehicle.

[0078] 2. Use the function: Call the fclassify function with the matching score and print the vehicle type label.

[0079] The rest is the same as in Example 1 and will not be described again here.

[0080] Through the above methods, new energy vehicles can be effectively identified, regardless of whether they have license plates or not. The design of these algorithms takes into account real-time performance and accuracy and is suitable for various complex environments and conditions.

[0081] Example 3 This embodiment discloses a charging pile occupancy determination device. Figure 2 As shown, it includes monitoring module, vehicle identification module, intelligent communication module, control module, display module and billing module. The specific introduction of each module is as follows: The monitoring module integrates two major parts: a camera and a distance sensing sensor to achieve comprehensive monitoring of the charging pile parking space. When the distance sensing sensor senses that a vehicle is approaching the preset area, and the distance between the vehicle and the distance sensor is less than or equal to the preset distance, the camera is started to monitor the vehicle. When the vehicle enters the corresponding parking space, the camera automatically captures the vehicle image. The camera is fixedly installed above or around the charging pile to ensure comprehensive monitoring of the charging pile parking space, and the camera module has night vision function and waterproof and dustproof design to ensure stable operation under severe weather conditions. The monitoring module also has a self-detection function, which can promptly detect and report potential camera and distance sensor failures.

[0082] The vehicle identification module can analyze the vehicle images captured by the monitoring module in real time and accurately extract the license plate number and other vehicle features. The module integrates advanced image stabilization technology and image enhancement algorithms to adapt to different lighting conditions and weather conditions. In addition, the module will support multiple trigger mechanisms, such as motion detection or preset time intervals, to achieve more flexible data collection.

[0083] For the license plate recognition algorithm, deep learning technology will be used to achieve fast and accurate license plate positioning and character recognition. Through training with a large amount of license plate image data, the model can recognize license plate numbers in different lighting, angles and backgrounds. Then, according to the license plate rules, it can be judged whether it is a new energy vehicle.

[0084] For vehicles without license plates, the algorithm will use other features of the vehicle, such as model, logo, window shape, light design, body color, grille, etc. These features can be analyzed through image recognition technology and matched with a large database to determine whether it is a new energy vehicle.

[0085] The specific process of identifying the vehicle by the vehicle identification module is the same as that in Embodiment 1 and Embodiment 2, and will not be repeated here.

[0086] The intelligent communication module receives status information of the charging pile, where the status information of the charging pile includes the status of the charging pile, including idle state, fault state, charging, charging completed, maintenance state, etc.

[0087] The control module is used to execute the occupancy judgment algorithm: matching the vehicle identification result with the charging pile status information, if the matching result shows that the vehicle is a non-charging vehicle and the charging pile is in an idle state, it is judged as an occupied vehicle. The judgment logic is implemented by the software algorithm in the control module.

[0088] The following table shows the corresponding relationship: The display module is used to display the occupancy judgment result: a warning message is sent to the owner of the occupied vehicle through the display module, and the display module includes a display screen on the charging pile and a mobile application.

[0089] The billing module is used to start the occupancy fee billing system when the occupancy behavior is determined: when the occupancy time exceeds the preset time, the occupancy fee billing system is started through the billing module, and the billing module calculates the occupancy fee based on the preset rate and occupancy time by multiplying the two.

[0090] Example 4 The embodiment of the present disclosure provides a charging pile occupancy judgment device 300, including a processor (processor) 304 and a memory (memory) 301. Optionally, the device may also include a communication interface (Communication Interface) 302 and a bus 303. Among them, the processor 304, the communication interface 302, and the memory 301 can communicate with each other through the bus 303. The communication interface 302 can be used for information transmission. The processor 304 can call the logic instructions in the memory 301 to execute the charging pile occupancy judgment method of the above embodiment.

[0091] In addition, the logic instructions in the memory 301 described above can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product.

[0092] The memory 301 is a computer-readable storage medium that can be used to store software programs and computer executable programs, such as program instructions / modules corresponding to the method in the embodiment of the present disclosure. The processor 304 executes the function application and data processing by running the program instructions / modules stored in the memory 301, that is, the charging pile occupancy judgment method in the above embodiment is implemented.

[0093] The memory 301 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and an application required for at least one function; the data storage area may store data created according to the use of the terminal device, etc. In addition, the memory 301 may include a high-speed random access memory and may also include a non-volatile memory. Example 5 An embodiment of the present disclosure provides a computer-readable storage medium storing computer-executable instructions, wherein the computer-executable instructions are configured to execute the above-mentioned charging pile occupancy determination method.

[0094] The computer-readable storage medium mentioned above may be a transient computer-readable storage medium or a non-transitory computer-readable storage medium.

[0095] The technical solution of the embodiment of the present disclosure can be embodied in the form of a software product, which is stored in a storage medium and includes one or more instructions for enabling a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in the embodiment of the present disclosure. The aforementioned storage medium may be a non-transient storage medium, including: a USB flash drive, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a disk or an optical disk, and other media that can store program codes, or a transient storage medium.

[0096] The above description and the accompanying drawings fully illustrate the embodiments of the present disclosure so that those skilled in the art can practice them. Other embodiments may include structural, logical, electrical, process and other changes. The embodiments represent only possible changes. Unless explicitly required, separate components and functions are optional, and the order of operations may vary. Parts and features of some embodiments may be included in or replace parts and features of other embodiments. Moreover, the words used in this application are only used to describe the embodiments and are not used to limit the claims. As used in the description of the embodiments and the claims, unless the context clearly indicates, the singular forms of "a", "an" and "the" are intended to include plural forms as well. Similarly, the term "and / or" as used in this application refers to any and all possible combinations of one or more associated listings. In addition, when used in this application, the term "comprise" and its variants "comprises" and / or comprising refer to the presence of stated features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or groups thereof. In the absence of further restrictions, the elements defined by the sentence "including one..." do not exclude the presence of other identical elements in the process, method or device including the elements. In this article, each embodiment may focus on the differences from other embodiments, and the same and similar parts between the embodiments may refer to each other. For the methods, products, etc. disclosed in the embodiments, if they correspond to the method part disclosed in the embodiments, then the relevant parts can refer to the description of the method part.

[0097] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software may depend on the specific application and design constraints of the technical solution. The technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the embodiments of the present disclosure. The technicians may clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here.

Claims

1. A method for determining the occupancy of a charging pile, characterized in that: The following steps are involved: S01, the camera collects the image of the vehicle driving into the parking space where the charging pile is located, and transmits the vehicle image to the vehicle identification module; S02, the vehicle identification module analyzes the received vehicle image in real time, and extracts the license plate number and vehicle features, and obtains the vehicle identification result based on the license plate number and vehicle features. The vehicle identification result is the type of vehicle driving into the parking space where the charging pile is located, and the vehicle type includes new energy vehicles and non-new energy vehicles; S03, uploading the status information of the charging pile to the control module, the status information including idle state, fault state, charging, charging completed, and maintenance state; S04. The control module makes an occupancy judgment based on the vehicle identification result obtained by the vehicle identification module and the charging pile status information, and matches the vehicle identification result with the charging pile status information. If the vehicle type is not a new energy vehicle, the charging status information is not considered and it is directly judged as a charging pile occupancy. If the vehicle type is a new energy vehicle, the charging pile status is idle, faulty, under maintenance, or exceeds the preset charging completion time, it is judged as a charging pile occupancy. If the charging pile status is charging or within the preset charging completion time, it is judged as a charging pile not occupancy.

2. The charging pile occupancy determination method according to claim 1, characterized in that: The process of analyzing vehicle images by the vehicle recognition module is as follows: S211, license plate positioning, using a deep learning model to locate the license plate. If the license plate is located, the deep learning model outputs the confidence and bounding box coordinates of the license plate area, where the license plate positioning function is: ， in fdetect is a function used to detect the location of the license plate in the vehicle image. is the vehicle image, is the bounding box combination, are the parameters of the deep learning model, including weights, biases, and optimizer parameters; S212, character segmentation, using a projection-based method to perform character segmentation on the located license plate area, the character segmentation function is: ， in fsegment Is a function used to perform character segmentation on the located license plate area. is the set of characters after segmentation; S213, character recognition, use the trained convolutional neural network to recognize each segmented character, the character recognition function is: , in frecognize is a function used to identify the character corresponding to each character image. is the set of recognized characters, are the parameters of the convolutional neural network, including weights, biases, pooling layer parameters, and normalization parameters; S214, judging the new energy license plate, judging whether the recognized characters are new energy license plates according to the format and character features of the license plate number, and the new energy license plate judging formula is: ， in fjudge It is a function used to determine whether the license plate is a new energy license plate based on the format and character characteristics of the license plate number. is the license plate type tag, Equal to 1 means a new energy license plate, Equal to 0 means it is not a new energy vehicle license plate.

3. The charging pile occupancy determination method according to claim 1, characterized in that: The process of the license plate recognition module analyzing the vehicle image is as follows: S221, license plate location, use the deep learning model to locate the license plate. If the license plate is not located, collect other features of the vehicle, including vehicle model, logo, window shape, headlight design, body color, and air intake grille. The vehicle feature extraction formula is: ， in fextract is a function used to extract key features from vehicle images. These features help identify and distinguish different vehicles. I is the vehicle image and F is the set of extracted vehicle features. S222, feature matching, matching the extracted vehicle features with the feature templates in the database to determine whether the vehicle is a new energy vehicle. The feature matching formula is: ， in fmatch is a function used to match the extracted vehicle features with the feature templates in the database and calculate the matching score. M is the matching score and T is the feature template set in the database. S223. Determine whether the vehicle is a new energy vehicle based on the matching score. The new energy vehicle model determination formula is: ， in fclassify It is a function used to determine whether a vehicle is a new energy vehicle based on the matching score and other vehicle features. V is the vehicle type label. A value of 1 indicates a new energy vehicle license plate, and a value of 0 indicates a non-new energy vehicle license plate.

4. The charging pile occupancy determination method according to claim 1, characterized in that: Multiple cameras with different acquisition angles are deployed around the charging pile, and the images collected by multiple cameras are fused and transmitted to the vehicle recognition module.

5. The charging pile occupancy determination method according to claim 1, characterized in that: During the training process of the vehicle recognition module, an adaptive learning rate algorithm is used to accelerate the convergence of the model. The formula of the adaptive learning rate algorithm is: ， in , are the model parameters at time steps t and t+1, respectively. is the adaptive learning rate, is the loss function.

6. The charging pile occupancy determination method according to claim 1, characterized in that: Distance sensors are also deployed around the charging piles. When the distance sensor senses that a vehicle is approaching a preset area and the distance between the vehicle and the distance sensor is less than a preset value, the camera is started to capture images.

7. The charging pile occupancy determination method according to claim 1, characterized in that: It also includes step S05 result display, sending a warning message to the owner of the occupied vehicle through the display module, and the display module includes a display screen on the charging pile and a mobile application.

8. The charging pile occupancy determination method according to claim 1, characterized in that: It also includes step S06 of occupancy billing. When the occupancy behavior is determined, the occupancy fee billing system is started. When the occupancy time exceeds the preset time, the occupancy fee billing system is started through the billing module. The billing module calculates the occupancy fee based on the preset rate and occupancy time by multiplying the two.

9. A charging pile occupancy judgment device, comprising a processor and a memory storing program instructions, characterized in that: The processor is configured to execute the charging pile occupancy determination method according to any one of claims 1 to 8 when running the program instructions.

10. A storage medium storing program instructions, characterized in that: When the program instructions are executed, the charging pile occupancy determination method according to any one of claims 1 to 8 is executed.

Images