A cloud computing-based charging pile management method and device

Abstract

This invention provides a cloud computing-based charging pile management method and apparatus. The method includes: acquiring image data uploaded by various edge devices through a cloud service platform; detecting and classifying vehicles in the image data to obtain a predicted vehicle capacity set after a preset time; and then sending this set to the charging piles for broadcast. The beneficial effects of this invention are: it enables the management of charging piles, making fuller use of charging pile resources; it also improves the user's charging experience and saves on the construction of charging facilities while meeting user charging needs.

Description

Technical Field

[0001] This invention relates to the field of new energy, and in particular to a charging pile management method and device based on cloud computing. Background Technology

[0002] With increasing environmental awareness and continuous technological advancements, the new energy electric vehicle market is developing rapidly. In recent years, the continuous development of new battery technologies such as lithium-ion batteries has led to a continuous increase in the driving range of electric vehicles and a reduction in charging time, further promoting the popularization of electric vehicles. Currently, charging infrastructure is a key focus; however, too many charging stations will increase infrastructure costs, while too few charging stations will easily cause charging difficulties for users. Existing technology lacks management of charging stations, resulting in insufficient utilization of charging station resources. Therefore, how to manage charging stations has become an important task. Summary of the Invention

[0003] The main objective of this invention is to provide a cloud computing-based charging pile management method and device, which aims to solve the problem of insufficient resource utilization of charging piles due to the lack of management of charging piles.

[0004] This invention provides a cloud computing-based charging pile management method. The method is implemented through a cloud service device, which includes multiple edge devices and a cloud service platform. The edge devices are deployed on multiple charging piles within a designated area, with each edge device corresponding to a specific charging pile. The edge devices connect to the charging piles and establish a communication connection with the cloud service platform to acquire image data transmitted by the charging piles. The charging pile management method includes:

[0005] The cloud service platform acquires image data uploaded by various edge devices;

[0006] Each image data is processed by frame extraction at a preset time interval to obtain a frame extraction dataset corresponding to each image data, and the frame extraction dataset is divided into multiple frame extraction data subsets in chronological order.

[0007] The vehicle detection results in each frame of the sampled data subset are obtained by detecting each frame of the sampled data subset according to the preset target detection algorithm.

[0008] The vehicles are classified according to the vehicle detection results; the vehicles in the subset of extracted data are tracked and labeled.

[0009] The threshold T for the same vehicle in the target area of ​​the vehicle detection result is determined by a preset algorithm;

[0010] If the target area threshold T is greater than threshold a, the vehicle is considered stationary; if the target area threshold T is less than threshold b, the vehicle is considered moving; if the target area threshold T is less than threshold a and greater than threshold b, the vehicle is considered locally moving.

[0011] Count the number of vehicles of each type in each sampled frame data subset to obtain the vehicle capacity set corresponding to each sampled frame dataset.

[0012] Predict the predicted vehicle capacity set after a preset time based on the vehicle capacity set;

[0013] The predicted vehicle capacity set is sent to the charging piles and broadcast at each charging pile, thereby enabling the management of the charging piles.

[0014] Furthermore, before the step of predicting the vehicle capacity set based on the vehicle capacity set after a preset time, the method further includes:

[0015] Obtain vehicle capacity sets from multiple subsets of sampled data. ;in, Indicates time as The vehicle capacity set of the i-th subset of extracted frame data, and , This refers to the interval time.

[0016] Input each vehicle capacity set into the fitting formula In the middle, the conversion parameters are obtained. ;in, The fitting parameters represent the vehicle capacity set corresponding to the i-th road segment. .

[0017] Furthermore, the step of determining the threshold T of the same vehicle in the target area of ​​the vehicle detection result using a preset algorithm further includes:

[0018] The images of the same vehicle in the (n+1)th frame, the nth frame, and the (n-1)th frame are denoted as Ln+1, Ln, and Ln-1, respectively, and the gray values ​​of the corresponding pixel points (x,y) in the three frames are denoted as Ln+1(x,y), Ln(x,y), and Ln-1(x,y).

[0019] The first difference image is obtained by subtracting the gray values ​​Ln(x,y) and Ln-1(x,y) and taking their absolute values. The second difference image is obtained by subtracting the gray values ​​Ln+1(x,y) and Ln(x,y) and taking their absolute values.

[0020] The first and second difference images are processed using the formula The target image is obtained through calculation; where, Represents the target image. This represents the first difference image. This represents the second difference image;

[0021] The target image is subjected to threshold decision processing for the target region to obtain the threshold T of the vehicle.

[0022] Further, the step of detecting each subset of sampled frame data according to a preset target detection algorithm to obtain the vehicle detection result in each frame of the sampled frame data subset includes:

[0023] Candidate regions are extracted from each frame in the subset of framed data using the Selective Search algorithm.

[0024] After determining the candidate regions, feature extraction is performed on the candidate regions, and feature classification is performed on the selected features of the candidate regions based on the SVM classifier;

[0025] The vehicle detection results are obtained by confirming candidate regions belonging to the same category using the regressor.

[0026] Furthermore, after the step of sending the predicted vehicle capacity set to the charging piles and broadcasting it at each charging pile to achieve management of the charging piles, the method further includes:

[0027] Receive the estimated charging time uploaded by the user;

[0028] After calculating the expected charging time based on the vehicle capacity set, the target vehicle capacity set of each charging pile is calculated.

[0029] Recommendation of target charging stations for the user based on each target vehicle capacity set.

[0030] This invention also provides a cloud-based charging pile management device. The charging pile management method is implemented through a cloud service device, wherein the cloud service device includes multiple edge devices and a cloud service platform. The multiple edge devices are deployed on multiple charging piles in a designated area, and each edge device corresponds one-to-one with a charging pile. The edge devices are connected to the charging piles and establish a communication connection with the cloud service platform to acquire image data transmitted by the charging piles. The charging pile management device includes:

[0031] The acquisition module is used to acquire image data uploaded by various edge devices;

[0032] The frame extraction module is used to perform frame extraction processing on each of the image data according to a preset time interval to obtain the frame extraction dataset corresponding to each image data, and divide the frame extraction dataset into multiple frame extraction data subsets according to time order.

[0033] The detection module is used to detect each subset of frame data according to a preset target detection algorithm, and obtain the vehicle detection result in each frame of the subset of frame data.

[0034] The classification module is used to classify vehicles based on the vehicle detection results; it tracks vehicles in the subset of sampled data and labels the vehicles.

[0035] The determination module is used to determine the threshold T of the same vehicle in the target area of ​​the vehicle detection result through a preset algorithm;

[0036] The identification module is used to identify a vehicle as stationary if the target area threshold T is greater than threshold a; to identify a vehicle as moving if the target area threshold T is less than threshold b; and to identify a vehicle as partially moving if the target area threshold T is less than threshold a and greater than threshold b.

[0037] The statistics module is used to count the number of vehicles of various types in each sampled frame data subset, and to obtain the vehicle capacity set corresponding to each sampled frame dataset.

[0038] The prediction module is used to predict the vehicle capacity set after a preset time based on the vehicle capacity set.

[0039] The broadcast module is used to send the predicted vehicle capacity set to the charging piles and broadcast it at each charging pile, thereby realizing the management of the charging piles.

[0040] Furthermore, the charging pile management device also includes:

[0041] The vehicle capacity set acquisition module is used to acquire vehicle capacity sets from multiple subsets of sampled data. ;in, Indicates time as The vehicle capacity set of the i-th subset of extracted frame data, and , This refers to the interval time.

[0042] The vehicle capacity set input module is used to input the fitting formula for each vehicle capacity set. In the middle, the conversion parameters are obtained. ;in, The fitting parameters represent the vehicle capacity set corresponding to the i-th road segment. .

[0043] Furthermore, the determining module also includes:

[0044] This is designated as a submodule, used to denote the images of the (n+1)th, nth, and (n-1)th frames of the same vehicle as Ln+1, Ln, and Ln-1 respectively, and the gray values ​​of the corresponding pixel points (x,y) in the three frames as Ln+1(x,y), Ln(x,y), and Ln-1(x,y).

[0045] The first calculation submodule is used to subtract the gray values ​​Ln(x,y) and Ln-1(x,y) and take their absolute values ​​to obtain the first difference image, and to subtract the gray values ​​Ln+1(x,y) and Ln(x,y) and take their absolute values ​​to obtain the difference image.

[0046] The second calculation submodule is used to process the first difference image and the second difference image using a formula. The target image is obtained through calculation; where, Represents the target image. This represents the first difference image. This represents the second difference image;

[0047] The decision submodule performs threshold decision processing on the target image to obtain the threshold T of the vehicle.

[0048] Furthermore, the detection module includes:

[0049] The extraction submodule is used to extract candidate regions from each frame in the subset of extracted frame data according to the Selective Search algorithm;

[0050] The feature extraction submodule is used to extract features from candidate regions after determining them, and to classify the features selected from the candidate regions according to the SVM classifier.

[0051] The confirmation submodule is used to confirm candidate regions belonging to the same category based on the regressor, and obtain vehicle detection results.

[0052] Furthermore, the charging pile management device also includes:

[0053] The charging time receiving module is used to receive the estimated charging time uploaded by the user;

[0054] The target vehicle capacity set calculation module is used to calculate the target vehicle capacity set of each charging pile after the expected charging time based on the vehicle capacity set.

[0055] The target charging station selection module is used to recommend target charging stations to the user based on the capacity sets of each target vehicle.

[0056] The beneficial effects of this invention are as follows: By acquiring image data uploaded by various edge devices through a cloud service platform, and by detecting and classifying vehicles in the image data, a predicted vehicle capacity set after a preset time is obtained, which is then sent to the charging pile for broadcast, thereby realizing the management of the charging pile and making fuller use of the charging pile resources. In addition, it also improves the user's charging experience and can save on the construction of charging facilities while meeting the user's charging needs. Attached Figure Description

[0057] Figure 1 This is a flowchart illustrating a cloud-based charging pile management method according to an embodiment of the present invention.

[0058] Figure 2 This is a schematic block diagram of a cloud computing-based charging pile management device according to an embodiment of the present invention.

[0059] Figure 3 This is a schematic block diagram of the structure of a computer device according to an embodiment of this application.

[0060] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0061] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0062] It should be noted that all directional indicators (such as up, down, left, right, front, back, etc.) in the embodiments of the present invention are only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicator will also change accordingly. The connection can be a direct connection or an indirect connection.

[0063] In this article, the term "and / or" is merely a description of the relationship between related objects, indicating that there can be three kinds of relationships. For example, A and B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone.

[0064] Furthermore, in this invention, descriptions involving "first," "second," etc., are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of that feature. Additionally, the technical solutions of the various embodiments can be combined with each other, but only on the basis of being achievable by those skilled in the art. When the combination of technical solutions is contradictory or impossible to implement, such a combination of technical solutions should be considered non-existent and not within the scope of protection claimed by this invention.

[0065] Reference Figure 1 This invention proposes a cloud computing-based charging pile management method. The method is implemented through a cloud service device, which includes multiple edge devices and a cloud service platform. The edge devices are deployed on multiple charging piles within a designated area, with each edge device corresponding to a specific charging pile. The edge devices connect to the charging piles and establish a communication connection with the cloud service platform to acquire image data transmitted by the charging piles. The charging pile management method includes:

[0066] S1: The cloud service platform acquires image data uploaded by each edge device;

[0067] S2: Perform frame extraction processing on each of the image data according to a preset time interval to obtain the frame extraction dataset corresponding to each image data, and divide the frame extraction dataset into multiple frame extraction data subsets according to time order.

[0068] S3: Detect each subset of frame data according to the preset target detection algorithm to obtain the vehicle detection result in each frame of the subset of frame data;

[0069] S4: Classify the vehicles based on the vehicle detection results; track the vehicles in the subset of extracted data and label them;

[0070] S5: Determine the threshold T for the same vehicle in the target area of ​​the vehicle detection result through a preset algorithm;

[0071] S6: If the target area threshold T is greater than the threshold a, the vehicle is considered to be stationary; if the target area threshold T is less than the threshold b, the vehicle is considered to be moving; if the target area threshold T is less than the threshold a and greater than the threshold b, the vehicle is considered to be locally moving.

[0072] S7: Count the number of vehicles of each type in each sampled frame data subset to obtain the vehicle capacity set corresponding to each sampled frame dataset.

[0073] S8: Predict the vehicle capacity set after a preset time based on the vehicle capacity set;

[0074] S9: The predicted vehicle capacity set is sent to the charging pile and broadcast at each charging pile to achieve management of the charging pile.

[0075] As described in step S1 above, the cloud service platform acquires image data uploaded by various edge devices. Acquiring image data from edge devices means using cloud computing technology to centrally store and process the image data transmitted from various edge devices in the cloud. This method allows edge devices to focus more on data acquisition, while the complex processing tasks are handled by the powerful computing resources in the cloud. Edge devices can be cameras.

[0076] As described in step S2 above, each image data is processed by frame extraction at preset time intervals to obtain a frame-extracted dataset corresponding to each image data. This dataset is then divided into multiple subsets of frame-extracted data in chronological order. Specific frames are extracted (or "extracted") from each image data at predetermined time intervals. This process can be used to reduce the amount of data, retaining only key or representative image frames. All frames extracted from the image data are integrated into a single "frame-extracted dataset." This dataset is further divided into multiple "frame-extracted data subsets" in chronological order. This allows for more ordered or systematic processing and analysis.

[0077] As described in step S3 above, each subset of sampled data is detected according to a preset target detection algorithm to obtain the vehicle detection results in each frame of the sampled data subset. The specific detection methods and target detection algorithms will be explained in detail later and will not be repeated here. Target detection algorithms, such as YOLO, Faster R-CNN, or SSD, are applied to the images in each subset of sampled data. This helps to find and identify all vehicles from the image.

[0078] As described in step S4 above, vehicle detection results are classified based on the vehicles to be detected; vehicles in the subset of sampled frames are tracked and labeled. Vehicle detection results need to be classified according to different criteria, namely, stationary vehicles, moving vehicles, and partially moving vehicles. Vehicle tracking is a continuous process; by tracking vehicles between consecutive frames, their motion trajectories can be obtained. Relevant tracking algorithms (such as Kalman filtering, multi-object tracking, etc.) can be used to achieve this. Once vehicles are tracked, each vehicle can be labeled, such as with a unique ID or other information, to identify and distinguish different vehicles in subsequent analysis.

[0079] As described in step S5 above, the threshold T of the same vehicle in the target area of ​​the vehicle detection result is determined by a preset algorithm. The steps for calculating the threshold T will be explained in detail later and will not be repeated here.

[0080] As described in step S6 above, if the target area threshold T is greater than threshold a, the vehicle is considered a stationary vehicle; if the target area threshold T is less than threshold b, the vehicle is considered a moving vehicle; if the target area threshold T is less than threshold a and greater than threshold b, the vehicle is considered a partially moving vehicle, thus classifying the vehicles. It should be noted that a moving vehicle may be a passing vehicle or a vehicle that has finished charging, while a partially moving vehicle is a vehicle that is slowly moving to find a charging location, which can be determined based on its speed. A stationary vehicle can be considered a vehicle that is charging.

[0081] As described in step S7 above, the number of vehicles of each type in each subset of frame data is counted to obtain the vehicle capacity set corresponding to each subset of frame data. Since the vehicles have been divided into stationary vehicles, moving vehicles, and partially moving vehicles as mentioned above, the vehicle capacity set corresponding to each subset of frame data can be obtained directly here.

[0082] As described in step S8 above, a predicted vehicle capacity set is generated based on the vehicle capacity set after a preset time. The prediction method can employ time series analysis methods, such as ARIMA (Autoregressive Integral Moving Average) or Prophet, to predict vehicle capacity at future points in time. Alternatively, machine learning methods, such as regression models or neural networks, can be used to build a predictive model based on historical data. This allows for the prediction of the vehicle capacity set.

[0083] As described in step S9 above, the predicted vehicle capacity set is sent to the charging piles and broadcast at each charging pile, thereby enabling the management of the charging piles. Once the charging piles receive the predicted vehicle capacity set, they can use broadcasting or other communication methods to disseminate this information to surrounding users or systems. This allows users to understand the charging pile's availability in advance and make better decisions. On the user side, a user interface can be designed to allow users to easily view the predicted vehicle capacity information and make charging plans based on this information. This achieves charging pile management, making fuller use of charging pile resources, improving the user's charging experience, and saving on the construction costs of charging infrastructure while meeting user charging needs.

[0084] In one embodiment, before step S8 of predicting the vehicle capacity set after a preset time based on the vehicle capacity set, the method further includes:

[0085] S701: Obtain vehicle capacity sets from multiple subsets of sampled frame data. ;in, Indicates time as The vehicle capacity set of the i-th subset of extracted frame data, and , This refers to the interval time.

[0086] S702: Input the various vehicle capacity sets into the fitting formula In the middle, the conversion parameters are obtained. ;in, The fitting parameters represent the vehicle capacity set corresponding to the i-th road segment. .

[0087] As described in steps S701-S702 above, Indicates time as The vehicle capacity set of the i-th subset of extracted frame data, and , The interval time is used; each element of each of the vehicle capacity sets is input into the fitting formula. Specifically, the charging and occupancy of vehicles exhibit periodic patterns. The charging time of each vehicle and the time when each user picks up their vehicle also follow certain regularities. Therefore, by statistically analyzing the status of vehicles within the range of each charging station at each time point, a predicted vehicle capacity set for a certain period of time can be obtained based on the current status, which facilitates the management of charging stations.

[0088] In one embodiment, step S5, which determines the threshold T of the target area of ​​the vehicle detection result using a preset algorithm, further includes:

[0089] S501: The images of the (n+1)th frame, the nth frame, and the (n-1)th frame of the same vehicle are denoted as Ln+1, Ln, and Ln-1, respectively, and the gray values ​​of the corresponding pixel points (x,y) in the three frames are denoted as Ln+1(x,y), Ln(x,y), and Ln-1(x,y).

[0090] S502: Subtract the gray values ​​Ln(x,y) and Ln-1(x,y) and take their absolute values ​​to obtain the first difference image; subtract the gray values ​​Ln+1(x,y) and Ln(x,y) and take their absolute values ​​to obtain the difference image.

[0091] S503: The first difference image and the second difference image are processed... The target image is obtained through calculation; where, Represents the target image. This represents the first difference image. This represents the second difference image;

[0092] S504: Perform threshold decision processing on the target image to obtain the threshold T of the vehicle.

[0093] As described in steps S501-S504 above, the threshold T for vehicles is obtained, that is, the threshold T of the target region is judged for each vehicle in the difference image to further improve the accuracy and stability of vehicle detection; the difference images of two consecutive frames and the background difference image in the dynamic image are processed, and then binarized to obtain the motion result; the target region is a value of a region of interest, so the purpose of the target region threshold T decision method is to determine which categories; in actual scenes, there is interference caused by changes in light; therefore, in the background difference method, the fixed background cannot be static; in order to ensure that the background image can change with the light and ensure the accuracy of detection, the background difference method and the inter-frame difference method are used as inspiration to process the difference images of two consecutive frames and the background difference image, and then the result is binarized to obtain the motion result. The motion results demonstrate how this increases the weight of the target vehicle while suppressing the static background, resulting in a motion detection image containing more vehicle information. Adding an additive term sensitive to overall illumination can further improve the accuracy of the decision, as pixel value changes under different lighting conditions affect the threshold selection. This sensitivity allows for adjustments based on the overall illumination of the image. Secondly, a difference image, obtained by subtracting two original images, is commonly used in target detection and tracking. However, due to signal issues, difference images often contain noise that can interfere with the accurate extraction of target information. Therefore, when using difference images for target detection and other tasks, noise suppression can be performed to obtain more accurate information. Finally, thresholding the target region T for different target types can further improve the accuracy of target detection. Different vehicle types may appear differently in images, so different thresholds T are needed for judgment. By judging the vehicle region thresholds T for different types of vehicles, information about different types of vehicles can be extracted more accurately. If the threshold T is greater than threshold a, the target is considered a stationary vehicle; if the threshold T is less than threshold b, the target is considered a moving vehicle, thereby reducing the output of false targets. The method of reducing the false detection rate is achieved by filtering the target detection results with spatiotemporal information.

[0094] In one embodiment, step S3, which involves detecting each subset of frame data according to a preset target detection algorithm to obtain the vehicle detection result in each frame of the subset of frame data, includes:

[0095] S301: Extract candidate regions for each frame in the subset of framed data according to the Selective Search algorithm;

[0096] S302: After determining the candidate region, extract features from the candidate region and classify the selected features of the candidate region according to the SVM classifier;

[0097] S303: Based on the regressor, candidate regions belonging to the same category are confirmed to obtain vehicle detection results.

[0098] As described in steps S301-S303 above, specifically, the Selective Search algorithm is used to extract multiple candidate regions from bottom to top in the image. Because the extracted regions are of different sizes, it is necessary to scale the image patches within each candidate region to the same size. Then, the entire image is input into the CNN. Multi-scale input to CNNs is achieved through SPP Net combined with the spatial pyramid method. A target region pooling layer is added to the CNN structure, allowing the input image to be of arbitrary size. SPP Net connects a pyramid pooling layer after the last convolutional layer to ensure that the input to the next fully connected layer is fixed. Specifically, in a CNN structure, the size of the input image is often fixed, and the output is a fixed-dimensional vector. SPP Net adds a target region pooling layer to the ordinary CNN structure, allowing the network's input image to be of arbitrary size, while the output remains a fixed-dimensional vector. In the SPP layer, each pooling filter adjusts its size according to the input, and the output of SPP is a fixed-dimensional vector, which is then fed to the fully connected (FC) layer. The Net performs a convolution calculation on the original image to obtain a convolutional feature map of the entire image. Then, each candidate region is mapped onto a patch on the feature map, and this patch is used as the convolutional feature of each candidate region. This patch is then input into the SPP layer and subsequent layers. The features extracted from the candidate regions are used by a classifier to determine whether they belong to a specific class. The regressor confirms the candidate regions that belong to the same class, thus obtaining the vehicle detection result.

[0099] In one embodiment, after step S9, which involves sending the predicted vehicle capacity set to the charging piles and broadcasting it at each charging pile to achieve management of the charging piles, the method further includes:

[0100] S1001: Receives the estimated charging time uploaded by the user;

[0101] S1002: Calculate the target vehicle capacity set for each charging pile after calculating the expected charging time based on the vehicle capacity set;

[0102] S1003: Recommend target charging stations for the user based on each of the target vehicle capacity sets.

[0103] As described in step S1001 above, receive the user's uploaded expected charging time and design a user interface that allows users to input their expected charging time. This can be an application, a web form, or other interactive interface. Provide a convenient time input method on the interface, allowing users to easily select or enter the specific time they wish to charge. You can use a time picker or text input box, etc. If the user inputs the expected charging time, ensure that this data is transmitted to the backend system for processing.

[0104] As described in step S1002 above, after calculating the expected charging time based on the vehicle capacity set, the target vehicle capacity set of each charging pile is calculated in the same way as the aforementioned method for calculating the predicted vehicle capacity set, and will not be repeated here.

[0105] As described in step S1003 above, target charging piles are selected and recommended for the user based on each target vehicle capacity set. That is, target charging piles are selected and recommended for the user based on each target vehicle capacity set. The specific method of recommendation is not limited. Since the target vehicle capacity sets of each charging pile are known, the optimal charging pile can be selected for recommendation, thereby improving the user experience.

[0106] Reference Figure 2 The present invention also provides a cloud computing-based charging pile management device. The charging pile management method is implemented through a cloud service device, wherein the cloud service device includes multiple edge devices and a cloud service platform. The multiple edge devices are deployed on multiple charging piles in a designated area, and each edge device corresponds one-to-one with a charging pile. The edge devices are connected to the charging piles and establish a communication connection with the cloud service platform to acquire image data transmitted by the charging piles. The charging pile management device includes:

[0107] The acquisition module 10 is used to acquire image data uploaded by each edge device;

[0108] The frame extraction module 20 is used to perform frame extraction processing on each of the image data according to a preset time interval to obtain the frame extraction dataset corresponding to each image data, and divide the frame extraction dataset into multiple frame extraction data subsets according to time order.

[0109] The detection module 30 is used to detect each subset of frame data according to a preset target detection algorithm to obtain the vehicle detection result in each frame of the subset of frame data.

[0110] The classification module 40 is used to classify vehicles based on the vehicle detection results of the vehicles to be detected; to track vehicles in the subset of frame data and to label the vehicles.

[0111] The determination module 50 is used to determine the threshold T of the same vehicle in the target area of ​​the vehicle detection result through a preset algorithm;

[0112] The identification module 60 is used to identify the vehicle as stationary if the target area threshold T is greater than the threshold a; to identify the vehicle as moving if the target area threshold T is less than the threshold b; and to identify the vehicle as partially moving if the target area threshold T is less than the threshold a and greater than the threshold b.

[0113] The statistics module 70 is used to count the number of vehicles of various types in each sampled frame data subset, and to obtain the vehicle capacity set corresponding to each sampled frame dataset.

[0114] Prediction module 80 is used to predict the vehicle capacity set after a preset time based on the vehicle capacity set.

[0115] The broadcast module 90 is used to send the predicted vehicle capacity set to the charging pile and broadcast it at each charging pile, thereby realizing the management of the charging pile.

[0116] In one embodiment, the charging pile management device further includes:

[0117] The vehicle capacity set acquisition module is used to acquire vehicles from multiple subsets of sampled data. In the middle, the conversion parameters are obtained. ;in, The fitting parameters represent the vehicle capacity set corresponding to the i-th road segment. ;in, Indicates time as The vehicle capacity set of the i-th subset of extracted frame data, and , This refers to the interval time.

[0118] The vehicle capacity set input module is used to input the fitting formula for each vehicle capacity set. In the middle, the conversion parameters are obtained. ;in, The fitting parameters represent the vehicle capacity set corresponding to the i-th road segment. .

[0119] In one embodiment, the determining module 50 further includes:

[0120] This is designated as a submodule, used to denote the images of the (n+1)th, nth, and (n-1)th frames of the same vehicle as Ln+1, Ln, and Ln-1 respectively, and the gray values ​​of the corresponding pixel points (x,y) in the three frames as Ln+1(x,y), Ln(x,y), and Ln-1(x,y).

[0121] The first calculation submodule is used to subtract the gray values ​​Ln(x,y) and Ln-1(x,y) and take their absolute values ​​to obtain the first difference image, and to subtract the gray values ​​Ln+1(x,y) and Ln(x,y) and take their absolute values ​​to obtain the difference image.

[0122] The second calculation submodule is used to process the first difference image and the second difference image using a formula. The target image is obtained through calculation; where, Represents the target image. This represents the first difference image. This represents the second difference image;

[0123] The decision submodule is used to perform threshold decision processing on the target image to obtain the threshold T of the vehicle. In one embodiment, the detection module 30 includes:

[0124] The extraction submodule is used to extract candidate regions from each frame in the subset of extracted frame data according to the Selective Search algorithm;

[0125] The feature extraction submodule is used to extract features from candidate regions after determining them, and to classify the features selected from the candidate regions according to the SVM classifier.

[0126] The confirmation submodule is used to confirm candidate regions belonging to the same category based on the regressor, and obtain vehicle detection results.

[0127] In one embodiment, the charging pile management device further includes:

[0128] The charging time receiving module is used to receive the estimated charging time uploaded by the user;

[0129] The target vehicle capacity set calculation module is used to calculate the target vehicle capacity set of each charging pile after the expected charging time based on the vehicle capacity set.

[0130] The target charging station selection module is used to recommend target charging stations to the user based on the capacity sets of each target vehicle.

[0131] The beneficial effects of this invention are as follows: By acquiring image data uploaded by various edge devices through a cloud service platform, and by detecting and classifying vehicles in the image data, a predicted vehicle capacity set after a preset time is obtained, which is then sent to the charging pile for broadcast, thereby realizing the management of the charging pile and making fuller use of the charging pile resources. In addition, it also improves the user's charging experience and can save on the construction of charging facilities while meeting the user's charging needs.

[0132] Reference Figure 3This application also provides a computer device, which may be a server, and its internal structure may be as follows: Figure 3 As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores various image data, etc. The network interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it can implement the cloud computing-based charging pile management method described in any of the above embodiments.

[0133] Those skilled in the art will understand that Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer equipment on which the present application is applied.

[0134] This application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, can implement the cloud computing-based charging pile management method described in any of the above embodiments.

[0135] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media provided in this application and used in the embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0136] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, apparatus, article, or method. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.

[0137] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence (AI) refers to the theories, methods, technologies, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0138] Foundational technologies for artificial intelligence generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies mainly encompass computer vision, robotics, biometrics, speech processing, natural language processing, and machine learning / deep learning.

[0139] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of the claims of the present invention.

Claims

1. A cloud computing-based charging pile management method, characterized in that, The charging pile management method is implemented through a cloud service device, wherein the cloud service device includes multiple edge devices and a cloud service platform. The edge devices are deployed on multiple charging piles in a designated area, and each edge device corresponds one-to-one with a charging pile. The edge devices connect to the charging piles and establish a communication connection with the cloud service platform to acquire image data transmitted by the charging piles. The charging pile management method includes: The cloud service platform acquires image data uploaded by various edge devices; Each image data is processed by frame extraction at a preset time interval to obtain a frame extraction dataset corresponding to each image data, and the frame extraction dataset is divided into multiple frame extraction data subsets in chronological order. The vehicle detection results in each frame of the sampled data subset are obtained by detecting each frame of the sampled data subset according to the preset target detection algorithm. The vehicles are classified according to the vehicle detection results; the vehicles in the subset of extracted data are tracked and labeled. The threshold T for the same vehicle in the target area of ​​the vehicle detection result is determined by a preset algorithm; If the target area threshold T is greater than threshold a, the vehicle is considered stationary; if the target area threshold T is less than threshold b, the vehicle is considered moving; if the target area threshold T is less than threshold a and greater than threshold b, the vehicle is considered locally moving. The number of vehicles of each type is counted in each subset of the sampled data to obtain the vehicle capacity set corresponding to each sampled data set. Predict the predicted vehicle capacity set after a preset time based on the vehicle capacity set; The predicted vehicle capacity set is sent to the charging piles and broadcast at each charging pile, thereby enabling the management of the charging piles.

2. The cloud computing-based charging pile management method as described in claim 1, characterized in that, The step of determining the threshold T of the same vehicle in the target area of ​​the vehicle detection result using a preset algorithm further includes: The images of the same vehicle in the (n+1)th frame, the nth frame, and the (n-1)th frame are denoted as Ln+1, Ln, and Ln-1, respectively, and the gray values ​​of the corresponding pixel points (x,y) in the three frames are denoted as Ln+1(x,y), Ln(x,y), and Ln-1(x,y). The first difference image is obtained by subtracting the gray values ​​Ln(x,y) and Ln-1(x,y) and taking their absolute values. The second difference image is obtained by subtracting the gray values ​​Ln+1(x,y) and Ln(x,y) and taking their absolute values. The first and second difference images are processed using the formula The target image is obtained through calculation; where, Represents the target image. This represents the first difference image. This represents the second difference image; The target image is subjected to threshold decision processing for the target region to obtain the threshold T of the vehicle.

3. The charging pile management method based on cloud computing as described in claim 1, characterized in that, The step of detecting each subset of frame data according to a preset target detection algorithm to obtain the vehicle detection result in each frame of the subset of frame data includes: Candidate regions are extracted from each frame in the subset of sampled data using the Selective Search algorithm; After determining the candidate regions, feature extraction is performed on the candidate regions, and feature classification is performed on the selected features of the candidate regions based on the SVM classifier; The vehicle detection results are obtained by confirming candidate regions belonging to the same category using the regressor.

4. The charging pile management method based on cloud computing as described in claim 1, characterized in that, After the step of sending the predicted vehicle capacity set to the charging piles and broadcasting it at each charging pile to achieve the management of the charging piles, the method further includes: Receive the estimated charging time uploaded by the user; After calculating the expected charging time based on the vehicle capacity set, the target vehicle capacity set of each charging pile is calculated. Recommendation of target charging stations for the user based on each target vehicle capacity set.

5. A cloud computing-based charging pile management device, characterized in that, The charging pile management method is implemented through a cloud service device, wherein the cloud service device includes multiple edge devices and a cloud service platform. The edge devices are deployed on multiple charging piles in a designated area, and each edge device corresponds one-to-one with a charging pile. The edge devices connect to the charging piles and establish a communication connection with the cloud service platform to acquire image data transmitted by the charging piles. The charging pile management device includes: The acquisition module is used to acquire image data uploaded by various edge devices; The frame extraction module is used to perform frame extraction processing on each of the image data according to a preset time interval to obtain the frame extraction dataset corresponding to each image data, and divide the frame extraction dataset into multiple frame extraction data subsets according to time order. The detection module is used to detect each subset of frame data according to a preset target detection algorithm, and obtain the vehicle detection result in each frame of the subset of frame data. The classification module is used to classify vehicles based on the vehicle detection results; it tracks vehicles in the subset of sampled data and labels the vehicles. The determination module is used to determine the threshold T of the same vehicle in the target area of ​​the vehicle detection result through a preset algorithm; The identification module is used to identify a vehicle as stationary if the target area threshold T is greater than threshold a; to identify a vehicle as moving if the target area threshold T is less than threshold b; and to identify a vehicle as partially moving if the target area threshold T is less than threshold a and greater than threshold b. The statistics module is used to count the number of vehicles of various types in each sampled frame data subset, and to obtain the vehicle capacity set corresponding to each sampled frame dataset. The prediction module is used to predict the vehicle capacity set after a preset time based on the vehicle capacity set. The broadcast module is used to send the predicted vehicle capacity set to the charging piles and broadcast it at each charging pile, thereby realizing the management of the charging piles.

6. The cloud-based charging pile management device as described in claim 5, characterized in that, The determining module further includes: This is designated as a submodule, used to denote the images of the (n+1)th, nth, and (n-1)th frames of the same vehicle as Ln+1, Ln, and Ln-1 respectively, and the gray values ​​of the corresponding pixel points (x,y) in the three frames as Ln+1(x,y), Ln(x,y), and Ln-1(x,y). The first calculation submodule is used to subtract the gray values ​​Ln(x,y) and Ln-1(x,y) and take their absolute values ​​to obtain the first difference image, and to subtract the gray values ​​Ln+1(x,y) and Ln(x,y) and take their absolute values ​​to obtain the difference image. The second calculation submodule is used to process the first difference image and the second difference image using a formula. The target image is obtained through calculation; where, Represents the target image. This represents the first difference image. This represents the second difference image; The decision submodule is used to perform threshold decision processing on the target image to obtain the threshold T of the vehicle.

7. The cloud-based charging pile management device as described in claim 5, characterized in that, The detection module includes: The extraction submodule is used to extract candidate regions from each frame in the subset of extracted frame data according to the Selective Search algorithm; The feature extraction submodule is used to extract features from candidate regions after determining them, and to classify the features selected from the candidate regions according to the SVM classifier. The confirmation submodule is used to confirm candidate regions belonging to the same category based on the regressor, and obtain vehicle detection results.

8. The cloud-based charging pile management device as described in claim 5, characterized in that, The charging pile management device also includes: The charging time receiving module is used to receive the estimated charging time uploaded by the user; The target vehicle capacity set calculation module is used to calculate the target vehicle capacity set of each charging pile after the expected charging time based on the vehicle capacity set. The target charging station selection module is used to recommend target charging stations to the user based on the capacity sets of each target vehicle.

Images