New energy automobile charging management method and system

Abstract

The invention discloses a new energy vehicle charging management method and system, and relates to the technical field of new energy vehicle charging management, and the method comprises the following steps: collecting a vehicle historical charging record of a current user, and defining a common charging position based on the collected vehicle historical charging record, constructing a user charging behavior habit model through a neural network in combination with historical charging records of the vehicle; when a user vehicle is charged, outputting a reference charging position and a reference charging electric quantity through a user charging behavior habit model based on the current charging position of the user and the electric quantity of a battery, and generating a reference charging position charging strategy and a common charging position charging strategy in combination with the common charging position of the current user vehicle; and saving amounts under different charging strategies are calculated. According to the new energy automobile charging management method and system, the personalized charging strategy can be generated based on the charging behavior habits of the user, the cost advantage can be clearly presented, and meanwhile the battery maintenance requirement is considered.

Description

Technical Field

[0001] This invention relates to the field of new energy vehicle charging management technology, and more specifically, to a new energy vehicle charging management method and system. Background Technology

[0002] With the rapid development of the new energy vehicle industry, charging services, as a core part of the use of new energy vehicles, directly affect the user's driving experience in terms of convenience, economy and adaptability. At present, the user group of new energy vehicles is constantly expanding, and users' charging needs are showing personalized and diversified characteristics. They not only pay attention to the control of charging costs, but also value the matching of charging methods with their own usage habits and battery maintenance needs. Existing charging management technologies mostly focus on the basic functions of charging devices, lacking in-depth analysis of users' charging habits. Different users have significant differences in their daily travel routes, charging times, and preferred charging locations. However, traditional management methods do not build charging habit models for individual users, and cannot output benchmark charging locations and benchmark charging capacities that match user habits. The generated charging strategies are often general and difficult to adapt to users' personalized usage scenarios. Furthermore, current charging management solutions do not fully consider the diverse needs of users and lack functional adaptability. Some users need to choose fast charging mode to improve efficiency based on their schedule, and also want to choose slow charging mode to ensure battery maintenance when they have enough time. However, existing technologies mostly do not systematically integrate slow and fast charging strategies, and cannot generate personalized charging strategies, thus failing to meet users' actual needs in different scenarios. Summary of the Invention

[0003] In view of the problems in related technologies, the present invention proposes a charging management method and system for new energy vehicles to overcome the above-mentioned technical problems existing in the existing related technologies.

[0004] Therefore, the specific technical solution adopted by the present invention is as follows: A method for managing the charging of new energy vehicles, the method comprising the following steps: S1. Collect the current user's vehicle charging history records, define frequently used charging locations based on the collected vehicle charging history records, and build a user charging behavior habit model through a neural network in conjunction with the vehicle charging history records. S2. When the user's vehicle is charging, based on the user's current charging location and battery level, the user's charging behavior habit model outputs the baseline charging location and baseline charging level. Combined with the user's current frequently used charging locations, the baseline charging location charging strategy and the frequently used charging location charging strategy are generated, and the amount saved under different charging strategies is calculated.

[0005] In a preferred embodiment, S1 includes the following steps: S11. Collect the current user's vehicle historical charging records, including the charging location, current battery level, charging amount, charging rate, charging time, and charging duration for each charging record. Based on the charging duration, charging rate, charging amount, and charging time of different charging locations, identify and define commonly used charging locations. S12. Based on the charging location, current battery level, charging level, and charging time in the historical charging records, construct a user charging behavior habit model through a neural network, learn the user's charging behavior decision patterns under different times, locations, and initial battery levels, and output typical decision results.

[0006] In a preferred embodiment, S11 includes the following steps: S111. For charging locations in historical charging records, aggregate the coordinates of all charging locations within a 500-meter radius into a single logical location ID and record it as... Where j is the location index, for each logical location Based on the set of all historical charging records that occurred at this location, the weight of each charging record is calculated based on the recording time of the charging record. : ; in, Represents the attenuation coefficient, and , These represent the current time and the charging time of the i-th charging record, respectively. S112, Combining weights Weighted total charging capacity at each charging location Weighted total charging time at each charging location and the weighted usage frequency of each charging location. The overall score for each charging position j is calculated. : ; in, Represents the weighting coefficient, and M represents the total number of charging locations, determined by... Sort the charging locations in descending order and select the top K charging locations as the most frequently used charging locations for the current user's vehicle. , Let represent the weight of the k-th charging record at charging position j and the charging amount of the k-th charging record at charging position j, respectively. The number of charging records at charging location j. , The charging duration for the k-th record at charging position j. N represents the total number of charging records.

[0007] In a preferred embodiment, S12 includes the following steps: S121. Based on the charging time, current battery level, charging location, and charging amount of different charging records in the historical data, a user charging behavior habit model is constructed through a neural network to output the habitual charging location and charging amount under different charging times and initial battery levels. The specific steps are as follows: For the logical position after aggregating the charging positions in the historical charging records, the position is converted into a numerical form through one-hot encoding. The charging time, current battery level, and charging level in the historical charging records are scaled using minimum-maximum normalization. The historical data is divided into training, validation, and test sets, with a ratio of 7:1.5:1.5. The inputs are charging time and current battery level. The charging location is output through a classification task, and the charging level is output through a regression task. A neural network model is constructed using multi-task learning. The task-specific output layer includes a location prediction head and a battery prediction head, which are activated by the Softmax activation function and the linear activation function, respectively. The location loss uses cross-entropy loss as the loss function, and the battery loss uses mean squared error loss as the loss function. The Adam optimizer is selected to update the parameters of the neural network. The model is trained and evaluated through training set, validation set, and test set to obtain a user charging behavior habit model.

[0008] In a preferred embodiment, step S2 includes the following steps: S21. After the current user's vehicle is connected to the charging pile, based on the current vehicle's charging location, current battery level, and current time, the reference charging location and reference charging amount are output through the user's charging behavior habit model. Based on the GIS system, the actual distance between the current vehicle's charging location and the reference charging location, and the actual distance between the current vehicle's charging location and the commonly used charging location are obtained. The estimated power consumption is calculated, and the charging strategy for the reference charging location and the charging strategy for the commonly used charging location are generated. Each charging strategy includes two modes: slow charging and fast charging. S22. Based on the charging rates at different charging locations, calculate the cost of different charging strategies, and combine this with the estimated cost of charging to full capacity or the baseline charging capacity at the current charging location to generate the savings amount for different charging strategies.

[0009] In a preferred embodiment, S21 includes the following steps: S211. Based on the current battery level of the vehicle and the current time, obtain the reference charging location through a user charging behavior habit model. Compared with the reference charging capacity Obtain the current vehicle location through a GIS system With reference charging position Common charging locations The actual distance between , ; S212, Based on the current average energy consumption rate of new energy vehicles and redundancy coefficient Calculate the estimated power consumption at the reference charging location Estimated power consumption at commonly used charging locations : ; ; S213. Combining the attributes of the charging pile at the current charging location and the vehicle's basic parameters, generate a benchmark charging location charging strategy and a common charging location charging strategy, including two modes: slow charging and fast charging. The slow charging mode of the benchmark charging location strategy involves charging the vehicle at the current charging pile using a slow charging current to estimate the energy consumption at the benchmark charging location, calculating the estimated charging time, the estimated time to move to the benchmark charging location, and the amount of energy charged to the benchmark charging location. The fast charging mode estimates the power consumption when charging at the current charging station to the reference charging location using a fast charging current, calculates the estimated charging time, the estimated time to move to the reference charging location, and the estimated amount of power to be charged to the reference charging location. The time consumed; The slow charging mode of the frequently used charging location strategy calculates the estimated power consumption of charging at the current charging station to the frequently used charging location, the estimated charging time, the estimated time to move to the frequently used charging location, and the time to fully charge. The fast charging mode calculates the estimated power consumption of charging at the current charging station to the frequently used charging location, the estimated charging time, the estimated time to move to the frequently used charging location, and the time to fully charge.

[0010] In a preferred embodiment, S22 includes the following steps: S221. For different charging strategies, combine the current charging location with the charging pile to charge normally to the reference charging level. Or, based on the full charge amount, calculate the savings for different charging strategies, specifically including the following steps: The charging strategy for the baseline charging location is combined with the charging rate at the current charging location. Down-charging Consumption amount The reference charging position is at the current battery level. Downcharge Consumption amount Calculate the amount of savings ,in Charging rates for the baseline charging location; Charging strategies for commonly used charging locations, combined with the current charging rate at the charging location. Down-charging Consumption amount The most common charging location is at the current battery level. Downcharge Consumption amount Calculate the amount of savings ,in Charging rates for frequently used charging locations, The battery is fully charged. S222. Display the amount saved under different charging strategies, and also display the estimated cost of regular charging at the current charging station.

[0011] A new energy vehicle charging management system includes a data acquisition and behavior modeling module and a charging strategy generation module; The data acquisition and behavior modeling module collects the current user's vehicle historical charging records, defines commonly used charging locations based on the collected vehicle historical charging records, and constructs a user charging behavior habit model through a neural network in conjunction with the vehicle historical charging records. The charging strategy generation module, after the current user's vehicle connects to the charging pile, outputs a benchmark charging location and benchmark charging amount based on the current vehicle's charging location, current battery level, and current time, using a user charging behavior habit model. It also obtains the actual distance between the current vehicle's charging location and the benchmark charging location, as well as the actual distance between the current vehicle's charging location and frequently used charging locations, based on a GIS system. The module calculates the estimated power consumption and generates charging strategies for the benchmark charging location and frequently used charging locations. Simultaneously, based on the charging rates for different charging locations, it calculates the cost of different charging strategies. Combining this with the estimated cost of charging to full capacity or the benchmark charging amount at the current charging location, the module generates the savings amount for different charging strategies.

[0012] The beneficial effects of this invention are as follows: 1. This invention analyzes users' vehicle charging habits to identify their frequently used charging locations and builds a user charging behavior habit model. Based on the user's charging location and battery level each time they charge, it outputs a baseline charging location and baseline charging level, and generates different slow and fast charging strategies based on the frequently used charging locations. It also displays the amount saved compared to charging normally at the current charging location, making it easier for users to choose a charging strategy based on their current situation. This helps users choose the most cost-effective charging solution based on their economic situation and charging needs, thereby reducing daily charging expenses. 2. This invention constructs a model by analyzing users' vehicle charging behavior habits, which can output a benchmark charging location and benchmark charging capacity that conforms to the user's habits. It fully considers each user's unique charging preferences and usage scenarios, making the generated charging strategy more in line with the user's actual needs and providing a personalized charging service experience. At the same time, it provides users with a variety of charging strategy options and clearly displays the amount saved by each strategy compared to traditional charging methods, making it convenient for users to choose the appropriate charging strategy based on their current time arrangement, power demand, economic situation and other factors, thus enhancing practicality. 3. By providing slow charging mode and fast charging mode under each charging strategy, this invention allows users to weigh battery maintenance based on actual needs, thus enhancing functionality. Attached Figure Description

[0013] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0014] Figure 1 This is a flowchart of a new energy vehicle charging management method according to an embodiment of the present invention. Detailed Implementation

[0015] To further illustrate the various embodiments, the present invention provides accompanying drawings, which are part of the disclosure of the present invention. These drawings are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these drawings, those skilled in the art should be able to understand other possible implementation methods and the advantages of the present invention. The components in the drawings are not drawn to scale, and similar component symbols are generally used to represent similar components.

[0016] According to an embodiment of the present invention, a method and system for managing the charging of new energy vehicles are provided.

[0017] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments: Example 1: like Figure 1 As shown, a new energy vehicle charging management method according to an embodiment of the present invention includes the following steps: S1. Collect the current user's vehicle charging history records, define frequently used charging locations based on the collected vehicle charging history records, and build a user charging behavior habit model through a neural network in conjunction with the vehicle charging history records. S11. Collect the current user's vehicle historical charging records, including the charging location, current battery level, charging amount, charging rate, charging time, and charging duration for each charging record. Based on the charging duration, charging rate, charging amount, and charging time of different charging locations, identify and define commonly used charging locations. S111. For charging locations in historical charging records, aggregate the coordinates of all charging locations within a 500-meter radius into a single logical location ID and record it as... Where j is the location index, for each logical location Based on the set of all historical charging records that occurred at this location, the weight of each charging record is calculated based on the recording time of the charging record. : ; in, Represents the attenuation coefficient, and , These represent the current time and the charging time of the i-th charging record, respectively. It should be noted that the attenuation coefficient can be used to measure the impact of data time on the overall data. The closer the data is to 1, the slower the weight of recent data decreases. The closer the value is to 0, the faster the weight of recent data decreases. The specific settings need to be determined by consulting experts in the field based on the actual amount of data.

[0018] S112, Combining weights Weighted total charging capacity at each charging location Weighted total charging time at each charging location and the weighted usage frequency of each charging location. The overall score for each charging position j is calculated. : ; in, Represents the weighting coefficient, and M represents the total number of charging locations, determined by... Sort the charging locations in descending order and select the top K charging locations as the most frequently used charging locations for the current user's vehicle. , Let represent the weight of the k-th charging record at charging position j and the charging amount of the k-th charging record at charging position j, respectively. The number of charging records at charging location j. , The charging duration for the k-th record at charging position j. N represents the total number of charging records.

[0019] It should be noted that, The weighting coefficient is used to measure the importance of the weighted total charging capacity, weighted total charging time, and weighted usage frequency. It is usually set to 0.4, 0.3, and 0.3, but can also be adjusted according to the actual usage and the charging pile at the charging location.

[0020] S12. Based on the charging location, current battery level, charging level, and charging time in the historical charging records, construct a user charging behavior habit model through a neural network, learn the user's charging behavior decision patterns under different times, locations, and initial battery levels, and output typical decision results. S121. Based on the charging time, current battery level, charging location, and charging amount of different charging records in the historical data, a user charging behavior habit model is constructed through a neural network to output the habitual charging location and charging amount under different charging times and initial battery levels. The specific steps are as follows: For the logical position after aggregating the charging positions in the historical charging records, the position is converted into a numerical form through one-hot encoding. The charging time, current battery level, and charging level in the historical charging records are scaled using minimum-maximum normalization. It should be noted that one-hot encoding is used to transform non-numerical data to facilitate input into the neural network model for model construction. Scaling is performed using min-max normalization, where the original data x is scaled to... , These represent the maximum and minimum values ​​of the data column, respectively.

[0021] The historical data is divided into training, validation, and test sets, with a ratio of 7:1.5:1.5. The inputs are charging time and current battery level. The charging location is output through a classification task, and the charging level is output through a regression task. It should be noted that when inputting charging time, the charging time can be encoded into hours and days of the week through periodic coding.

[0022] A neural network model is constructed using multi-task learning. The task-specific output layer includes a location prediction head and a battery prediction head, which are activated by the Softmax activation function and the linear activation function, respectively. The location loss uses cross-entropy loss as the loss function, and the battery loss uses mean squared error loss as the loss function. The Adam optimizer is selected to update the parameters of the neural network. The model is trained and evaluated through training set, validation set, and test set to obtain a user charging behavior habit model.

[0023] It should be noted that the number of neurons in the input layer is equal to the dimension of the input feature vector. The shared hidden layer includes 2-3 fully connected layers, and each layer is followed by a ReLU activation function and an optional Dropout layer to prevent overfitting. During the training of the neural network model, a batch of training data is input into the network to obtain the location probability and predicted battery level. Then, the total loss is calculated based on the network output and the true label. The gradient of the loss function with respect to all network parameters is calculated, and the network parameters are updated based on the gradient using the Adam optimizer. This process is repeated until all training data has been traversed. At the same time, after each training data, the model performance is evaluated on the validation set. When the validation set loss no longer decreases for multiple consecutive training data, training is stopped, and the model parameters are restored to the model of the training data with the best performance on the validation set to prevent overfitting. For location prediction evaluation, accuracy and Top-k accuracy are used, where accuracy is the proportion of correctly predicted locations, and Top-k accuracy represents the proportion of the true location appearing in the top k locations with the highest predicted probability. For battery level prediction evaluation, root mean square error, mean absolute error, and mean absolute percentage error are used for evaluation.

[0024] Example 2: S2. When the user's vehicle is charging, based on the user's current charging location and battery level, the user's charging behavior habit model outputs the baseline charging location and baseline charging level. Combined with the user's current frequently used charging locations, the baseline charging location charging strategy and the frequently used charging location charging strategy are generated, and the amount saved under different charging strategies is calculated. S21. After the current user's vehicle is connected to the charging pile, based on the current vehicle's charging location, current battery level, and current time, the reference charging location and reference charging amount are output through the user's charging behavior habit model. Based on the GIS system, the actual distance between the current vehicle's charging location and the reference charging location, and the actual distance between the current vehicle's charging location and the commonly used charging location are obtained. The estimated power consumption is calculated, and the charging strategy for the reference charging location and the charging strategy for the commonly used charging location are generated. Each charging strategy includes two modes: slow charging and fast charging. S211. Based on the current battery level of the vehicle and the current time, obtain the reference charging location through a user charging behavior habit model. Compared with the reference charging capacity Obtain the current vehicle location through a GIS system With reference charging position Common charging locations The actual distance between , ; It should be noted that the reference charging position and common charging locations It can be multiple or a single, depending on the user behavior analysis based on step S1. The actual distance between various locations in the current city is obtained through the GIS system, which facilitates the calculation of subsequent power consumption and generates multiple charging strategies.

[0025] S212, Based on the current average energy consumption rate of new energy vehicles and redundancy coefficient Calculate the estimated power consumption at the reference charging location Estimated power consumption at commonly used charging locations : ; ; It should be noted that the redundancy factor can prevent battery damage caused by different road conditions. The redundancy factor is usually set to 1.1, but it can also be set according to the actual vehicle performance.

[0026] S213. Combining the attributes of the charging pile at the current charging location and the vehicle's basic parameters, generate a benchmark charging location charging strategy and a common charging location charging strategy, including two modes: slow charging and fast charging. The slow charging mode of the benchmark charging location strategy involves charging the vehicle at the current charging pile using a slow charging current to estimate the energy consumption at the benchmark charging location, calculating the estimated charging time, the estimated time to move to the benchmark charging location, and the amount of energy charged to the benchmark charging location. The fast charging mode estimates the power consumption when charging at the current charging station to the reference charging location using a fast charging current, calculates the estimated charging time, the estimated time to move to the reference charging location, and the estimated amount of power to be charged to the reference charging location. The time consumed; The slow charging mode of the frequently used charging location strategy calculates the estimated power consumption of charging at the current charging station to the frequently used charging location, the estimated charging time, the estimated time to move to the frequently used charging location, and the time to fully charge. The fast charging mode calculates the estimated power consumption of charging at the current charging station to the frequently used charging location, the estimated charging time, the estimated time to move to the frequently used charging location, and the time to fully charge.

[0027] It should be noted that different charging locations and charging piles have different slow and fast charging currents for vehicles, and different currents cause varying degrees of battery wear. By providing different charging strategies and modes, we can meet the actual needs of users. By intuitively displaying the time taken for different charging strategies, users can flexibly choose different charging modes according to their current situation, or install the normal charging mode at the current charging pile. Basic vehicle parameters can be entered after checking the parameters in the vehicle's user manual or official website. The estimated driving time can be obtained from the estimated time of the vehicle's navigation system, and the charging time can be obtained from the time displayed on the vehicle's infotainment system.

[0028] S22. Based on the charging rate at different charging locations, calculate the cost of different charging strategies, and combine the estimated cost of charging to full capacity or the base charging capacity at the current charging location to generate the savings amount for different charging strategies. S221. For different charging strategies, combine the current charging location with the charging pile to charge normally to the reference charging level. Or, based on the full charge amount, calculate the savings for different charging strategies, specifically including the following steps: The charging strategy for the baseline charging location is combined with the charging rate at the current charging location. Down-charging Consumption amount The reference charging position is at the current battery level. Downcharge Consumption amount Calculate the amount of savings ,in Charging rates for the baseline charging location; Charging strategies for commonly used charging locations, combined with the current charging rate at the charging location. Down-charging Consumption amount The most common charging location is at the current battery level. Downcharge Consumption amount Calculate the amount of savings ,in Charging rates for frequently used charging locations, The battery is fully charged. S222. Display the amount saved under different charging strategies, and also display the estimated cost of regular charging at the current charging station.

[0029] It should be noted that by displaying the amount saved under different charging strategies, users can easily choose the charging strategy according to their actual situation. The current charging station's normal charging means that no charging current control is performed, and the highest specification charging is performed directly according to the parameters of the charging station at the current charging location and the vehicle parameters.

[0030] Example 3: A new energy vehicle charging management system includes a data acquisition and behavior modeling module and a charging strategy generation module; The data acquisition and behavior modeling module collects the current user's vehicle history charging records, defines commonly used charging locations based on the collected vehicle history charging records, and builds a user charging behavior habit model through neural networks in combination with the vehicle history charging records. The charging strategy generation module, after the current user's vehicle connects to the charging pile, outputs a baseline charging location and baseline charging capacity based on the current vehicle's charging location, current battery level, and current time, using a user charging behavior habit model. It also obtains the actual distance between the current vehicle's charging location and the baseline charging location, as well as the actual distance between the current vehicle's charging location and frequently used charging locations, based on a GIS system. The module calculates the estimated power consumption and generates charging strategies for the baseline and frequently used charging locations. Simultaneously, based on the charging rates for different charging locations, it calculates the cost of different charging strategies. Combining this with the estimated cost of charging to full capacity or the baseline charging capacity at the current charging location, the module generates the savings amount for different charging strategies.

[0031] In summary, this invention analyzes users' vehicle charging habits to identify their frequently used charging locations and constructs a user charging behavior habit model. Based on the user's charging location and battery level each time they charge, and combined with the user charging behavior habit model, it outputs a baseline charging location and baseline charging capacity. Furthermore, it generates different slow and fast charging strategies based on frequently used charging locations, while displaying the savings compared to normal charging at the current location. This allows users to choose a charging strategy based on their current situation, helping them select the most cost-effective charging solution according to their economic circumstances and charging needs, thus reducing daily charging expenses. By analyzing users' vehicle charging behavior habits to construct a model, it can output baseline charging locations and baseline charging capacities that conform to user habits, fully considering each user's unique charging preferences and usage scenarios. This makes the generated charging strategies more closely aligned with users' actual needs, providing a personalized charging service experience. It also offers users multiple charging strategy options and clearly displays the savings of each strategy compared to traditional charging methods, allowing users to choose the appropriate charging strategy based on their current schedule, battery demand, and economic situation.

[0032] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A new energy vehicle charging management method, characterized in that, The method comprises the following steps: S1, collecting vehicle historical charging records of a current user, defining a common charging position based on the collected vehicle historical charging records, and constructing a user charging behavior habit model through a neural network in combination with the vehicle historical charging records; S2, when the user's vehicle is charging, outputting a reference charging position and a reference charging capacity through the user charging behavior habit model based on the user's current charging position and battery capacity, generating a reference charging position charging strategy and a common charging position charging strategy in combination with the common charging position of the user's vehicle, and calculating the amount of money saved under different charging strategies.

2. The new energy vehicle charging management method according to claim 1, characterized in that, The S1 comprises the following steps: S11, collecting vehicle historical charging records of a current user, including the charging position, current battery capacity, charging capacity, charging rate, charging time and charging duration of each charging record, identifying and defining a common charging position based on the charging duration, charging rate, charging capacity and charging time of different charging positions; S12, constructing a user charging behavior habit model through a neural network according to the charging position, current battery capacity, charging capacity and charging time in the historical charging records, learning the charging behavior decision-making mode of the user under different times, positions and initial capacities, and outputting typical decision results.

3. The new energy vehicle charging management method according to claim 1, characterized in that, The S11 comprises the following steps: S111, for the charging location in the historical charging record, all charging location coordinates within a distance of 500 meters are aggregated into one logical location ID and recorded as where j is the position index, for each logical location Based on the set of all historical charging records occurring at the location, the weight of each charging record is calculated based on the record time of the charging record : ； wherein, represents an attenuation coefficient, and , respectively represent the current time and the charging time of the i-th charging record. S112, combine the weights total charging amount with weights for each charging position total charging duration with weights for each charging position and usage frequency with weights for each charging position to calculate a comprehensive score for each charging position j : ； wherein, represents a weight coefficient, and M is the total number of charging locations, and the first K charging locations are selected as the frequently used charging locations of the current user vehicle by arranging the charging locations in descending order, wherein , , respectively represent the weight of the kth charging record of the charging location j and the charging capacity of the kth charging record of the charging location j, is the number of charging records possessed by the charging location j, , is the charging duration of the kth record of the charging location j, N represents the total number of charging records.

4. The new energy vehicle charging management method according to claim 3, characterized in that, The S12 comprises the following steps: S121, constructing a user charging behavior habit model through a neural network based on the charging time, current battery capacity, charging position and charging capacity of different charging records in the historical records, to output the habit charging position and charging capacity under different charging times and initial capacities, and the specific steps are as follows: The logical position aggregated after the charging position in the historical charging records is converted into a numerical form through one-hot encoding, and the charging time, current battery capacity and charging capacity in the historical charging records are scaled through min-max normalization; The historical data is divided into a training set, a validation set and a test set, wherein the ratio of the training set, the validation set and the test set is 7:1.5:1.5, the input is the charging time and current battery capacity, the charging position is output through a classification task, and the charging capacity is output through a regression task; A neural network model is constructed through multi-task learning, the task-specific output layer includes a position prediction head and a capacity prediction head, the Softmax activation function and the linear activation function are used respectively, the cross-entropy loss is used as the loss function for the position loss, the mean square error loss is used as the loss function for the capacity loss, the Adam optimizer is selected to update the parameters of the neural network, and the model is trained and evaluated through the training set, the validation set and the test set to obtain the user charging behavior habit model.

5. The new energy vehicle charging management method according to claim 4, characterized in that, The S2 comprises the following steps: S21, after the current user vehicle connects the charging pile, based on the current vehicle charging position, the current battery capacity and the current time, the reference charging position and the reference charging capacity are output through the user charging behavior habit model, the actual distance between the current vehicle charging position and the reference charging position and the actual distance between the current vehicle charging position and the commonly used charging position are obtained based on the GIS system, the estimated consumption capacity is calculated, and the charging strategy of the reference charging position and the charging strategy of the commonly used charging position are generated, wherein each charging strategy includes slow charging and fast charging modes; S22, based on the charging rate of different charging positions, the consumption amount of different charging strategies is calculated, and the estimated consumption amount of charging to full power or the reference charging capacity at the current charging position is combined to generate the saving amount of different charging strategies.

6. The new energy vehicle charging management method according to claim 5, characterized in that, The S21 includes the following steps: S211, obtaining a reference charging position based on the current battery level of the current vehicle and the current time through a user charging behavior habit model With the reference charging power , obtaining the current vehicle position through a GIS system With the reference charging position , the commonly used charging position The actual distance between , ; S212、based on the average energy consumption rate of the current new energy vehicle and the redundancy coefficient obtained by calculation and the estimated consumption of the common charging position : ； ； S213、combine the current charging location charging pile attribute and the basic parameters of the vehicle, generate the benchmark charging location charging strategy and the common charging location charging strategy, including slow charging and fast charging two modes, wherein the slow charging mode of the benchmark charging location charging strategy is to charge the benchmark charging location with the slow charging current, estimate the consumed power, calculate the expected charging time, and estimate the expected time consumed for moving to the benchmark charging location and the time consumed for charging to the benchmark charging amount The fast charging mode is to charge the benchmark charging location with the fast charging current, estimate the consumed power, calculate the expected charging time, and estimate the expected time consumed for moving to the benchmark charging location and the time consumed for charging to the benchmark charging amount ​ The slow charging mode of the commonly used charging position charging strategy is to charge the commonly used charging position estimated consumption capacity at the current charging pile with slow charging current, calculate the estimated charging time and the estimated time consumption of moving to the commonly used charging position and the time consumption of charging to full power, and the fast charging mode is to charge the commonly used charging position estimated consumption capacity at the current charging pile with fast charging current, calculate the estimated charging time and the estimated time consumption of moving to the commonly used charging position and the time consumption of charging to full power.

7. The new energy vehicle charging management method according to claim 6, characterized in that, The S22 includes the following steps: S221、For different charging strategies, combine the current charging position charging pile normal charging to the reference charging capacity Or the amount of full power, calculate the saving amount of different charging strategies, which includes the following steps: For the reference charging location charging strategy, combine the current charging location charging rate Lower charging The consumption amount , the reference charging location is charged to The consumption amount , calculate the saving amount , wherein The charging rate of the reference charging location ; For a charging strategy of a common charging location, a current charging location charging rate is combined under the current battery power under the current battery power , the common charging location is charged to under the current battery power under the current battery power , a saving amount is calculated , wherein is a charging rate of the common charging location, is a full battery power; S222, display the saving amount under different charging strategies, and display the estimated consumption amount of the current charging pile regular charging.

8. A new energy vehicle charging management system, characterized in that, The system adopts a new energy vehicle charging management method according to any one of claims 1-7, including a data acquisition and behavior modeling module, a charging strategy generation module; The data acquisition and behavior modeling module collects the current user's vehicle historical charging records, defines the commonly used charging position based on the collected vehicle historical charging records, and constructs the user charging behavior habit model through the neural network combined with the vehicle historical charging records; The charging strategy generation module, after the current user vehicle connects the charging pile, based on the current vehicle charging position, the current battery capacity and the current time, the reference charging position and the reference charging capacity are output through the user charging behavior habit model, the actual distance between the current vehicle charging position and the reference charging position and the actual distance between the current vehicle charging position and the commonly used charging position are obtained based on the GIS system, the estimated consumption capacity is calculated, and the charging strategy of the reference charging position and the charging strategy of the commonly used charging position are generated, and based on the charging rate of different charging positions, the consumption amount of different charging strategies is calculated, and the estimated consumption amount of charging to full power or the reference charging capacity at the current charging position is combined to generate the saving amount of different charging strategies.

Images