Intelligent charging method and system for charging pile

Abstract

The invention provides an intelligent charging method and system for a charging pile in the technical field of charging piles. The method comprises the following steps: S1, creating a charging strategy generation model; s2, acquiring a large amount of historical charging data to construct a data set; s3, training the charging strategy generation model through the data set, and deploying the trained charging strategy generation model to the charging pile; s4, the charging pile obtains an input charging instruction, after the charging instruction is verified, the charging instruction is analyzed to obtain a charging task, and the charging task is input into the charging strategy generation model to obtain a charging strategy; s5, the charging pile executes charging operation on the charging pile based on the charging strategy, and a charging bill is generated and displayed after charging is completed; and S6, the charging pile records the charging log in real time, the charging log is encrypted into an encrypted log, and the encrypted log is stored and backed up. The method has the advantage that the economical efficiency and safety of charging pile operation are greatly improved.

Description

Technical Field

[0001] The present invention relates to the technical field of charging piles, and in particular to an intelligent charging method and system for charging piles. Background Art

[0002] With the increasing popularity of electric vehicles, the demand for charging stations is also increasing. Charging stations transmit electricity to electric vehicles to charge their battery packs, and after charging, they are charged based on the amount of electricity used. However, existing charging stations have the following problems in operation:

[0003] 1. Electricity cost fluctuations are primarily due to three factors: First, the peak-to-valley variation in grid load leads to significant fluctuations in time-of-use electricity prices. The marginal cost of purchasing electricity for charging stations during high-load periods can be two to three times that of off-peak periods. Second, charging stations must pay a base electricity fee to the grid company based on maximum demand. Simultaneous high-power operation of multiple charging stations may trigger tiered billing. Third, there is a lack of dynamic optimization models for allocating fixed costs such as transformer and line losses during power transmission. Furthermore, user charging behavior exhibits uneven temporal and spatial distribution, including: concentrated charging during commuting hours leading to local grid overloads, harmonic mitigation costs caused by high-power fast charging, and additional energy consumption due to battery preheating in low temperatures. This dynamic imbalance between supply and demand results in uncontrollable fluctuations of 10%-25% in charging station operating costs, impacting the economic viability of charging station operations. 2. Regarding data security, user identity information, payment credentials, and charging behavior data collected during the charging process are transmitted in plain text, making them vulnerable to man-in-the-middle attacks.

[0004] Therefore, how to provide a smart charging method and system for charging piles to improve the economy and safety of charging pile operations has become a technical problem that needs to be solved urgently. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide an intelligent charging method and system for charging piles, so as to improve the economy and safety of charging pile operations.

[0006] In a first aspect, the present invention provides a method for intelligent charging of a charging pile, comprising the following steps:

[0007] Step S1: creating a charging strategy generation model for outputting a charging strategy, and setting a loss function of the charging strategy generation model;

[0008] Step S2: Acquire a large amount of historical charging data, perform preprocessing including data cleaning and data labeling on each of the historical charging data, and then construct a data set;

[0009] Step S3: training a charging strategy generation model using the data set and the loss function, and deploying the trained charging strategy generation model to a charging pile;

[0010] Step S4: The charging pile obtains an input charging instruction, verifies the charging instruction, parses the charging instruction to obtain a charging task, and inputs the charging task into a charging strategy generation model to obtain a charging strategy;

[0011] Step S5: The charging pile performs a charging operation on the charging pile based on the charging strategy, and generates and displays a charging bill after charging is completed;

[0012] Step S6: The charging pile records the charging log in real time, encrypts the charging log into an encrypted log, and stores and backs up the encrypted log.

[0013] Furthermore, in step S1, the charging strategy generation model is constructed based on an input module, a feature extraction module, a prediction module, and a strategy generation module; the input module, the feature extraction module, the prediction module, and the strategy generation module are connected in sequence;

[0014] The input module is used to perform data cleaning and data normalization on the input charging data; the feature extraction module is used to extract charging behavior features, electricity price features, load features, and environmental features from the charging data; the prediction module is constructed based on a long short-term memory network and is used to output charging behavior prediction data, electricity price prediction data, load prediction data, and environmental prediction data based on the charging behavior features, electricity price features, load features, and environmental features; the strategy generation module is used to output a charging strategy based on the charging behavior prediction data, electricity price prediction data, load prediction data, and environmental prediction data; the charging strategy includes at least a charging time period, a charging power, and a charging amount;

[0015] The formula of the loss function is:

[0016]

[0017] Among them, L represents the loss value of the loss function; α and β both represent weight coefficients; P t represents the electricity price at time t; E t Indicates the charge capacity at time t; L t represents the grid load at time t; T represents the charging time.

[0018] Furthermore, the step S2 is specifically as follows:

[0019] Obtain a large amount of historical charging data including at least charging time, charging duration, real-time electricity price, charging power, charging amount, vehicle entry time, vehicle departure time, ambient temperature, ambient humidity, regional vehicle density, power loss rate, grid load, and charging demand;

[0020] performing data cleaning on each of the historical charging data, including at least filling missing values, removing outliers, merging duplicate data, unifying uppercase and lowercase letters, and unifying date formats; annotating each of the historical charging data after data cleaning with a charging strategy to complete preprocessing of each of the historical charging data; and constructing a data set based on each of the preprocessed historical charging data;

[0021] The step S3 is specifically as follows:

[0022] Based on the eight-fold cross-validation method, the data set is divided into a training set and a validation set. The charging strategy generation model is trained using the training set. During the training process, the hyperparameters of the charging strategy generation model, including at least the learning rate, random dropout rate, batch size, and number of iterations, are continuously optimized until the loss value of the loss function is less than the preset loss threshold. The trained charging strategy generation model is then verified using the validation set to determine whether the accuracy of the charging strategy generation is greater than the preset accuracy threshold. If not, the verification fails, and the training set is expanded to continue training. If so, the verification is successful, the training is terminated, and the trained charging strategy generation model is deployed to the charging pile.

[0023] Furthermore, the step S4 is specifically as follows:

[0024] The charging pile receives an input charging instruction carrying a charging task, a timestamp, and a hash value, where the hash value is obtained by hashing the charging task and the timestamp; the charging task includes at least one of a charging cutoff condition, a charging start time, a cost upper limit, a power upper limit, and a power percentage; the charging cutoff condition is that the total charging cost reaches the cost upper limit, the charged power reaches the power upper limit, or the current power of the battery pack reaches the power percentage;

[0025] The charging pile parses the charging instruction to obtain a charging task, a timestamp, and a hash value, performs an integrity check on the charging task and the timestamp using the hash value, and then performs an aging check using the timestamp;

[0026] The charging pile obtains real-time electricity price, ambient temperature, ambient humidity, regional vehicle density, power loss rate and grid load, and inputs the real-time electricity price, ambient temperature, ambient humidity, regional vehicle density, power loss rate, grid load and charging task into a charging strategy generation model to obtain a charging strategy.

[0027] Furthermore, the step S5 is specifically as follows:

[0028] The charging pile performs a charging operation on the charging pile based on the charging strategy, generates a charging bill based on the real-time electricity price, charging amount, and power loss rate after charging is completed, displays the charging bill on a display screen, encrypts the charging bill into an encrypted bill using an AES algorithm, and pushes the encrypted bill to an associated client via the HTTPS protocol;

[0029] The step S6 is specifically as follows:

[0030] The real-time record of the charging station shall include at least the user account, vehicle model, vehicle serial number, charging task and charging log of the charging bill;

[0031] The charging pile creates a pair of public and private keys through the RSA algorithm, maps the public key through a preset mapping rule to obtain a first key, encrypts the first key through the IDEA algorithm to obtain a second key, performs MAC calculation on the charging log to obtain a MAC value, encrypts the charging log, the second key, and the MAC value through the RC6 algorithm to obtain first-level encrypted data, shifts each character of the first-level encrypted data to the right by 3 bits to obtain second-level encrypted data, encrypts the second-level encrypted data through the ECDSA algorithm to obtain third-level encrypted data, shifts each character of the third-level encrypted data to the left by 7 bits to obtain fourth-level encrypted data, encrypts the fourth-level encrypted data through the XTEA algorithm to obtain an encrypted log, and stores and distributes the encrypted log.

[0032] In a second aspect, the present invention provides a charging pile intelligent charging system, comprising the following modules:

[0033] a charging strategy generation model creation module, configured to create a charging strategy generation model for outputting a charging strategy, and to set a loss function of the charging strategy generation model;

[0034] A data set construction module is used to obtain a large amount of historical charging data, and construct a data set after preprocessing each of the historical charging data including data cleaning and data labeling;

[0035] A charging strategy generation model training module is used to train the charging strategy generation model using the data set and the loss function, and deploy the trained charging strategy generation model to the charging pile;

[0036] A charging strategy generation module is used to obtain a charging instruction input by the charging pile, verify the charging instruction, parse the charging instruction to obtain a charging task, and input the charging task into a charging strategy generation model to obtain a charging strategy;

[0037] A charging module, configured to cause the charging pile to perform charging operations on the charging pile based on the charging strategy, and to generate and display a charging bill after charging is completed;

[0038] The charging log management module is used for the charging pile to record the charging log in real time, encrypt the charging log into an encrypted log, and store and back up the encrypted log.

[0039] Furthermore, in the charging strategy generation model creation module, the charging strategy generation model is constructed based on an input module, a feature extraction module, a prediction module, and a strategy generation module; the input module, the feature extraction module, the prediction module, and the strategy generation module are connected in sequence;

[0040] The input module is used to perform data cleaning and data normalization on the input charging data; the feature extraction module is used to extract charging behavior features, electricity price features, load features, and environmental features from the charging data; the prediction module is constructed based on a long short-term memory network and is used to output charging behavior prediction data, electricity price prediction data, load prediction data, and environmental prediction data based on the charging behavior features, electricity price features, load features, and environmental features; the strategy generation module is used to output a charging strategy based on the charging behavior prediction data, electricity price prediction data, load prediction data, and environmental prediction data; the charging strategy includes at least a charging time period, a charging power, and a charging amount;

[0041] The formula of the loss function is:

[0042]

[0043] Among them, L represents the loss value of the loss function; α and β both represent weight coefficients; P t represents the electricity price at time t; E t Indicates the charge capacity at time t; L t represents the grid load at time t; T represents the charging time.

[0044] Furthermore, the dataset construction module is specifically used to:

[0045] Obtain a large amount of historical charging data including at least charging time, charging duration, real-time electricity price, charging power, charging amount, vehicle entry time, vehicle departure time, ambient temperature, ambient humidity, regional vehicle density, power loss rate, grid load, and charging demand;

[0046] performing data cleaning on each of the historical charging data, including at least filling missing values, removing outliers, merging duplicate data, unifying uppercase and lowercase letters, and unifying date formats; annotating each of the historical charging data after data cleaning with a charging strategy to complete preprocessing of each of the historical charging data; and constructing a data set based on each of the preprocessed historical charging data;

[0047] The charging strategy generation model training module is specifically used to:

[0048] Based on the eight-fold cross-validation method, the data set is divided into a training set and a validation set. The charging strategy generation model is trained using the training set. During the training process, the hyperparameters of the charging strategy generation model, including at least the learning rate, random dropout rate, batch size, and number of iterations, are continuously optimized until the loss value of the loss function is less than the preset loss threshold. The trained charging strategy generation model is then verified using the validation set to determine whether the accuracy of the charging strategy generation is greater than the preset accuracy threshold. If not, the verification fails, and the training set is expanded to continue training. If so, the verification is successful, the training is terminated, and the trained charging strategy generation model is deployed to the charging pile.

[0049] Furthermore, the charging strategy generation module is specifically used to:

[0050] The charging pile receives an input charging instruction carrying a charging task, a timestamp, and a hash value, where the hash value is obtained by hashing the charging task and the timestamp; the charging task includes at least one of a charging cutoff condition, a charging start time, a cost upper limit, a power upper limit, and a power percentage; the charging cutoff condition is that the total charging cost reaches the cost upper limit, the charged power reaches the power upper limit, or the current power of the battery pack reaches the power percentage;

[0051] The charging pile parses the charging instruction to obtain a charging task, a timestamp, and a hash value, performs an integrity check on the charging task and the timestamp using the hash value, and then performs an aging check using the timestamp;

[0052] The charging pile obtains real-time electricity price, ambient temperature, ambient humidity, regional vehicle density, power loss rate and grid load, and inputs the real-time electricity price, ambient temperature, ambient humidity, regional vehicle density, power loss rate, grid load and charging task into a charging strategy generation model to obtain a charging strategy.

[0053] Furthermore, the charging module is specifically used for:

[0054] The charging pile performs a charging operation on the charging pile based on the charging strategy, generates a charging bill based on the real-time electricity price, charging amount, and power loss rate after charging is completed, displays the charging bill on a display screen, encrypts the charging bill into an encrypted bill using an AES algorithm, and pushes the encrypted bill to an associated client via the HTTPS protocol;

[0055] The charging log management module is specifically used to:

[0056] The real-time record of the charging station shall include at least the user account, vehicle model, vehicle serial number, charging task and charging log of the charging bill;

[0057] The charging pile creates a pair of public and private keys through the RSA algorithm, maps the public key through a preset mapping rule to obtain a first key, encrypts the first key through the IDEA algorithm to obtain a second key, performs MAC calculation on the charging log to obtain a MAC value, encrypts the charging log, the second key, and the MAC value through the RC6 algorithm to obtain first-level encrypted data, shifts each character of the first-level encrypted data to the right by 3 bits to obtain second-level encrypted data, encrypts the second-level encrypted data through the ECDSA algorithm to obtain third-level encrypted data, shifts each character of the third-level encrypted data to the left by 7 bits to obtain fourth-level encrypted data, encrypts the fourth-level encrypted data through the XTEA algorithm to obtain an encrypted log, and stores and distributes the encrypted log.

[0058] The advantages of the present invention are:

[0059] 1. Create a charging strategy generation model for outputting charging strategies and set the loss function of the charging strategy generation model; then obtain a large amount of historical charging data, perform pre-processing on each historical charging data including data cleaning and data labeling, and then construct a data set; train the charging strategy generation model through the data set and the loss function, and deploy the trained charging strategy generation model to the charging pile; the charging pile obtains the input charging instruction, verifies the charging instruction, parses the charging instruction to obtain the charging task, inputs the charging task into the charging strategy generation model to obtain the charging strategy, performs charging operations on the charging pile based on the charging strategy, and generates and displays the charging bill after charging is completed; the charging pile records the charging log in real time, encrypts the charging log into an encrypted log, and stores and backs up the encrypted log; that is, the electric vehicle is charged based on the charging strategy generated by the pre-trained charging strategy generation model, because the charging strategy generation model is constructed based on the input module, feature extraction module, prediction module and strategy generation module; input module Used to perform data cleaning and data normalization on the input charging data; the feature extraction module is used to extract charging behavior features, electricity price features, load features and environmental features from the charging data; the prediction module is constructed based on the long short-term memory network, and is used to output charging behavior prediction data, electricity price prediction data, load prediction data and environmental prediction data based on the charging behavior features, electricity price features, load characteristics and environmental characteristics; the strategy generation module is used to output the charging strategy based on the charging behavior prediction data, electricity price prediction data, load prediction data and environmental prediction data, that is, the charging behavior features, electricity price features, load characteristics and environmental features are extracted from the charging data to generate the charging strategy, and the loss function of the charging strategy generation model combines the charging cost and the power burden (grid load), so that the generated charging strategy can better balance the charging cost and the power burden, and security measures are taken for the transmission and storage of charging instructions, charging bills and charging logs, which ultimately greatly improves the economy and safety of charging pile operations.

[0060] 2. The charging strategy generation model is trained using a dataset constructed from historical charging data including charging time, charging duration, real-time electricity price, charging power, charging amount, vehicle entry time, vehicle departure time, ambient temperature, ambient humidity, regional vehicle density, power loss rate, grid load, and charging demand. This allows the charging strategy generation model to learn multi-dimensional features from the dataset, thereby greatly improving the quality of charging strategy generation.

[0061] 3. By performing data cleaning on each historical charging data, including at least filling missing values, removing outliers, merging duplicate data, unifying uppercase and lowercase letters, and unifying date formats, the quality of the dataset is greatly improved, thereby greatly improving the training effect of the charging strategy generation model.

[0062] 4. The dataset is divided into training and validation sets using the 8-fold cross-validation method. This method divides the dataset into 8 subsets (or "folds") of roughly equal size. In each iteration, 7 subsets are used as training sets and the remaining 1 subset is used as the validation set. This process is repeated 8 times, with a different subset selected as the validation set each time. This effectively reduces the bias caused by a single random partition, thereby greatly improving the performance of the charging strategy generation model.

[0063] 5. By setting the charging instruction to carry the charging task, timestamp and hash value, the integrity check can be performed through the hash value and the timeliness check can be performed through the timestamp to avoid performing charging operations based on incorrect or tampered charging instructions, thereby greatly improving the safety of charging.

[0064] 6. The charging bill is encrypted into an encrypted bill using the AES algorithm, and the encrypted bill is pushed to the associated client via the HTTPS protocol to prevent the charging bill from being stolen in plain text during transmission. The HTTPS protocol is a secure transmission protocol, and dual security measures are taken, which greatly improves the security of charging bill transmission.

[0065] 7. By recording the charging log in real time, including at least the user account, vehicle model, vehicle serial number, charging task and charging bill, it is convenient for later tracing.

[0066] 8. Create a public key and private key pair using the RSA algorithm. Map the public key using the mapping rule to obtain the first key. Encrypt the first key using the IDEA algorithm to obtain the second key. Perform a MAC calculation on the charging log to obtain a MAC value. Encrypt the charging log, the second key, and the MAC value using the RC6 algorithm to obtain level 1 encrypted data. Shift each character of the level 1 encrypted data right by 3 bits to obtain level 2 encrypted data. Encrypt the level 2 encrypted data using the ECDSA algorithm to obtain level 3 encrypted data. Shift each character of the level 3 encrypted data left by 7 bits to obtain level 4 encrypted data. Encrypt the level 4 encrypted data using the XTEA algorithm to obtain an encrypted log. The encrypted log is stored and distributedly backed up. This means that multiple encryption algorithms and data transformation rules are used to encrypt the charging log. Without knowing any of the encryption algorithms or data transformation rules, the encrypted log cannot be decrypted. At least 10 security measures are implemented (RSA algorithm, mapping rule, IDEA algorithm, MAC calculation, RC6 algorithm, right shift by 3 bits, ECDSA algorithm, left shift by 7 bits, XTEA algorithm, and distributed backup), greatly improving the security of charging log storage. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0068] Figure 1 It is a flow chart of a charging pile intelligent charging method of the present invention.

[0069] Figure 2 It is a structural diagram of an intelligent charging system for charging piles of the present invention. DETAILED DESCRIPTION

[0070] The technical solution in the embodiment of the present application has the following overall idea: the electric vehicle is charged based on the charging strategy generated by the charging strategy generation model constructed based on the input module, feature extraction module, prediction module and strategy generation module; the input module is used to perform data cleaning and data normalization on the input charging data; the feature extraction module is used to extract charging behavior characteristics, electricity price characteristics, load characteristics and environmental characteristics from the charging data; the prediction module is constructed based on the long short-term memory network, and is used to output charging behavior prediction data, electricity price prediction data, load prediction data and environmental prediction data based on the charging behavior characteristics, electricity price characteristics, load characteristics and environmental characteristics; the strategy generation module is used to output the charging strategy based on the charging behavior prediction data, electricity price prediction data, load prediction data and environmental prediction data, that is, the charging behavior characteristics, electricity price characteristics, load characteristics and environmental characteristics are extracted from the charging data to generate the charging strategy, and the loss function of the charging strategy generation model combines the charging cost and the power burden, so that the generated charging strategy can better balance the charging cost and the power burden, and security measures are taken for the transmission and storage of charging instructions, charging bills and charging logs, thereby improving the economy and safety of charging pile operations.

[0071] Please refer to Figures 1 to 2 As shown, a preferred embodiment of the intelligent charging method of a charging pile of the present invention includes the following steps:

[0072] Step S1: creating a charging strategy generation model for outputting a charging strategy, and setting a loss function of the charging strategy generation model;

[0073] Step S2: Acquire a large amount of historical charging data, perform preprocessing including data cleaning and data labeling on each of the historical charging data, and then construct a data set;

[0074] Step S3: training a charging strategy generation model using the data set and the loss function, and deploying the trained charging strategy generation model to a charging pile;

[0075] Step S4: The charging pile obtains an input charging instruction, verifies the charging instruction, parses the charging instruction to obtain a charging task, and inputs the charging task into a charging strategy generation model to obtain a charging strategy;

[0076] Step S5: The charging pile performs a charging operation on the charging pile based on the charging strategy, and generates and displays a charging bill after charging is completed;

[0077] Step S6: The charging pile records the charging log in real time, encrypts the charging log into an encrypted log, and stores and backs up the encrypted log.

[0078] In step S1, the charging strategy generation model is constructed based on an input module, a feature extraction module, a prediction module, and a strategy generation module; the input module, the feature extraction module, the prediction module, and the strategy generation module are connected in sequence;

[0079] The input module is used to perform data cleaning and data normalization on the input charging data; the feature extraction module is used to extract charging behavior features, electricity price features, load features, and environmental features from the charging data; the prediction module is constructed based on a long short-term memory network and is used to output charging behavior prediction data, electricity price prediction data, load prediction data, and environmental prediction data based on the charging behavior features, electricity price features, load features, and environmental features; the strategy generation module is used to output a charging strategy based on the charging behavior prediction data, electricity price prediction data, load prediction data, and environmental prediction data; the charging strategy includes at least a charging time period, a charging power, and a charging amount;

[0080] The formula of the loss function is:

[0081]

[0082] Among them, L represents the loss value of the loss function; α and β both represent weight coefficients; P t represents the electricity price at time t; E t Indicates the charge capacity at time t; L t represents the grid load at time t; T represents the charging time.

[0083] The step S2 is specifically as follows:

[0084] Obtain a large amount of historical charging data including at least charging time, charging duration, real-time electricity price, charging power, charging amount, vehicle entry time, vehicle departure time, ambient temperature, ambient humidity, regional vehicle density, power loss rate, grid load, and charging demand;

[0085] The charging strategy generation model is trained using a dataset constructed from historical charging data including charging time, charging duration, real-time electricity price, charging power, charging amount, vehicle entry time, vehicle departure time, ambient temperature, ambient humidity, regional vehicle density, power loss rate, grid load, and charging demand. This allows the charging strategy generation model to learn multi-dimensional features from the dataset, thereby greatly improving the quality of charging strategy generation.

[0086] performing data cleaning on each of the historical charging data, including at least filling missing values, removing outliers, merging duplicate data, unifying uppercase and lowercase letters, and unifying date formats; annotating each of the historical charging data after data cleaning with a charging strategy to complete preprocessing of each of the historical charging data; and constructing a data set based on each of the preprocessed historical charging data;

[0087] By performing data cleaning on each historical charging data, including at least filling missing values, removing outliers, merging duplicate data, unifying uppercase and lowercase letters, and unifying date formats, the quality of the dataset is greatly improved, thereby greatly improving the training effect of the charging strategy generation model.

[0088] The step S3 is specifically as follows:

[0089] Based on the eight-fold cross-validation method, the data set is divided into a training set and a validation set. The charging strategy generation model is trained using the training set. During the training process, the hyperparameters of the charging strategy generation model, including at least the learning rate, random dropout rate, batch size, and number of iterations, are continuously optimized until the loss value of the loss function is less than the preset loss threshold. The trained charging strategy generation model is then verified using the validation set to determine whether the accuracy of the charging strategy generation is greater than the preset accuracy threshold. If not, the verification fails, and the training set is expanded to continue training. If so, the verification is successful, the training is terminated, and the trained charging strategy generation model is deployed to the charging pile.

[0090] The dataset is divided into a training set and a validation set using the eight-fold cross-validation method. That is, the dataset is divided into eight subsets (or "folds") of roughly equal size. In each iteration, seven subsets are used as training sets and the remaining one subset is used as a validation set. This process is repeated eight times, with a different subset selected as the validation set each time. This can effectively reduce the deviation caused by a single random partition, thereby greatly improving the performance of the charging strategy generation model.

[0091] The step S4 is specifically as follows:

[0092] The charging pile receives an input charging instruction carrying a charging task, a timestamp, and a hash value, where the hash value is obtained by hashing the charging task and the timestamp; the charging task includes at least one of a charging cutoff condition, a charging start time, a cost upper limit, a power upper limit, and a power percentage; the charging cutoff condition is that the total charging cost reaches the cost upper limit, the charged power reaches the power upper limit, or the current power of the battery pack reaches the power percentage;

[0093] By setting the charging instruction to carry the charging task, timestamp and hash value, the integrity check can be performed through the hash value and the timeliness check can be performed through the timestamp to avoid executing charging operations based on erroneous or tampered charging instructions, thereby greatly improving the safety of charging.

[0094] The charging pile parses the charging instruction to obtain a charging task, a timestamp, and a hash value, performs an integrity check on the charging task and the timestamp using the hash value, and then performs an aging check using the timestamp;

[0095] The charging pile obtains real-time electricity price, ambient temperature, ambient humidity, regional vehicle density, power loss rate and grid load, and inputs the real-time electricity price, ambient temperature, ambient humidity, regional vehicle density, power loss rate, grid load and charging task into a charging strategy generation model to obtain a charging strategy.

[0096] The step S5 is specifically as follows:

[0097] The charging pile performs a charging operation on the charging pile based on the charging strategy, generates a charging bill based on the real-time electricity price, charging amount, and power loss rate after charging is completed, displays the charging bill on a display screen, encrypts the charging bill into an encrypted bill using an AES algorithm, and pushes the encrypted bill to an associated client via the HTTPS protocol;

[0098] The charging bill is encrypted into an encrypted bill using the AES algorithm and pushed to the associated client via the HTTPS protocol to prevent the charging bill from being stolen in plain text during transmission. The HTTPS protocol is a secure transmission protocol, and dual security measures are taken, which greatly improves the security of charging bill transmission.

[0099] The step S6 is specifically as follows:

[0100] The real-time record of the charging station shall include at least the user account, vehicle model, vehicle serial number, charging task and charging log of the charging bill;

[0101] By recording charging logs that include at least user account, vehicle model, vehicle serial number, charging tasks, and charging bills in real time, it is easy to trace the source later.

[0102] The charging pile creates a pair of public and private keys through the RSA algorithm, maps the public key through a preset mapping rule to obtain a first key, encrypts the first key through the IDEA algorithm to obtain a second key, performs MAC calculation on the charging log to obtain a MAC value, encrypts the charging log, the second key, and the MAC value through the RC6 algorithm to obtain first-level encrypted data, shifts each character of the first-level encrypted data to the right by 3 bits to obtain second-level encrypted data, encrypts the second-level encrypted data through the ECDSA algorithm to obtain third-level encrypted data, shifts each character of the third-level encrypted data to the left by 7 bits to obtain fourth-level encrypted data, encrypts the fourth-level encrypted data through the XTEA algorithm to obtain an encrypted log, and stores and distributes the encrypted log.

[0103] A pair of public and private keys is created using the RSA algorithm. The public key is mapped using a mapping rule to obtain a first key. The first key is encrypted using the IDEA algorithm to obtain a second key. A MAC calculation is performed on the charging log to obtain a MAC value. The charging log, the second key, and the MAC value are encrypted using the RC6 algorithm to obtain first-level encrypted data. Each character of the first-level encrypted data is shifted right by 3 bits to obtain second-level encrypted data. The second-level encrypted data is encrypted using the ECDSA algorithm to obtain third-level encrypted data. Each character of the third-level encrypted data is shifted left by 7 bits to obtain fourth-level encrypted data. The fourth-level encrypted data is encrypted using the XTEA algorithm to obtain an encrypted log. The encrypted log is stored and distributedly backed up. This means that multiple encryption algorithms and data transformation rules are used to encrypt the charging log. Without knowing any of the encryption algorithms or data transformation rules, the encrypted log cannot be decrypted. At least 10 security measures are implemented (RSA algorithm, mapping rule, IDEA algorithm, MAC calculation, RC6 algorithm, right shift by 3 bits, ECDSA algorithm, left shift by 7 bits, XTEA algorithm, and distributed backup), greatly improving the security of charging log storage.

[0104] A preferred embodiment of the intelligent charging system for charging piles of the present invention includes the following modules:

[0105] a charging strategy generation model creation module, configured to create a charging strategy generation model for outputting a charging strategy, and to set a loss function of the charging strategy generation model;

[0106] A data set construction module is used to obtain a large amount of historical charging data, and construct a data set after preprocessing each of the historical charging data including data cleaning and data labeling;

[0107] A charging strategy generation model training module is used to train the charging strategy generation model using the data set and the loss function, and deploy the trained charging strategy generation model to the charging pile;

[0108] A charging strategy generation module is used to obtain a charging instruction input by the charging pile, verify the charging instruction, parse the charging instruction to obtain a charging task, and input the charging task into a charging strategy generation model to obtain a charging strategy;

[0109] A charging module, configured to cause the charging pile to perform charging operations on the charging pile based on the charging strategy, and to generate and display a charging bill after charging is completed;

[0110] The charging log management module is used for the charging pile to record the charging log in real time, encrypt the charging log into an encrypted log, and store and back up the encrypted log.

[0111] In the charging strategy generation model creation module, the charging strategy generation model is constructed based on an input module, a feature extraction module, a prediction module, and a strategy generation module; the input module, the feature extraction module, the prediction module, and the strategy generation module are connected in sequence;

[0112] The input module is used to perform data cleaning and data normalization on the input charging data; the feature extraction module is used to extract charging behavior features, electricity price features, load features, and environmental features from the charging data; the prediction module is constructed based on a long short-term memory network and is used to output charging behavior prediction data, electricity price prediction data, load prediction data, and environmental prediction data based on the charging behavior features, electricity price features, load features, and environmental features; the strategy generation module is used to output a charging strategy based on the charging behavior prediction data, electricity price prediction data, load prediction data, and environmental prediction data; the charging strategy includes at least a charging time period, a charging power, and a charging amount;

[0113] The formula of the loss function is:

[0114]

[0115] Among them, L represents the loss value of the loss function; α and β both represent weight coefficients; P t represents the electricity price at time t; E t Indicates the charge capacity at time t; L t represents the grid load at time t; T represents the charging time.

[0116] The dataset construction module is specifically used for:

[0117] Obtain a large amount of historical charging data including at least charging time, charging duration, real-time electricity price, charging power, charging amount, vehicle entry time, vehicle departure time, ambient temperature, ambient humidity, regional vehicle density, power loss rate, grid load, and charging demand;

[0118] The charging strategy generation model is trained using a dataset constructed from historical charging data including charging time, charging duration, real-time electricity price, charging power, charging amount, vehicle entry time, vehicle departure time, ambient temperature, ambient humidity, regional vehicle density, power loss rate, grid load, and charging demand. This allows the charging strategy generation model to learn multi-dimensional features from the dataset, thereby greatly improving the quality of charging strategy generation.

[0119] performing data cleaning on each of the historical charging data, including at least filling missing values, removing outliers, merging duplicate data, unifying uppercase and lowercase letters, and unifying date formats; annotating each of the historical charging data after data cleaning with a charging strategy to complete preprocessing of each of the historical charging data; and constructing a data set based on each of the preprocessed historical charging data;

[0120] By performing data cleaning on each historical charging data, including at least filling missing values, removing outliers, merging duplicate data, unifying uppercase and lowercase letters, and unifying date formats, the quality of the dataset is greatly improved, thereby greatly improving the training effect of the charging strategy generation model.

[0121] The charging strategy generation model training module is specifically used to:

[0122] Based on the eight-fold cross-validation method, the data set is divided into a training set and a validation set. The charging strategy generation model is trained using the training set. During the training process, the hyperparameters of the charging strategy generation model, including at least the learning rate, random dropout rate, batch size, and number of iterations, are continuously optimized until the loss value of the loss function is less than the preset loss threshold. The trained charging strategy generation model is then verified using the validation set to determine whether the accuracy of the charging strategy generation is greater than the preset accuracy threshold. If not, the verification fails, and the training set is expanded to continue training. If so, the verification is successful, the training is terminated, and the trained charging strategy generation model is deployed to the charging pile.

[0123] The dataset is divided into a training set and a validation set using the eight-fold cross-validation method. That is, the dataset is divided into eight subsets (or "folds") of roughly equal size. In each iteration, seven subsets are used as training sets and the remaining one subset is used as a validation set. This process is repeated eight times, with a different subset selected as the validation set each time. This can effectively reduce the deviation caused by a single random partition, thereby greatly improving the performance of the charging strategy generation model.

[0124] The charging strategy generation module is specifically used to:

[0125] The charging pile receives an input charging instruction carrying a charging task, a timestamp, and a hash value, where the hash value is obtained by hashing the charging task and the timestamp; the charging task includes at least one of a charging cutoff condition, a charging start time, a cost upper limit, a power upper limit, and a power percentage; the charging cutoff condition is that the total charging cost reaches the cost upper limit, the charged power reaches the power upper limit, or the current power of the battery pack reaches the power percentage;

[0126] By setting the charging instruction to carry the charging task, timestamp and hash value, the integrity check can be performed through the hash value and the timeliness check can be performed through the timestamp to avoid executing charging operations based on erroneous or tampered charging instructions, thereby greatly improving the safety of charging.

[0127] The charging pile parses the charging instruction to obtain a charging task, a timestamp, and a hash value, performs an integrity check on the charging task and the timestamp using the hash value, and then performs an aging check using the timestamp;

[0128] The charging pile obtains real-time electricity price, ambient temperature, ambient humidity, regional vehicle density, power loss rate and grid load, and inputs the real-time electricity price, ambient temperature, ambient humidity, regional vehicle density, power loss rate, grid load and charging task into a charging strategy generation model to obtain a charging strategy.

[0129] The charging module is specifically used for:

[0130] The charging pile performs a charging operation on the charging pile based on the charging strategy, generates a charging bill based on the real-time electricity price, charging amount, and power loss rate after charging is completed, displays the charging bill on a display screen, encrypts the charging bill into an encrypted bill using an AES algorithm, and pushes the encrypted bill to an associated client via the HTTPS protocol;

[0131] The charging bill is encrypted into an encrypted bill using the AES algorithm and pushed to the associated client via the HTTPS protocol to prevent the charging bill from being stolen in plain text during transmission. The HTTPS protocol is a secure transmission protocol, and dual security measures are taken, which greatly improves the security of charging bill transmission.

[0132] The charging log management module is specifically used to:

[0133] The real-time record of the charging station shall include at least the user account, vehicle model, vehicle serial number, charging task and charging log of the charging bill;

[0134] By recording charging logs that include at least user account, vehicle model, vehicle serial number, charging tasks, and charging bills in real time, it is easy to trace the source later.

[0135] The charging pile creates a pair of public and private keys through the RSA algorithm, maps the public key through a preset mapping rule to obtain a first key, encrypts the first key through the IDEA algorithm to obtain a second key, performs MAC calculation on the charging log to obtain a MAC value, encrypts the charging log, the second key, and the MAC value through the RC6 algorithm to obtain first-level encrypted data, shifts each character of the first-level encrypted data to the right by 3 bits to obtain second-level encrypted data, encrypts the second-level encrypted data through the ECDSA algorithm to obtain third-level encrypted data, shifts each character of the third-level encrypted data to the left by 7 bits to obtain fourth-level encrypted data, encrypts the fourth-level encrypted data through the XTEA algorithm to obtain an encrypted log, and stores and distributes the encrypted log.

[0136] A pair of public and private keys is created using the RSA algorithm. The public key is mapped using a mapping rule to obtain a first key. The first key is encrypted using the IDEA algorithm to obtain a second key. A MAC calculation is performed on the charging log to obtain a MAC value. The charging log, the second key, and the MAC value are encrypted using the RC6 algorithm to obtain first-level encrypted data. Each character of the first-level encrypted data is shifted right by 3 bits to obtain second-level encrypted data. The second-level encrypted data is encrypted using the ECDSA algorithm to obtain third-level encrypted data. Each character of the third-level encrypted data is shifted left by 7 bits to obtain fourth-level encrypted data. The fourth-level encrypted data is encrypted using the XTEA algorithm to obtain an encrypted log. The encrypted log is stored and distributedly backed up. This means that multiple encryption algorithms and data transformation rules are used to encrypt the charging log. Without knowing any of the encryption algorithms or data transformation rules, the encrypted log cannot be decrypted. At least 10 security measures are implemented (RSA algorithm, mapping rule, IDEA algorithm, MAC calculation, RC6 algorithm, right shift by 3 bits, ECDSA algorithm, left shift by 7 bits, XTEA algorithm, and distributed backup), greatly improving the security of charging log storage.

[0137] In summary, the advantages of the present invention are:

[0138] 1. Create a charging strategy generation model for outputting charging strategies and set the loss function of the charging strategy generation model; then obtain a large amount of historical charging data, perform pre-processing on each historical charging data including data cleaning and data labeling, and then construct a data set; train the charging strategy generation model through the data set and the loss function, and deploy the trained charging strategy generation model to the charging pile; the charging pile obtains the input charging instruction, verifies the charging instruction, parses the charging instruction to obtain the charging task, inputs the charging task into the charging strategy generation model to obtain the charging strategy, performs charging operations on the charging pile based on the charging strategy, and generates and displays the charging bill after charging is completed; the charging pile records the charging log in real time, encrypts the charging log into an encrypted log, and stores and backs up the encrypted log; that is, the electric vehicle is charged based on the charging strategy generated by the pre-trained charging strategy generation model, because the charging strategy generation model is constructed based on the input module, feature extraction module, prediction module and strategy generation module; input module Used to perform data cleaning and data normalization on the input charging data; the feature extraction module is used to extract charging behavior features, electricity price features, load features and environmental features from the charging data; the prediction module is constructed based on the long short-term memory network, and is used to output charging behavior prediction data, electricity price prediction data, load prediction data and environmental prediction data based on the charging behavior features, electricity price features, load characteristics and environmental characteristics; the strategy generation module is used to output the charging strategy based on the charging behavior prediction data, electricity price prediction data, load prediction data and environmental prediction data, that is, the charging behavior features, electricity price features, load characteristics and environmental features are extracted from the charging data to generate the charging strategy, and the loss function of the charging strategy generation model combines the charging cost and the power burden (grid load), so that the generated charging strategy can better balance the charging cost and the power burden, and security measures are taken for the transmission and storage of charging instructions, charging bills and charging logs, which ultimately greatly improves the economy and safety of charging pile operations.

[0139] 2. The charging strategy generation model is trained using a dataset constructed from historical charging data including charging time, charging duration, real-time electricity price, charging power, charging amount, vehicle entry time, vehicle departure time, ambient temperature, ambient humidity, regional vehicle density, power loss rate, grid load, and charging demand. This allows the charging strategy generation model to learn multi-dimensional features from the dataset, thereby greatly improving the quality of charging strategy generation.

[0140] 3. By performing data cleaning on each historical charging data, including at least filling missing values, removing outliers, merging duplicate data, unifying uppercase and lowercase letters, and unifying date formats, the quality of the dataset is greatly improved, thereby greatly improving the training effect of the charging strategy generation model.

[0141] 4. The dataset is divided into training and validation sets using the 8-fold cross-validation method. This method divides the dataset into 8 subsets (or "folds") of roughly equal size. In each iteration, 7 subsets are used as training sets and the remaining 1 subset is used as the validation set. This process is repeated 8 times, with a different subset selected as the validation set each time. This effectively reduces the bias caused by a single random partition, thereby greatly improving the performance of the charging strategy generation model.

[0142] 5. By setting the charging instruction to carry the charging task, timestamp and hash value, the integrity check can be performed through the hash value and the timeliness check can be performed through the timestamp to avoid performing charging operations based on incorrect or tampered charging instructions, thereby greatly improving the safety of charging.

[0143] 6. The charging bill is encrypted into an encrypted bill using the AES algorithm, and the encrypted bill is pushed to the associated client via the HTTPS protocol to prevent the charging bill from being stolen in plain text during transmission. The HTTPS protocol is a secure transmission protocol, and dual security measures are taken, which greatly improves the security of charging bill transmission.

[0144] 7. By recording the charging log in real time, including at least the user account, vehicle model, vehicle serial number, charging task and charging bill, it is convenient for later tracing.

[0145] 8. Create a public key and private key pair using the RSA algorithm. Map the public key using the mapping rule to obtain the first key. Encrypt the first key using the IDEA algorithm to obtain the second key. Perform a MAC calculation on the charging log to obtain a MAC value. Encrypt the charging log, the second key, and the MAC value using the RC6 algorithm to obtain level 1 encrypted data. Shift each character of the level 1 encrypted data right by 3 bits to obtain level 2 encrypted data. Encrypt the level 2 encrypted data using the ECDSA algorithm to obtain level 3 encrypted data. Shift each character of the level 3 encrypted data left by 7 bits to obtain level 4 encrypted data. Encrypt the level 4 encrypted data using the XTEA algorithm to obtain an encrypted log. The encrypted log is stored and distributedly backed up. This means that multiple encryption algorithms and data transformation rules are used to encrypt the charging log. Without knowing any of the encryption algorithms or data transformation rules, the encrypted log cannot be decrypted. At least 10 security measures are implemented (RSA algorithm, mapping rule, IDEA algorithm, MAC calculation, RC6 algorithm, right shift by 3 bits, ECDSA algorithm, left shift by 7 bits, XTEA algorithm, and distributed backup), greatly improving the security of charging log storage.

[0146] Although the specific embodiments of the present invention are described above, those skilled in the art should understand that the specific embodiments described are merely illustrative and are not intended to limit the scope of the present invention. Equivalent modifications and changes made by those skilled in the art in accordance with the spirit of the present invention should be included within the scope of protection of the claims of the present invention.

Claims

1. A charging pile intelligent charging method, characterized by: The steps include: Step S1: creating a charging strategy generation model for outputting a charging strategy, and setting a loss function of the charging strategy generation model; Step S2: Acquire a large amount of historical charging data, perform preprocessing including data cleaning and data labeling on each of the historical charging data, and then construct a data set; Step S3: training a charging strategy generation model using the data set and the loss function, and deploying the trained charging strategy generation model to a charging pile; Step S4: The charging pile obtains an input charging instruction, verifies the charging instruction, parses the charging instruction to obtain a charging task, and inputs the charging task into a charging strategy generation model to obtain a charging strategy; Step S5: The charging pile performs a charging operation on the charging pile based on the charging strategy, and generates and displays a charging bill after charging is completed; Step S6: The charging pile records the charging log in real time, encrypts the charging log into an encrypted log, and stores and backs up the encrypted log.

2. The intelligent charging method for a charging pile according to claim 1, characterized in that: In step S1, the charging strategy generation model is constructed based on an input module, a feature extraction module, a prediction module, and a strategy generation module; the input module, the feature extraction module, the prediction module, and the strategy generation module are connected in sequence; The input module is used to perform data cleaning and data normalization on the input charging data; the feature extraction module is used to extract charging behavior features, electricity price features, load features and environmental features from the charging data; The prediction module is constructed based on a long short-term memory network and is used to output charging behavior prediction data, electricity price prediction data, load prediction data, and environmental prediction data based on charging behavior characteristics, electricity price characteristics, load characteristics, and environmental characteristics; the strategy generation module is used to output a charging strategy based on the charging behavior prediction data, electricity price prediction data, load prediction data, and environmental prediction data; the charging strategy includes at least a charging time period, a charging power, and a charging amount; The formula of the loss function is: Among them, L represents the loss value of the loss function; α and β both represent weight coefficients; P t represents the electricity price at time t; E t Indicates the charge capacity at time t; L t represents the grid load at time t; T represents the charging time.

3. The intelligent charging method for a charging pile according to claim 1, wherein: The step S2 is specifically as follows: Obtain a large amount of historical charging data including at least charging time, charging duration, real-time electricity price, charging power, charging amount, vehicle entry time, vehicle departure time, ambient temperature, ambient humidity, regional vehicle density, power loss rate, grid load, and charging demand; performing data cleaning on each of the historical charging data, including at least filling missing values, removing outliers, merging duplicate data, unifying uppercase and lowercase letters, and unifying date formats; annotating each of the historical charging data after data cleaning with a charging strategy to complete preprocessing of each of the historical charging data; and constructing a data set based on each of the preprocessed historical charging data; The step S3 is specifically as follows: Based on the eight-fold cross-validation method, the data set is divided into a training set and a validation set. The charging strategy generation model is trained using the training set. During the training process, the hyperparameters of the charging strategy generation model, including at least the learning rate, random dropout rate, batch size, and number of iterations, are continuously optimized until the loss value of the loss function is less than the preset loss threshold. The trained charging strategy generation model is then verified using the validation set to determine whether the accuracy of the charging strategy generation is greater than the preset accuracy threshold. If not, the verification fails, and the training set is expanded to continue training. If so, the verification is successful, the training is terminated, and the trained charging strategy generation model is deployed to the charging pile.

4. The intelligent charging method for a charging pile according to claim 1, wherein: The step S4 is specifically as follows: The charging pile receives an input charging instruction carrying a charging task, a timestamp, and a hash value, where the hash value is obtained by hashing the charging task and the timestamp; the charging task includes at least one of a charging cutoff condition, a charging start time, a cost upper limit, a power upper limit, and a power percentage; the charging cutoff condition is that the total charging cost reaches the cost upper limit, the charged power reaches the power upper limit, or the current power of the battery pack reaches the power percentage; The charging pile parses the charging instruction to obtain a charging task, a timestamp, and a hash value, performs an integrity check on the charging task and the timestamp using the hash value, and then performs an aging check using the timestamp; The charging pile obtains real-time electricity price, ambient temperature, ambient humidity, regional vehicle density, power loss rate and grid load, and inputs the real-time electricity price, ambient temperature, ambient humidity, regional vehicle density, power loss rate, grid load and charging task into a charging strategy generation model to obtain a charging strategy.

5. The intelligent charging method for a charging pile according to claim 1, characterized in that: The step S5 is specifically as follows: The charging pile performs a charging operation on the charging pile based on the charging strategy, generates a charging bill based on the real-time electricity price, charging amount, and power loss rate after charging is completed, displays the charging bill on a display screen, encrypts the charging bill into an encrypted bill using an AES algorithm, and pushes the encrypted bill to an associated client via the HTTPS protocol; The step S6 is specifically as follows: The real-time record of the charging station shall include at least the user account, vehicle model, vehicle serial number, charging task and charging log of the charging bill; The charging pile creates a pair of public and private keys through the RSA algorithm, maps the public key through a preset mapping rule to obtain a first key, encrypts the first key through the IDEA algorithm to obtain a second key, performs MAC calculation on the charging log to obtain a MAC value, encrypts the charging log, the second key, and the MAC value through the RC6 algorithm to obtain first-level encrypted data, shifts each character of the first-level encrypted data to the right by 3 bits to obtain second-level encrypted data, encrypts the second-level encrypted data through the ECDSA algorithm to obtain third-level encrypted data, shifts each character of the third-level encrypted data to the left by 7 bits to obtain fourth-level encrypted data, encrypts the fourth-level encrypted data through the XTEA algorithm to obtain an encrypted log, and stores and distributes the encrypted log.

6. A charging pile intelligent charging system, characterized by: Includes the following modules: a charging strategy generation model creation module, configured to create a charging strategy generation model for outputting a charging strategy, and to set a loss function of the charging strategy generation model; A data set construction module is used to obtain a large amount of historical charging data, and construct a data set after preprocessing each of the historical charging data including data cleaning and data labeling; A charging strategy generation model training module is used to train the charging strategy generation model using the data set and the loss function, and deploy the trained charging strategy generation model to the charging pile; A charging strategy generation module is used to obtain a charging instruction input by the charging pile, verify the charging instruction, parse the charging instruction to obtain a charging task, and input the charging task into a charging strategy generation model to obtain a charging strategy; A charging module, configured to cause the charging pile to perform charging operations on the charging pile based on the charging strategy, and to generate and display a charging bill after charging is completed; The charging log management module is used for the charging pile to record the charging log in real time, encrypt the charging log into an encrypted log, and store and back up the encrypted log.

7. The intelligent charging system for charging piles according to claim 6, characterized in that: In the charging strategy generation model creation module, the charging strategy generation model is constructed based on an input module, a feature extraction module, a prediction module, and a strategy generation module; the input module, the feature extraction module, the prediction module, and the strategy generation module are connected in sequence; The input module is used to perform data cleaning and data normalization on the input charging data; the feature extraction module is used to extract charging behavior features, electricity price features, load features and environmental features from the charging data; The prediction module is constructed based on a long short-term memory network and is used to output charging behavior prediction data, electricity price prediction data, load prediction data, and environmental prediction data based on charging behavior characteristics, electricity price characteristics, load characteristics, and environmental characteristics; the strategy generation module is used to output a charging strategy based on the charging behavior prediction data, electricity price prediction data, load prediction data, and environmental prediction data; the charging strategy includes at least a charging time period, a charging power, and a charging amount; The formula of the loss function is: Among them, L represents the loss value of the loss function; α and β both represent weight coefficients; P t represents the electricity price at time t; E t Indicates the charge capacity at time t; L t represents the grid load at time t; T represents the charging time.

8. The intelligent charging system for charging piles according to claim 6, characterized in that: The dataset construction module is specifically used for: Obtain a large amount of historical charging data including at least charging time, charging duration, real-time electricity price, charging power, charging amount, vehicle entry time, vehicle departure time, ambient temperature, ambient humidity, regional vehicle density, power loss rate, grid load, and charging demand; performing data cleaning on each of the historical charging data, including at least filling missing values, removing outliers, merging duplicate data, unifying uppercase and lowercase letters, and unifying date formats; annotating each of the historical charging data after data cleaning with a charging strategy to complete preprocessing of each of the historical charging data; and constructing a data set based on each of the preprocessed historical charging data; The charging strategy generation model training module is specifically used to: Based on the eight-fold cross-validation method, the data set is divided into a training set and a validation set. The charging strategy generation model is trained using the training set. During the training process, the hyperparameters of the charging strategy generation model, including at least the learning rate, random dropout rate, batch size, and number of iterations, are continuously optimized until the loss value of the loss function is less than the preset loss threshold. The trained charging strategy generation model is then verified using the validation set to determine whether the accuracy of the charging strategy generation is greater than the preset accuracy threshold. If not, the verification fails, and the training set is expanded to continue training. If so, the verification is successful, the training is terminated, and the trained charging strategy generation model is deployed to the charging pile.

9. The intelligent charging system for charging piles according to claim 6, characterized in that: The charging strategy generation module is specifically used to: The charging pile receives an input charging instruction carrying a charging task, a timestamp, and a hash value, where the hash value is obtained by hashing the charging task and the timestamp; the charging task includes at least one of a charging cutoff condition, a charging start time, a cost upper limit, a power upper limit, and a power percentage; the charging cutoff condition is that the total charging cost reaches the cost upper limit, the charged power reaches the power upper limit, or the current power of the battery pack reaches the power percentage; The charging pile parses the charging instruction to obtain a charging task, a timestamp, and a hash value, performs an integrity check on the charging task and the timestamp using the hash value, and then performs an aging check using the timestamp; The charging pile obtains real-time electricity price, ambient temperature, ambient humidity, regional vehicle density, power loss rate and grid load, and inputs the real-time electricity price, ambient temperature, ambient humidity, regional vehicle density, power loss rate, grid load and charging task into a charging strategy generation model to obtain a charging strategy.

10. The intelligent charging system for charging piles according to claim 6, characterized in that: The charging module is specifically used for: The charging pile performs a charging operation on the charging pile based on the charging strategy, generates a charging bill based on the real-time electricity price, charging amount, and power loss rate after charging is completed, displays the charging bill on a display screen, encrypts the charging bill into an encrypted bill using an AES algorithm, and pushes the encrypted bill to an associated client via the HTTPS protocol; The charging log management module is specifically used to: The real-time record of the charging station shall include at least the user account, vehicle model, vehicle serial number, charging task and charging log of the charging bill; The charging pile creates a pair of public and private keys through the RSA algorithm, maps the public key through a preset mapping rule to obtain a first key, encrypts the first key through the IDEA algorithm to obtain a second key, performs MAC calculation on the charging log to obtain a MAC value, encrypts the charging log, the second key, and the MAC value through the RC6 algorithm to obtain first-level encrypted data, shifts each character of the first-level encrypted data to the right by 3 bits to obtain second-level encrypted data, encrypts the second-level encrypted data through the ECDSA algorithm to obtain third-level encrypted data, shifts each character of the third-level encrypted data to the left by 7 bits to obtain fourth-level encrypted data, encrypts the fourth-level encrypted data through the XTEA algorithm to obtain an encrypted log, and stores and distributes the encrypted log.

Images