Alternating current ordered charging demand acquisition method and device, storage medium and product

Abstract

The invention discloses an alternating current ordered charging demand acquisition method and device, a storage medium and a product, relates to the field of ordered charging of new energy vehicles, and realizes accurate balance between charging demands and pile network loads through vehicle pile digital interaction and intelligent analysis. The method specifically comprises the steps that after a charging pile is connected with a vehicle-mounted charger, communication is established through a PWM signal with a specific duty ratio, and static parameters of an electric vehicle, charging pile operation data and user interaction data are synchronously collected; carrying out abnormal value elimination and normalization processing, and constructing a charging behavior feature vector fused with multi-dimensional features; a K-means algorithm is adopted to divide a scene into high, medium and low priorities, and a correction factor is introduced to construct a differentiated charging power and duration model; and finally integrating the demand list, pushing the integrated demand list to a management platform, and dynamically updating according to a fixed period. According to the method, the problems of poor vehicle pile adaptation, one-sided data, insufficient scene adaptation and the like in the prior art are solved, and the communication compatibility and the data accuracy are improved.

Description

Technical Field

[0001] This invention relates to the field of orderly charging of new energy vehicles, specifically to a method, device, storage medium, and product for obtaining AC orderly charging demand. Background Technology

[0002] With the global energy structure transformation and the advancement of "dual-carbon" goals, the new energy electric vehicle industry has experienced explosive growth, with its ownership continuing to climb, and the construction of charging infrastructure has entered a stage of large-scale expansion. However, the contradiction between the concentrated surge in charging demand and the limited capacity of the power grid is becoming increasingly prominent. Disorderly charging behavior has become a key bottleneck restricting the high-quality development of the industry. Currently, communication between AC charging piles and electric vehicles is achieved through CC / CP. CC is a connection confirmation function, ensuring a reliable connection between the charging plug and the socket, and determining the maximum current carrying capacity of the cable based on the resistance. CP is a control guidance function, determining the status of the power supply device and the vehicle controller based on the CP voltage. During energy transmission, the PWM duty cycle of CP is used to determine the maximum current that the power supply device can provide. The vehicle controller adjusts the charging current according to this information to ensure that it does not exceed the power supply capacity of the power supply device. It is evident that AC charging lacks information exchange on charging demand. During peak hours, a large number of electric vehicles charge simultaneously, which can easily lead to problems such as local power grid overload, voltage drop, and power factor imbalance. This not only affects the safe and stable operation of the power grid but may also trigger a chain reaction such as decreased charging efficiency and increased equipment wear. Conversely, during off-peak hours, charging resources are idle, resulting in a waste of electricity resources and further amplifying the temporal and spatial mismatch between charging supply and demand. Traditional solutions often employ fixed priority classifications or simple rule-based judgments, failing to build a multi-dimensional feature system integrating vehicles, charging piles, and users. This makes it impossible to dynamically adapt to different users' charging scenarios and grid operating conditions. Scenario classification lacks quantitative indicators, resulting in insufficient targeting of charging solutions and difficulty in balancing user needs with grid load optimization goals. Existing charging strategies do not set differentiated goals based on scenario priorities: they cannot guarantee an emergency charging experience for high-demand users, nor can they achieve peak shaving and valley filling for the grid through dynamic power adjustments. More importantly, most solutions are statically generated, lacking real-time update mechanisms and unable to respond to dynamic factors such as charging pile load fluctuations and battery status changes, making it difficult to balance charging efficiency with grid operating efficiency.

[0003] In summary, existing AC charging demand acquisition technologies have significant shortcomings in terms of communication adaptability, comprehensive data collection, intelligent scenario classification, solution differentiation, and dynamic adjustment capabilities, making it difficult to meet the multi-party balancing needs of large-scale charging scenarios. Therefore, developing an AC orderly charging demand acquisition method with high compatibility, accurate data support, intelligent scenario adaptation, and dynamic adjustment capabilities has become an urgent technological requirement in the field of new energy vehicle charging. Summary of the Invention

[0004] The purpose of this invention is to provide a method, device, storage medium, and product for acquiring orderly charging demand. It establishes digital communication between the charging pile and the on-board charger using a specific duty cycle PWM signal, collects static parameters of electric vehicles, charging pile operation data, and user interaction data, constructs feature vectors through outlier removal and normalization, uses the K-means algorithm to classify charging scenarios into high, medium, and low priorities, and then introduces correction factors to construct differentiated charging power and duration models for different priorities. Finally, it integrates the demand list and pushes it to the charging management platform for regular dynamic updates.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A method for obtaining AC orderly charging demand, characterized by comprising the following steps:

[0007] S1: After the charging pile is connected to the on-board charger, it initiates a digital communication request through a PWM signal with a specific duty cycle; the on-board charger needs to complete more than 5 switching state changes that meet the interval requirements within 100 milliseconds and remain disconnected in response to the communication request.

[0008] S2: After the charging pile recognizes and responds, a connection is established. Data acquisition commands are sent to the on-board charger via serial port data format to collect static parameters of the electric vehicle, and charging pile operation data and user interaction data are collected simultaneously.

[0009] S3: Perform outlier removal and normalization on the raw data, and construct a user charging behavior feature vector based on the processed data;

[0010] S4: Based on the user charging behavior feature vector, the K-means algorithm is used to dynamically classify charging scenarios. By determining the scenario adaptation coefficient, the charging scenarios are divided into three categories: high priority, medium priority, and low priority.

[0011] S5: Set differentiated targets for three different priorities, introduce charging demand correction factors, and build charging power and duration demand models respectively;

[0012] S6: Charging power requirements, charging time requirements, and scenario priority information are integrated into an orderly charging demand list, which is pushed to the charging management platform in real time and updated according to a fixed cycle. The entire process is repeated periodically to achieve dynamic adjustment of demand.

[0013] After the charging pile establishes a connection with the on-board charger in step S1, it sends a PWM signal with a 5% duty cycle on the CP line for 200 milliseconds, and then maintains a 100% duty cycle output, that is, continuously outputs a high level, indicating that digital communication is required. If the on-board charger supports this function, it must complete the on-board charger switch state switching more than 5 times within 100 milliseconds, and the state switching interval is not less than 1 millisecond. After completion, it remains in the disconnected state. If the charging pile does not detect the on-board charger switch state switching within 100 milliseconds after sending the 5% signal, it enters the standard charging process without digital communication.

[0014] After recognizing the response from the on-board charger, the charging pile sends a data acquisition command to the on-board charger via a serial port data format, consisting of 1 start bit, 8 data bits, 1 stop bit, and a baud rate of 600. The command content includes the command code, command length, command data, and check code. After the command is sent, the output remains high.

[0015] If the on-board charger fails to decode and verify the received command, it sends a failure message; otherwise, it sends the data required by the command.

[0016] After completing data transmission, the on-board charger keeps its switch off and waits to check if the information exchange has ended. If the charging pile successfully decodes the information returned by the on-board charger, it will proceed to the next stage. If it fails, it will retry the reading instruction process, with a maximum of 3 retries. After completing data reading and obtaining the charging demand information, the charging pile will switch the CP voltage more than 5 times within 100 milliseconds, with an interval of no less than 5 milliseconds between each voltage state switch. After completion, it will maintain a high level and then enter the standard charging process.

[0017] Step S2 specifically involves establishing a connection after the charging pile recognizes and responds, sending data acquisition commands to the on-board charger via serial port data format to collect static parameters of the electric vehicle, and simultaneously collecting charging pile operation data and user interaction data. The charging pile operation data includes real-time line voltage. Power factor and node load rate The static parameters of the electric vehicle include the rated capacity of the battery. Current remaining battery power Battery health and the rated current of the charging interface The user interaction data includes the user's scheduled charging start time. Expected charging end time and minimum expected remaining power .

[0018] Step S3 specifically involves outlier removal and normalization of the original data, and constructing a user charging behavior feature vector based on the processed data. The expression for the feature vector V is:

[0019] ;

[0020] In the formula, Feature vector of user charging behavior; Let be all feature weight coefficients, and their sum satisfies The specific value is determined by combining the analytic hierarchy process (AHP) with the entropy weight method. ; This refers to the current remaining battery power of the electric vehicle. This refers to the maximum capacity of the electric vehicle's battery. The desired end point of charging for the user; The scheduled charging start time for users; The standard charging time is set to 8 hours. The value represents the health status of the electric vehicle battery and ranges from [0,1]. The rated current for the electric vehicle charging interface; This refers to the maximum allowable charging current for the charging station. This represents the real-time load rate of the charging pile, with a value ranging from [0,1].

[0021] Step S4 specifically involves dynamically classifying charging scenarios using the K-means algorithm based on user charging behavior feature vectors. Specifically, it involves pre-setting three clustering center vectors for charging scenarios at different times, for different user types, and for different charging pile load states, based on historical charging data. (High-priority scenario center) (Medium-priority scenario center) (Low-priority scene center) corresponds to three target scene categories: "high", "medium", and "low". For each newly collected and constructed feature vector V, the distance between it and the three cluster centers is calculated using Euclidean distance. The smaller the distance, the higher the fit between V and the scene category.

[0022] All feature vectors are assigned to the category of the nearest cluster center, and then the mean vector of each category is recalculated as the new cluster center. The "assignment-update" process is repeated until the cluster centers no longer change, ensuring that the classification results are stable.

[0023] After convergence, each feature vector will be classified into one of the three categories: high, medium, and low, providing a basis for subsequent priority determination.

[0024] After the K-means algorithm completes the initial classification, it further quantifies the scene priority by determining the scene fitness coefficient λ, avoiding the boundary ambiguity problem caused by simple clustering. The calculation logic of λ is as follows:

[0025] ;

[0026] In the formula, This is the scene adaptation coefficient, with a value ranging from [0,1]. The number of cluster centers is a fixed value, corresponding to the three types of charging scenarios: high, medium, and low. It is uniformly set to 3 and has no unit. Euclidean distance is used to calculate the feature vector of user charging behavior. With the Scene clustering center vector The similarity between them; For the first The cluster center vectors of the scene types are obtained by K-means clustering analysis; This represents the kernel function bandwidth, a fixed value, uniformly set to 0.15. For the first Charging pile capacity coefficient for different scenarios, with high-priority scenarios corresponding to Medium priority scenarios correspond to Low-priority scenarios correspond to .

[0027] Step S5 sets differentiated targets for three different priority categories, introduces charging demand correction factors, and constructs charging power and duration demand models respectively, specifically including:

[0028] For high-priority scenarios: prioritizing user needs and ensuring the safety of charging stations, charging power requirements... and charging time requirements The formula for calculation is:

[0029] ;

[0030] ;

[0031] In the formula, To meet the real-time charging power requirements in high-priority scenarios; This refers to the real-time line voltage of the charging station. Rated current for electric vehicle charging interface; This is the real-time power factor of the charging pile, with a value ranging from [0,1]. The maximum allowable charging current for the charging station; Real-time load rate of charging piles; The load impact factor for charging piles ranges from [0.5, 1], and its calculation formula is as follows: ; To meet the charging time requirements in high-priority scenarios; The rated capacity of the electric vehicle battery; The minimum expected remaining battery level set for the user, which is the minimum battery level the user needs to reach after charging; This refers to the current remaining battery power of the electric vehicle. To meet the real-time charging power requirements in high-priority scenarios; This is the charging efficiency correction factor, with a value ranging from [0.92, 0.95]. Its calculation formula is as follows: ;

[0032] For medium-priority scenarios: with the goal of balancing charging pile load and user demand, the charging power requirement... and charging time requirements The formula for calculation is:

[0033] ;

[0034] ;

[0035] In the formula, For real-time charging power requirements in medium-priority scenarios; This refers to the real-time line voltage of the charging pile. The maximum allowable charging current for the charging station; Pi; Real-time; Schedule the start time for charging for users; The expected end point of charging for the user; This refers to the real-time power factor of the charging pile. For charging time requirements in medium-priority scenarios; For electric vehicle battery rated capacity, For users' minimum expected remaining battery power, The current remaining battery power of the electric vehicle, This is a correction factor for charging efficiency; For real-time charging power requirements in medium-priority scenarios; The elastic duration correction is calculated using the following formula: ;

[0036] For low-priority scenarios: Prioritizing optimal charging station load, charging power demand... and charging time requirements The formula for calculation is:

[0037] ;

[0038] ;

[0039] In the formula, For real-time charging power requirements in low-priority scenarios; For the real-time line voltage of the charging pile, For the maximum allowable charging current of the charging pile, For real-time load rate of charging piles, This refers to the real-time power factor of the charging pile. This represents the average daily load rate of the charging pile, with a value ranging from [0,1). For charging time requirements in low-priority scenarios; For electric vehicle battery rated capacity, For users' minimum expected remaining battery power, The current remaining battery power of the electric vehicle, For low-priority scenarios, real-time charging power requirements For charging efficiency correction factor, This is the adjustment amount for the elastic duration; Double the elastic duration for low-priority scenarios.

[0040] A computer device includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, cause the at least one processor to perform the method according to any one of claims 1 to 7.

[0041] A non-transitory computer-readable storage medium storing computer instructions that, when executed by at least one processor, cause the at least one processor to perform the method as described in any one of claims 1 to 7.

[0042] A computer program product comprising a computer program stored on a non-transitory computer-readable storage medium, the computer program including program instructions that, when executed by a computer, cause the computer to perform the method as described in any one of claims 1 to 7.

[0043] The core mechanism of this orderly charging demand acquisition method is to establish a communication link through vehicle-charging pile digital interaction, integrate multi-dimensional data of vehicles, charging piles, and people, and construct a differentiated charging demand model through intelligent analysis and scenario adaptation. Finally, it dynamically outputs an orderly charging solution to achieve a precise balance between user charging demand and charging pile load.

[0044] Its specific working logic is as follows: First, after the charging pile connects to the on-board charger, it initiates a digital communication request by sending a PWM signal with a specific duty cycle (5% duty cycle for 200 milliseconds followed by switching to 100% high level) through the CP line. The on-board charger needs to complete at least 5 switching state transitions with an interval of no less than 1 millisecond within 100 milliseconds and remain disconnected to verify communication compatibility. If no valid response is detected, it automatically enters the standard charging process to ensure communication reliability and scenario adaptability. After communication is established, the charging pile sends a data acquisition command through a serial port format with 1 start bit, 8 data bits, 1 stop bit, and a baud rate of 600, simultaneously acquiring static parameters of the electric vehicle, charging pile operating data, and user interaction data, and ensuring the accuracy of data acquisition through a 3-retry mechanism.

[0045] Subsequently, outlier removal and normalization were performed on the raw data to eliminate data noise and scale differences. Based on the feature weight coefficients determined by the analytic hierarchy process and the entropy weight method, a charging behavior feature vector integrating battery status, user demand, and charging pile load was constructed, transforming multi-dimensional data into quantifiable analytical indicators. Based on this feature vector, the K-means algorithm was used for dynamic clustering. By calculating the Euclidean distance between the feature vector and the cluster centers of the three scenarios, and combining the kernel function and the charging pile load coefficient, the scenario adaptation coefficient λ was obtained. High, medium, and low priority scenarios were then classified according to the thresholds λ≥0.75, 0.4≤λ<0.75, and λ<0.4, achieving accurate classification of charging scenarios.

[0046] For different priority scenarios, key parameters such as load impact factor, charging efficiency correction coefficient, and flexible duration correction are introduced to construct differentiated charging power and duration demand models: High-priority scenarios focus on user needs and charging pile safety, taking the minimum power under the constraints of interface rated current and charging pile network allowable current, and accurately calculating the duration to meet the minimum power demand; Medium-priority scenarios take into account both load balancing and user needs, dynamically adjusting power through a sine function, and adding flexible duration to adapt to load fluctuations; Low-priority scenarios aim for optimal charging pile network load, optimizing power allocation based on the ratio of real-time load rate to daily average load rate, and setting double the flexible duration to adapt to off-peak periods of the power grid.

[0047] Ultimately, the charging power requirements, duration requirements, and priority information of the three scenarios are integrated into an ordered charging demand list, which is pushed to the charging management platform in real time and repeated throughout the entire process at a fixed cycle. This achieves dynamic updates and closed-loop optimization of charging demands, ensuring both the user's charging experience and improving the operating efficiency of charging piles and the grid's acceptance capacity.

[0048] Compared with the prior art, the beneficial effects of the present invention are:

[0049] 1. Compared with the problems of insufficient vehicle-charging digital communication compatibility and ambiguous response mechanism in the existing technology, the present invention initiates communication requests through the PWM signal with a specific duty cycle of the CP line, clarifies the response standard of the on-board charger, and sets up a 3-retry mechanism and a standard charging process fallback scheme. This not only ensures efficient interaction between vehicles and charging piles that support digital communication functions, but also is compatible with traditional charging scenarios that do not support this function, greatly improving the adaptability and communication stability between different devices.

[0050] 2. Existing technologies often focus on single-dimensional data collection. This invention simultaneously integrates static parameters of electric vehicles, charging pile operation data, and user interaction data. It also ensures data accuracy through serial port standardized data format and verification mechanism, providing multi-dimensional and high-quality data support for charging demand modeling and avoiding demand judgment bias caused by single data dimensions.

[0051] 3. Existing technologies often employ fixed charging strategies or simple priority divisions, lacking dynamic adaptation capabilities. This invention determines feature weights through the analytic hierarchy process and entropy weight method, constructs a charging behavior feature vector that integrates battery status, user needs, and charging pile network load, and then combines the K-means algorithm with the scenario adaptation coefficient λ to achieve dynamic and accurate classification of charging scenarios. Compared with traditional classification methods, this approach better meets the charging needs of different users and the operating status of charging piles, improving the targeting of scenario adaptation.

[0052] 4. Existing technologies often adopt a "one-size-fits-all" approach to setting charging power and duration, which makes it difficult to balance user demand and grid load. This invention sets differentiated targets for three priority scenarios and introduces key parameters such as load impact factors and flexible duration correction to build a dedicated model. This not only ensures the charging experience for users with high demand, but also reduces the peak load of charging piles through dynamic power adjustment, thereby improving the grid's acceptance capacity and operating efficiency, achieving a win-win situation for users, charging piles, and the grid. Attached Figure Description

[0053] Figure 1 This is a flowchart of a method for obtaining AC orderly charging demand according to the present invention;

[0054] Figure 2 This is a waveform diagram illustrating the information interaction and negotiation process for a method of obtaining orderly charging demand according to the present invention.

[0055] Figure 3 This is a waveform diagram illustrating the information interaction of a method for obtaining orderly charging demand according to the present invention.

[0056] Figure 4 This is a waveform diagram showing the end of information interaction in the AC orderly charging demand acquisition method of the present invention.

[0057] Figure 5This invention provides a detailed waveform diagram of the information exchange and charging pile transmission method for obtaining orderly charging demand.

[0058] Figure 6 This invention provides a detailed waveform diagram of the OBC response in an AC-ordered charging demand acquisition method.

[0059] Figure 7 This is a complete waveform diagram of information interaction for the AC orderly charging demand acquisition method of the present invention; Detailed Implementation

[0060] The technical solutions of the present invention will now be described in detail with reference to the accompanying drawings.

[0061] like Figure 1-7 As shown, a method for obtaining AC orderly charging demand is characterized by including the following steps:

[0062] S1: After the charging pile is connected to the on-board charger, it initiates a digital communication request through a PWM signal with a specific duty cycle; the on-board charger needs to complete more than 5 switching state changes that meet the interval requirements within 100 milliseconds and remain disconnected in response to the communication request.

[0063] S2: After the charging pile recognizes and responds, a connection is established. Data acquisition commands are sent to the on-board charger via serial port data format to collect static parameters of the electric vehicle, and charging pile operation data and user interaction data are collected simultaneously.

[0064] S3: Perform outlier removal and normalization on the raw data, and construct a user charging behavior feature vector based on the processed data;

[0065] S4: Based on the user charging behavior feature vector, the K-means algorithm is used to dynamically classify charging scenarios. By determining the scenario adaptation coefficient, the charging scenarios are divided into three categories: high priority, medium priority, and low priority.

[0066] S5: Set differentiated targets for three different priorities, introduce charging demand correction factors, and build charging power and duration demand models respectively;

[0067] S6: Charging power requirements, charging time requirements, and scenario priority information are integrated into an orderly charging demand list, which is pushed to the charging management platform in real time and updated according to a fixed cycle. The entire process is repeated periodically to achieve dynamic adjustment of demand.

[0068] After the charging pile establishes a connection with the on-board charger in step S1, it sends a PWM signal with a 5% duty cycle on the CP line for 200 milliseconds, and then maintains a 100% duty cycle output, that is, continuously outputs a high level, indicating that digital communication is required. If the on-board charger supports this function, it must complete the on-board charger switch state switching more than 5 times within 100 milliseconds, and the state switching interval is not less than 1 millisecond. After completion, it remains in the disconnected state. If the charging pile does not detect the on-board charger switch state switching within 100 milliseconds after sending the 5% signal, it enters the standard charging process without digital communication.

[0069] After recognizing the response from the on-board charger, the charging pile sends a data acquisition command to the on-board charger via a serial port data format, consisting of 1 start bit, 8 data bits, 1 stop bit, and a baud rate of 600. The command content includes the command code, command length, command data, and check code. After the command is sent, the output remains high.

[0070] If the on-board charger fails to decode and verify the received command, it sends a failure message; otherwise, it sends the data required by the command.

[0071] After completing data transmission, the on-board charger keeps its switch off and waits to check if the information exchange has ended. If the charging pile successfully decodes the information returned by the on-board charger, it will proceed to the next stage. If it fails, it will retry the reading instruction process, with a maximum of 3 retries. After completing data reading and obtaining the charging demand information, the charging pile will switch the CP voltage more than 5 times within 100 milliseconds, with an interval of no less than 5 milliseconds between each voltage state switch. After completion, it will maintain a high level and then enter the standard charging process.

[0072] Step S2 specifically involves establishing a connection after the charging pile recognizes and responds, sending data acquisition commands to the on-board charger via serial port data format to collect static parameters of the electric vehicle, and simultaneously collecting charging pile operation data and user interaction data. The charging pile operation data includes real-time line voltage. Power factor and node load rate The static parameters of the electric vehicle include the rated capacity of the battery. Current remaining battery power Battery health and the rated current of the charging interface The user interaction data includes the user's scheduled charging start time. Expected charging end time and minimum expected remaining power .

[0073] Step S3 specifically involves outlier removal and normalization of the original data, and constructing a user charging behavior feature vector based on the processed data. The expression for the feature vector V is:

[0074] ;

[0075] In the formula, Feature vector of user charging behavior; Let be all feature weight coefficients, and their sum satisfies The specific value is determined by combining the analytic hierarchy process (AHP) with the entropy weight method. ; This refers to the current remaining battery power of the electric vehicle. This refers to the maximum capacity of the electric vehicle's battery. The desired end point of charging for the user; The scheduled charging start time for users; The standard charging time is set to 8 hours. The value represents the health status of the electric vehicle battery and ranges from [0,1]. The rated current for the electric vehicle charging interface; This refers to the maximum allowable charging current for the charging station. This represents the real-time load rate of the charging pile, with a value ranging from [0,1].

[0076] Step S4 specifically involves using the K-means algorithm to dynamically classify charging scenarios based on the user's charging behavior feature vector, determining the scenario adaptation coefficient λ. Scenarios are classified as high-priority scenarios when λ ≥ 0.75, medium-priority scenarios when 0.4 ≤ λ < 0.75, and low-priority scenarios when λ < 0.4. The calculation formula for the scenario adaptation coefficient λ is:

[0077] ;

[0078] In the formula, This is the scene adaptation coefficient, with a value ranging from [0,1]. The number of cluster centers is a fixed value, corresponding to the three types of charging scenarios: high, medium, and low. It is uniformly set to 3 and has no unit. Euclidean distance is used to calculate the feature vector of user charging behavior. With the Scene clustering center vector The similarity between them; For the first The cluster center vectors of the scene types are obtained by K-means clustering analysis; This represents the kernel function bandwidth, a fixed value, uniformly set to 0.15. For the first Charging pile capacity coefficient for different scenarios, with high-priority scenarios corresponding to Medium priority scenarios correspond to Low-priority scenarios correspond to .

[0079] Step S5 sets differentiated targets for three different priority categories, introduces charging demand correction factors, and constructs charging power and duration demand models respectively, specifically including:

[0080] For high-priority scenarios: prioritizing user needs and ensuring the safety of charging stations, charging power requirements... and charging time requirements The formula for calculation is:

[0081] ;

[0082] ;

[0083] In the formula, To meet the real-time charging power requirements in high-priority scenarios; This refers to the real-time line voltage of the charging station. Rated current for electric vehicle charging interface; This is the real-time power factor of the charging pile, with a value ranging from [0,1]. The maximum allowable charging current for the charging station; Real-time load rate of charging piles; The load impact factor for charging piles ranges from [0.5, 1], and its calculation formula is as follows: ; To meet the charging time requirements in high-priority scenarios; The rated capacity of the electric vehicle battery; The minimum expected remaining battery level set for the user, which is the minimum battery level the user needs to reach after charging; This refers to the current remaining battery power of the electric vehicle. To meet the real-time charging power requirements in high-priority scenarios; This is the charging efficiency correction factor, with a value ranging from [0.92, 0.95]. Its calculation formula is as follows: ;

[0084] For medium-priority scenarios: with the goal of balancing charging pile load and user demand, the charging power requirement... and charging time requirements The formula for calculation is:

[0085] ;

[0086] ;

[0087] In the formula, For real-time charging power requirements in medium-priority scenarios; This refers to the real-time line voltage of the charging pile. The maximum allowable charging current for the charging station; Pi; Real-time; Schedule the start time for charging for users; The expected end point of charging for the user; This refers to the real-time power factor of the charging pile. For charging time requirements in medium-priority scenarios; For electric vehicle battery rated capacity, For users' minimum expected remaining battery power, The current remaining battery power of the electric vehicle, This is a correction factor for charging efficiency; For real-time charging power requirements in medium-priority scenarios; The elastic duration correction is calculated using the following formula: ;

[0088] For low-priority scenarios: Prioritizing optimal charging station load, charging power demand... and charging time requirements The formula for calculation is:

[0089] ;

[0090] ;

[0091] In the formula, For real-time charging power requirements in low-priority scenarios; For the real-time line voltage of the charging pile, For the maximum allowable charging current of the charging pile, For real-time load rate of charging piles, This refers to the real-time power factor of the charging pile. This represents the average daily load rate of the charging pile, with a value ranging from [0,1). For charging time requirements in low-priority scenarios; For electric vehicle battery rated capacity, For users' minimum expected remaining battery power, The current remaining battery power of the electric vehicle, For low-priority scenarios, real-time charging power requirements For charging efficiency correction factor, This is the adjustment amount for the elastic duration; Double the elastic duration for low-priority scenarios.

[0092] In practical applications of community charging stations, after the charging station connects to the on-board charger of the electric vehicle, it first sends a PWM signal with a 5% duty cycle on the CP line for 200 milliseconds, and then maintains a high level to initiate a digital communication request. If the on-board charger supports this function, it needs to complete at least 5 switching state transitions within 100 milliseconds, with each transition having an interval of no less than 1 millisecond. After switching, it remains disconnected. If the charging station does not detect a valid response within 100 milliseconds, it automatically enters the standard charging process. After successful communication verification, the charging station sends a collection command containing the instruction code, instruction length, instruction data, and checksum in a serial port format with 1 start bit, 8 data bits, 1 stop bit, and a baud rate of 600. The on-board charger returns data after successful verification and sends a failure message if it fails. The charging station can retry up to 3 times. After data acquisition is completed, the charging station will switch the CP voltage more than 5 times within 100 milliseconds, with each transition having an interval of no less than 5 milliseconds, and then maintain a high level to enter the subsequent process. During this process, the system simultaneously collects data on the electric vehicle's battery rated capacity, current remaining charge, battery health, and charging interface rated current; the charging pile's real-time line voltage, power factor, and real-time load rate; and the user's scheduled charging start time, expected end time, and minimum expected remaining charge. After outlier removal and normalization of these raw data, a user charging behavior feature vector is constructed with weights of 0.32, 0.28, 0.20, 0.10, and 0.10, respectively. The standard charging duration is set to 8 hours. Based on this feature vector, a K-means algorithm is used for dynamic classification. The number of cluster centers is fixed at 3, the kernel function bandwidth is 0.15, the charging pile load factor for high-priority scenarios is 0.9, for medium-priority scenarios it is 0.6, and for low-priority scenarios it is 0.3. Scenario adaptation coefficients of 0.75 or higher are considered high-priority, between 0.4 and 0.75 are medium-priority, and below 0.4 are low-priority. For high-priority scenarios, with the core focus on meeting user needs and ensuring charging pile safety, the charging efficiency correction coefficient is between 0.92 and 0.95, the charging pile load impact factor is between 0.5 and 1, and the charging power is the smaller value between the power corresponding to the rated current of the charging interface and the power under the constraint of the charging pile network's allowable current. The charging time is calculated based on the battery's rated capacity, the difference between the current remaining power and the minimum expected remaining power, combined with the determined charging power and the charging efficiency correction coefficient. For medium-priority scenarios, the charging pile load balance and user needs are taken into account. The charging power is calculated based on the real-time line voltage, the maximum allowable current of the charging pile network, the power factor, and the coefficient that dynamically changes with the charging period. The flexible time correction amount is determined based on the length of the scheduled charging period and the scenario adaptation coefficient. The charging time is increased by this flexible time correction amount on the basic calculated time. For low-priority scenarios, the goal is to optimize the charging pile network load. The charging power is calculated based on the real-time line voltage, the maximum allowable current of the charging pile network, the real-time load rate, the daily average load rate, and the power factor. The charging time is increased by double the flexible time correction amount on the basic calculated time.Finally, the charging power, charging time, and priority information of the three scenarios are integrated into an orderly charging demand list, which is pushed to the community charging management platform in real time. The entire process from communication establishment, data collection, processing and classification to model building is repeated at fixed intervals to continuously and dynamically adjust the charging demand plan to adapt to the charging needs of different users and changes in grid load.

Claims

1. A method for obtaining AC orderly charging demand, characterized in that, Includes the following steps: S1: After the charging pile is connected to the on-board charger, it initiates a digital communication request through a PWM signal with a specific duty cycle; the on-board charger needs to complete more than 5 switching state changes that meet the interval requirements within 100 milliseconds and remain disconnected in response to the communication request. S2: After the charging pile recognizes and responds, a connection is established. Data acquisition commands are sent to the on-board charger via serial port data format to collect static parameters of the electric vehicle, and charging pile operation data and user interaction data are collected simultaneously. S3: Perform outlier removal and normalization on the original data, and construct a user charging behavior feature vector based on the processed data; S4: Based on the user charging behavior feature vector, the K-means algorithm is used to dynamically classify charging scenarios. By determining the scenario adaptation coefficient, the charging scenarios are divided into three categories: high priority, medium priority, and low priority. S5: Set differentiated targets for three different priorities, introduce charging demand correction factors, and build charging power and duration demand models respectively; S6: Charging power requirements, charging time requirements, and scenario priority information are integrated into an orderly charging demand list, which is pushed to the charging management platform in real time and updated according to a fixed cycle. The entire process is repeated periodically to achieve dynamic adjustment of demand.

2. The method for obtaining AC orderly charging demand according to claim 1, characterized in that, After the charging pile establishes a connection with the on-board charger in step S1, it sends a PWM signal with a 5% duty cycle on the CP line for 200 milliseconds, and then maintains a 100% duty cycle output, that is, continuously outputs a high level, indicating that digital communication is required. If the on-board charger supports this function, it must complete the on-board charger switch state switching more than 5 times within 100 milliseconds, and the state switching interval is not less than 1 millisecond. After completion, it remains in the disconnected state. If the charging pile does not detect the on-board charger switch state switching within 100 milliseconds after sending the 5% signal, it enters the standard charging process without digital communication. After recognizing the response from the on-board charger, the charging pile sends a data acquisition command to the on-board charger via a serial port data format, consisting of 1 start bit, 8 data bits, 1 stop bit, and a baud rate of 600. The command content includes the command code, command length, command data, and check code. After the command is sent, the output remains high. If the on-board charger fails to decode and verify the received command, it will send a failure message; otherwise, it will send the data required by the command.

3. The method for obtaining AC orderly charging demand according to claim 2, characterized in that, After completing the data transmission, the on-board charger keeps its switch off and waits to see if the information exchange has ended. If the charging pile successfully decodes the information returned by the on-board charger, it will proceed to the next stage. If it fails, it will retry the reading instruction process. The maximum number of retries is 3. After completing data reading and obtaining charging demand information, the charging pile switches the CP voltage more than 5 times within 100 milliseconds, with an interval of no less than 5 milliseconds between each voltage state switch. After completion, it maintains a high level and then enters the standard charging process.

4. The method for obtaining AC orderly charging demand according to claim 1, characterized in that, Step S2 specifically involves establishing a connection after the charging pile recognizes and responds, sending data acquisition commands to the on-board charger via serial port data format to collect static parameters of the electric vehicle, and simultaneously collecting charging pile operation data and user interaction data. The charging pile operation data includes real-time line voltage. Power factor and node load rate The static parameters of the electric vehicle include the rated capacity of the battery. Current remaining battery power Battery health and the rated current of the charging interface ; The user interaction data includes the user's scheduled charging start time. Expected charging end time and minimum expected remaining power .

5. The method for obtaining AC orderly charging demand according to claim 1, characterized in that, Step S3 specifically involves outlier removal and normalization of the original data, and constructing a user charging behavior feature vector based on the processed data. The expression for the feature vector V is: ; In the formula, Feature vector of user charging behavior; Let be all feature weight coefficients, and their sum satisfies The specific value is determined by combining the analytic hierarchy process (AHP) with the entropy weight method. ; This refers to the current remaining battery power of the electric vehicle. This refers to the maximum capacity of the electric vehicle's battery. The desired end point of charging for the user; The scheduled charging start time for the user; The standard charging time is set to 8 hours. The value represents the health status of the electric vehicle battery and ranges from [0,1]. The rated current for the electric vehicle charging interface; This refers to the maximum allowable charging current for the charging station. This represents the real-time load rate of the charging pile, with a value ranging from [0,1].

6. The method for obtaining AC orderly charging demand according to claim 1, characterized in that, Step S4 specifically involves using the K-means algorithm to dynamically classify charging scenarios based on the user's charging behavior feature vector, determining the scenario adaptation coefficient λ. Scenarios are classified as high-priority scenarios when λ ≥ 0.75, medium-priority scenarios when 0.4 ≤ λ < 0.75, and low-priority scenarios when λ < 0.

4. The calculation formula for the scenario adaptation coefficient λ is: ； In the formula, This is the scene adaptation coefficient, with a value ranging from [0,1]. The number of cluster centers is a fixed value, corresponding to the three types of charging scenarios: high, medium, and low. It is uniformly set to 3 and has no unit. Euclidean distance is used to calculate the feature vector of user charging behavior. With the Scene-based cluster center vector The similarity between them; For the first The cluster center vectors of the scene types are obtained by K-means clustering analysis; This represents the kernel function bandwidth, a fixed value, uniformly set to 0.

15. For the first Charging pile capacity coefficient for different scenarios, with high-priority scenarios corresponding to Medium priority scenarios correspond to Low-priority scenarios correspond to .

7. The method for obtaining AC orderly charging demand according to claim 1, characterized in that, Step S5 sets differentiated targets for three different priority categories, introduces charging demand correction factors, and constructs charging power and duration demand models respectively, specifically including: For high-priority scenarios: prioritizing user needs and ensuring the safety of charging stations, charging power requirements... and charging time requirements The formula for calculation is: ; ; In the formula, To meet the real-time charging power requirements in high-priority scenarios; This refers to the real-time line voltage of the charging station. Rated current for electric vehicle charging interface; This is the real-time power factor of the charging pile, with a value ranging from [0,1]. The maximum allowable charging current for the charging station; Real-time load rate of charging piles; The load impact factor for charging piles ranges from [0.5, 1], and its calculation formula is as follows: ; To meet the charging time requirements in high-priority scenarios; The rated capacity of the electric vehicle battery; The minimum expected remaining battery level set for the user, which is the minimum battery level the user needs to reach after charging; This refers to the current remaining battery power of the electric vehicle. To meet the real-time charging power requirements in high-priority scenarios; This is the charging efficiency correction factor, with a value ranging from [0.92, 0.95]. Its calculation formula is as follows: ; For medium-priority scenarios: with the goal of balancing charging pile load and user demand, the charging power requirement... and charging time requirements The formula for calculation is: ; ; In the formula, For real-time charging power requirements in medium-priority scenarios; This refers to the real-time line voltage of the charging pile. The maximum allowable charging current for the charging station; Pi; Real-time; Schedule the start time for charging for users; The expected end point of charging for the user; This refers to the real-time power factor of the charging pile. For charging time requirements in medium-priority scenarios; For electric vehicle battery rated capacity, For users' minimum expected remaining battery power, The current remaining battery power of the electric vehicle, This is a correction factor for charging efficiency; For real-time charging power requirements in medium-priority scenarios; The elastic duration correction is calculated using the following formula: ; For low-priority scenarios: The goal is to optimize the charging station load, and the charging power requirement is... and charging time requirements The formula for calculation is: ; ; In the formula, For real-time charging power requirements in low-priority scenarios; For the real-time line voltage of the charging pile, For the maximum allowable charging current of the charging pile, For real-time load rate of charging piles, This refers to the real-time power factor of the charging pile. This represents the average daily load rate of the charging pile, with a value ranging from [0,1). For charging time requirements in low-priority scenarios; For electric vehicle battery rated capacity, For users' minimum expected remaining battery power, The current remaining battery power of the electric vehicle, For low-priority scenarios, real-time charging power requirements For charging efficiency correction factor, This is the adjustment amount for the elastic duration; Double the elastic duration for low-priority scenarios.

8. A computer device, the device comprising: At least one processor; And a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, cause the at least one processor to perform the method of any one of claims 1 to 7.

9. A non-transitory computer-readable storage medium storing computer instructions that, when executed by at least one processor, cause the at least one processor to perform the method as described in any one of claims 1 to 7.

10. A computer program product comprising a computer program stored on a non-transitory computer-readable storage medium, the computer program including program instructions that, when executed by a computer, cause the computer to perform the method as described in any one of claims 1 to 7.

Images