A charging scheduling system and method based on Actor-Critic analysis of automobile electricity prices and carbon emissions

Abstract

The present invention provides an Actor-Critic-based charging scheduling system and method for automobile electricity prices and carbon emissions, which is applied to the field of data processing technology. The application processes the dynamic electricity price information and carbon emission signal information of the power grid based on the privacy protection parameter information of federated learning to generate benefit evaluation parameter information; processes the user charging demand information based on the historical charging and power grid load problem information to generate charging strategy adjustment reference information; processes the charging strategy adjustment reference information based on the privacy protection parameter information of federated learning and the power grid information of different regions to generate a cross-regional collaborative optimization factor; processes the user charging demand information and the regional information corresponding to the user charging demand information based on the cross-regional collaborative optimization factor to generate charging demand classification information and priority information; and processes the scheduling model based on the target Actor-Critic algorithm to generate charging scheduling allocation result information.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to an Actor-Critic based vehicle electricity price and carbon emission charging scheduling system and method. Background Art

[0002] With the global emphasis on environmental protection and sustainable energy, new energy vehicles (NEVs), a key alternative to traditional fuel-powered vehicles, have seen explosive growth in their ownership. However, the large-scale charging of NEVs has placed enormous pressure on the power grid. Traditional power grids were not designed to fully account for such large-scale EV charging demand. Consequently, during peak hours, when large numbers of NEVs are charging simultaneously, grid overloads are a high risk, severely impacting grid stability and reliability, and potentially even causing power outages.

[0003] At the same time, energy costs and carbon emissions are becoming increasingly prominent. Charging costs are significantly affected by the dynamic electricity prices of the power grid. These prices vary significantly across time periods and regions, making it difficult for users to balance charging costs with electricity needs. Furthermore, the carbon emissions generated by the electricity consumed by charging new energy vehicles cannot be ignored. In the context of achieving carbon reduction targets, effectively reducing carbon emissions in charging scheduling has become a pressing issue.

[0004] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present disclosure, and therefore includes information that does not constitute prior art known to ordinary technicians in this field. Summary of the Invention

[0005] The purpose of this application is to provide an Actor-Critic based charging scheduling system and method for automobile electricity prices and carbon emissions, which at least to a certain extent overcomes the problems existing in the existing technology. By comprehensively considering the dynamic electricity price information of the power grid, carbon emission signal information, user charging demand information, historical charging and power grid load problem information, power grid information in different regions, and federated learning privacy protection parameter information, the charging scheduling system and method are of great significance for optimizing the charging management of new energy vehicles, improving the operation efficiency of the power grid, and reducing energy costs and carbon emissions.

[0006] Other features and advantages of the present application will become apparent from the following detailed description, or may be learned in part by practice of the invention.

[0007] According to one aspect of the present application, a method for charging scheduling of automobile electricity prices and carbon emissions based on Actor-Critic is provided, comprising: obtaining dynamic grid electricity price information, carbon emission signal information, user charging demand information, historical charging and grid load problem information, grid information in different regions, and federated learning privacy protection parameter information; processing the dynamic grid electricity price information and carbon emission signal information based on the federated learning privacy protection parameter information to generate benefit evaluation parameter information; processing the user charging demand information based on historical charging and grid load problem information to generate charging strategy adjustment reference information; processing the charging strategy adjustment reference information based on the federated learning privacy protection parameter information and grid information in different regions to generate a cross-regional collaborative optimization factor; processing the user charging demand information and regional information corresponding to the user charging demand information based on the cross-regional collaborative optimization factor to generate charging demand classification information and priority information, wherein the user charging demand information is used to characterize the user's vehicle type, charging time preference, battery power information, and grid access point information; and inputting the benefit evaluation parameter information, charging demand classification information and priority information, and historical charging and grid load problem information into a scheduling model based on a target Actor-Critic algorithm for processing to generate charging scheduling allocation result information.

[0008] Another aspect of the present application is a charging scheduling device for automobile electricity prices and carbon emissions based on Actor-Critic, characterized in that it includes: an acquisition module for acquiring dynamic power grid electricity price information, carbon emission signal information, user charging demand information, historical charging and power grid load problem information, power grid information in different regions, and federated learning privacy protection parameter information; a processing module for processing the dynamic power grid electricity price information and carbon emission signal information based on the federated learning privacy protection parameter information to generate benefit evaluation parameter information; processing the user charging demand information based on the historical charging and power grid load problem information to generate charging strategy adjustment reference information; based on the federated learning privacy protection parameter information, a charging scheduling device for automobile electricity prices and carbon emission emissions based on Actor-Critic, characterized in that it includes: an acquisition module for acquiring dynamic power grid electricity price information, carbon emission signal information, user charging demand information, historical charging and power grid load problem information, power grid information in different regions, and federated learning privacy protection parameter information; a processing module for processing the dynamic power grid electricity price information and carbon emission signal information based on the federated learning privacy protection parameter information to generate benefit evaluation parameter information; based on the historical charging and power grid load problem information, a charging strategy adjustment reference information is generated; based on the federated learning privacy protection parameter information, a charging scheduling device for automobile electricity prices and carbon emission emissions based on Actor-Critic, characterized in that it includes: an acquisition module for acquiring dynamic power grid electricity price information, carbon emission signal information, user charging demand information, historical charging and power grid load problem information, power grid information in different regions, and federated learning privacy protection parameter information; a processing module for processing the dynamic power grid electricity price information and carbon emission signal information based on the federated learning privacy protection parameter information to generate benefit evaluation parameter information; based on the federated learning privacy protection parameter information, ... The protection parameter information and the information of the power grid in different regions are used to process the reference information for charging strategy adjustment to generate a cross-regional collaborative optimization factor. Based on the cross-regional collaborative optimization factor, the user charging demand information and the regional information corresponding to the user charging demand information are processed to generate charging demand classification information and priority information, wherein the user charging demand information is used to characterize the user's vehicle type, charging time preference, battery power information and grid access point information. The benefit evaluation parameter information, charging demand classification information and priority information, as well as historical charging and grid load problem information, are input into the scheduling model based on the target Actor-Critic algorithm for processing to generate charging scheduling allocation result information.

[0009] According to another aspect of the present application, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a second processor, the above-mentioned Actor-Critic-based automobile electricity price and carbon emission charging scheduling method is implemented.

[0010] This application provides an Actor-Critic-based charging scheduling system and method for vehicle electricity prices and carbon emissions. A server acquires information such as dynamic grid electricity prices, carbon emission signals, user charging demands, historical charging and grid load issues, regional grids, and federated learning privacy-preserving parameters. Next, the server extracts and quantitatively analyzes the dynamic grid electricity prices and carbon emission signals, and combines them with privacy-preserving parameters to generate benefit evaluation parameter information. Simultaneously, historical charging and grid load issue information and user charging demand information are processed to generate reference information for charging strategy adjustment. This charging strategy adjustment reference information is then processed based on the federated learning privacy-preserving parameters and regional grid information to generate cross-regional collaborative optimization factors, which are used to generate charging demand classification and priority information. Finally, the benefit evaluation parameters, charging demand classification and priority information, and historical charging and grid load issue information are input into a target model. After analysis, strategy vector optimization, and deviation vector analytical conversion, the system generates charging scheduling and allocation results, encompassing allocated grid resources and target quantitative indicators. This system achieves charging scheduling optimization that comprehensively considers cost, efficiency, and grid stability, improving the charging management of new energy vehicles.

[0011] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 A flowchart illustrating a method for charging scheduling based on automobile electricity prices and carbon emissions based on Actor-Critic provided in one embodiment of the present application is shown;

[0013] Figure 2 A schematic diagram of the structure of a charging scheduling device for automobile electricity prices and carbon emissions based on Actor-Critic is shown in an embodiment of the present application. DETAILED DESCRIPTION

[0014] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0015] The following combination Figure 1 To describe the charging scheduling method of automobile electricity price and carbon emission based on Actor-Critic according to the exemplary embodiment of the present application. Figure 1 As shown, the method is applied to the server and includes:

[0016] S101, obtain dynamic grid electricity price information, carbon emission signal information, user charging demand information, historical charging and grid load problem information, grid information in different regions, and federated learning privacy protection parameter information.

[0017] In one implementation, taking a regional power grid and its new energy vehicle users as an example, the regional power grid operator adjusts electricity prices in real time based on factors such as power supply and demand, generation costs, and other factors. The power grid data monitoring system captures information such as the weekday peak electricity price of 0.8 yuan / kWh during the daytime (8:00 AM - 6:00 PM), 0.3 yuan / kWh during the nighttime off-peak (11:00 PM - 7:00 AM), and 0.7 yuan / kWh during the weekend peak (10:00 AM - 4:00 PM). This time-varying price data represents the grid's dynamic electricity price information. Environmental protection authorities collaborate with energy monitoring agencies to monitor carbon emissions from electricity production in the region. The data shows that the carbon emission intensity of thermal power generation is 0.8 kg / kWh, while that of hydropower generation is almost zero, and that of wind power generation is 0.05 kg / kWh. At the same time, based on regional environmental protection policies and carbon emission targets, information such as the total amount of carbon emissions allowed in different time periods and the real-time carbon emission remaining amount can also be obtained, which constitutes carbon emission signal information.

[0018] New energy vehicle users upload their charging requirements to the server via a mobile charging app or in-vehicle system. For example, user A's vehicle is a pure electric sedan with a 60 kWh battery capacity and a current battery charge of 30%. They plan to start charging after get off work (6:30 PM) and hope to fully charge within two hours. Their primary charging location is a private charging station at home in the city. User B's vehicle is an electric SUV with an 80 kWh battery capacity and a remaining charge of 20%. They plan to charge before their weekend road trip (9:00 AM on Saturday) and want to charge to 80%. Their preferred charging location is a public charging station at a highway service area. This user's vehicle type, charging time preference, battery charge level, and grid access point information are aggregated to form user charging demand information. This information is extracted from the grid operation historical database. Over the past month, the large number of new energy vehicles charging in the region during weekday evenings (7:00 PM to 10:00 PM) caused overload in some areas of the grid, leading to transformer overheating alarms. Furthermore, during certain periods, irregular charging habits among users caused significant fluctuations in charging power, impacting grid stability. Information such as power fluctuations, grid load overload, and abnormal charging time that occurred during these historical charging processes constitutes historical charging and grid load problem information.

[0019] Furthermore, the region is comprised of multiple subgrids, including the city center grid, suburban grid, and industrial park grid. The city center grid primarily relies on a mix of thermal and hydropower, resulting in a heavy load and stable demand. The suburban grid, with a higher proportion of wind power, has a relatively light load, but can also experience power shortages during peak nighttime residential electricity demand. The industrial park grid primarily ensures power for industrial production, requiring high power stability. Different regional grids differ in terms of power source composition, load characteristics, and power supply capacity. This information is the regional grid information.

[0020] In multi-institutional collaborations, federated learning technology is employed to protect the data privacy of all parties. For example, power grid operators, new energy vehicle manufacturers, and energy research institutions jointly participate in data collaborations. The privacy protection parameters set by each party for federated learning include encryption algorithm type (e.g., homomorphic encryption to ensure data is encrypted during computations), data sharing scope (only aggregated and anonymized statistical information is shared, excluding sensitive data from specific users or grid nodes), and participant permission management (e.g., grid operators are limited to viewing data related to grid load, while vehicle manufacturers are limited to accessing aggregated data related to vehicle charging behavior). These parameters ensure the privacy and security of all parties involved during data sharing and collaborative processing.

[0021] S102: Processing the dynamic electricity price information and carbon emission signal information of the power grid based on the federated learning privacy protection parameter information to generate benefit evaluation parameter information.

[0022] In one embodiment, dynamic grid electricity price information is subjected to feature extraction and processing to generate characteristics for electricity price fluctuation amplitude, peak and valley period characteristics, and regional price difference characteristics. During a specific week, dynamic grid electricity price monitoring data for the region showed that on Monday, the peak daytime electricity price reached a high of 0.9 yuan / kWh, while the low nighttime electricity price was as low as 0.25 yuan / kWh. Calculation revealed that the price fluctuation amplitude for that day was (0.9-0.25) ÷ 0.25 = 2.6, indicating a significant daily price fluctuation. The price fluctuation amplitude varied from day to day within a week. Statistical analysis of the weekly data yielded a characteristic value for the average price fluctuation amplitude for that week, reflecting the severity of fluctuations in the dynamic grid electricity price for the region.

[0023] Long-term monitoring has revealed that weekday electricity price peaks in the region typically occur between 8:00 AM and 10:00 AM and 7:00 PM and 9:00 PM, when electricity demand is high and prices are higher. The off-peak period is from 11:00 PM to 6:00 AM the following day, when demand is low and prices are lower. Weekend peaks occur between 10:00 AM and 4:00 PM, while the off-peak period is from 10:00 PM to 8:00 AM. These fixed peak and off-peak periods create the peak-off-peak price characteristics of the region, providing valuable insights for new energy vehicle users to optimize charging schedules. The region is divided into three main areas: A, B, and C. Area A, the central business district, has higher electricity supply costs, and its peak-off-peak electricity prices are generally 0.1-0.2 yuan / kWh higher than other areas. Area B is a residential area with relatively stable electricity prices. Area C, an industrial park, has high industrial electricity consumption and special electricity policies, resulting in even greater price differences between peak and off-peak periods, with off-peak prices even lower than in Area B by 0.15 yuan / kWh. These price differences between different areas constitute the regional variation in electricity prices.

[0024] Feature extraction and processing of carbon emission signal information generates carbon emission intensity variation characteristics, carbon emission peak period characteristics, and carbon emission trend characteristics for different regions. The main sources of electricity in this region are thermal power, hydropower, and wind power. Over the past year, with improvements in environmental protection technology, the carbon emission intensity of thermal power has decreased from 0.85 kg / kWh to 0.8 kg / kWh, showing a gradually decreasing trend. The carbon emission intensity of hydropower has remained at an extremely low level, almost negligible. The carbon emission intensity of wind power is 0.05 kg / kWh, which is relatively stable. By monitoring the changes in the carbon emission intensity of different power sources, the overall carbon emission intensity variation characteristics of the region are derived, reflecting the changes in carbon emission intensity over time during the power production process.

[0025] Analysis of regional carbon emissions data reveals that carbon emissions typically peak between December and February each year, due to increased thermal power generation during the winter heating season. Peak carbon emissions often occur during peak electricity consumption hours, such as between 7:00 PM and 9:00 PM on weekdays. These fixed time periods characterize peak carbon emissions and help us understand the temporal distribution of carbon emissions.

[0026] Region A, with its dense commercial activity and high electricity consumption, primarily relies on thermal power, resulting in relatively high but stable carbon emissions. Region B, with residential electricity consumption and a gradually increasing proportion of hydropower and wind power, has seen its carbon emissions slowly decline. Region C, despite high industrial electricity consumption, is also experiencing a gradual decline in carbon emissions as the green industrial transition and the increased use of clean energy are driving its carbon emissions. These differences in carbon emission trends across regions contribute to the distinct regional carbon emission trend characteristics.

[0027] The characteristics of electricity price fluctuation amplitude, peak and valley period characteristics, and regional differences in electricity prices were quantitatively analyzed and processed to generate quantitative characteristic values ​​for electricity prices. The daily electricity price fluctuation amplitudes within a week were weighted averaged, with different weights assigned to weekday and weekend electricity consumption patterns. The final quantitative characteristic value for the weekly electricity price fluctuation amplitude was 2.3. For the peak and valley period characteristics, a comprehensive quantitative characteristic value for peak and valley electricity prices was calculated based on the proportion of electricity consumption and price differences during peak and valley periods (taking into account factors such as the duration of peak and valley periods, price differences, and the proportion of electricity consumption in each period). For the regional differences in electricity prices, by comparing the degree of price differences between different regions, a quantitative characteristic value for regional differences was determined to be 0.18 (reflecting the relative size of price differences between regions). These quantitative characteristic values ​​were integrated to form a complete set of quantitative characteristic values ​​for electricity prices.

[0028] We quantitatively analyzed and processed the characteristics of carbon emission intensity changes, the time periods during which carbon emission peaks occur, and the carbon emission trend characteristics of different regions to generate quantitative carbon emission characteristic values. Regarding the characteristics of carbon emission intensity changes, we calculated the average rate of decline in thermal power carbon emission intensity over the past year, resulting in a quantitative characteristic value of -0.05 (a negative sign indicates a decline). Regarding the time periods during which carbon emission peaks occur, we calculated a quantitative characteristic value of 0.6, reflecting the importance of carbon emissions during peak periods, based on the proportion of carbon emissions during the winter heating period and peak electricity consumption periods within a day. We quantified the characteristics of carbon emission trends in different regions by taking a weighted average of the carbon emission change rates for each region, resulting in a quantitative characteristic value of -0.03 for the overall regional carbon emission trend. These values ​​constitute the set of quantitative carbon emission characteristic values.

[0029] Based on the privacy-preserving parameters of federated learning, the quantitative characteristics of electricity prices and carbon emissions are processed to generate comprehensive benefit assessment features and corresponding weight calculation results. These comprehensive benefit assessment features and corresponding weight calculation results are processed to generate benefit assessment parameters, which characterize the impact of dynamic grid electricity prices and carbon emission signals on the benefits of new energy vehicle charging. Within the federated learning framework, power grid operators, new energy vehicle manufacturers, and energy research institutions jointly participate in data collaboration. Based on the set privacy-preserving parameters of federated learning, such as using homomorphic encryption algorithms to ensure data is encrypted during computation and only sharing aggregated and anonymized statistical information, each participant leverages their own data advantages to collaboratively analyze the quantitative characteristics of electricity prices and carbon emissions. Power grid operators provide grid cost and power supply stability data, vehicle manufacturers provide vehicle energy consumption and user charging habits data, and energy research institutions provide data assessing the environmental impact of carbon emissions. Through the fusion analysis of these data, comprehensive benefit assessment features are generated. For example, we found that electricity price fluctuations are closely related to user charging costs, making this an important feature in comprehensive benefit assessment. Changes in carbon emission intensity have a significant impact on environmental and energy policies, so they are also included in comprehensive benefit assessment. Furthermore, we calculated weights for each feature based on its impact on new energy vehicle charging benefits. For example, the weight for electricity price fluctuations is 0.4, and the weight for carbon emission intensity changes is 0.3, etc., resulting in weighted calculation results.

[0030] The comprehensive benefit assessment characteristics and their corresponding weight calculation results are further processed. Each comprehensive benefit assessment characteristic is multiplied by its corresponding weight and then accumulated to obtain a comprehensive benefit assessment parameter value. Assume that the calculated benefit assessment parameter value is 0.7 (this value comprehensively reflects the impact of dynamic grid electricity prices and carbon emission signals on the charging benefits of new energy vehicles). This benefit assessment parameter information can help new energy vehicle users understand the charging benefits under different electricity prices and carbon emission conditions and also provide an important basis for the charging scheduling system to formulate reasonable charging strategies. If the benefit assessment parameter value is high, it means that the current electricity price and carbon emission conditions are more favorable for charging benefits; otherwise, the charging strategy needs to be adjusted to improve benefits.

[0031] S103: Processing user charging demand information based on historical charging and grid load problem information to generate charging strategy adjustment reference information.

[0032] In one embodiment, historical charging and grid load problem information is extracted and classified to generate historical charging power fluctuation information, grid load overload information, and abnormal charging duration information. By analyzing the historical operating data of the regional power grid, historical charging and grid load problem information from the past month is extracted. For example, significant fluctuations in charging power were observed during certain periods. On a certain weekday evening between 8:00 PM and 9:00 PM, due to the concentrated charging of a large number of new energy vehicles, the charging power in a certain area rapidly increased from 500 kilowatts to 1200 kilowatts within an hour, then dropped to 800 kilowatts within half an hour, resulting in significant power fluctuations. This constitutes historical charging power fluctuation information. Over the past month, the regional power grid experienced grid overload on multiple weekday evenings (7:00 PM to 10:00 PM). For example, at 8:30 PM on a certain weekday, the grid load in some areas exceeded 20% of its rated load, causing some transformers to overheat and issue overload alarms. These conditions constitute grid overload information. Abnormal charging durations were also detected. Under normal circumstances, it takes 1-1.5 hours to fully charge a certain model of pure electric car using fast charging equipment. However, during certain periods, due to unstable grid voltage or charging pile failure, the charging time of some vehicles is extended to 3-4 hours. This is abnormal charging time information.

[0033] Based on the user's charging demand information, which includes the user's vehicle type, charging time preference, battery charge level, and grid access point, historical charging power fluctuations, grid overloads, and abnormal charging durations are processed to generate charging problem type and severity assessment information. User A's vehicle type is a pure electric sedan with a 60 kWh battery capacity and a current battery charge of 30%. They plan to start charging after get off work (6:30 PM) and hope to fully charge their vehicle within two hours at a private charging station at their home in the city. Reviewing historical charging power fluctuation information reveals significant charging power fluctuations in the area during User A's planned charging time (6:30 PM-8:30 PM). Combined with User A's charging needs, this power fluctuation is determined to extend charging time, impacting their usage plans, and is therefore classified as a "charging time impact" type of charging problem.

[0034] Regarding the severity assessment of this issue, considering that user A expects a full charge in two hours, and power fluctuations may extend charging time by one to two hours, severely impacting user convenience, the issue is assessed as "moderately severe." User B, on the other hand, has an 80 kWh electric SUV with 20% remaining charge. They plan to charge their vehicle to 80% before departing on their weekend road trip (9:00 AM on Saturday) at a public charging station at a highway service area. Grid overload information indicates that the grid in this highway service area frequently experiences overload on weekend mornings (9:00 AM to 11:00 AM). Considering User B's charging needs, this overload could slow charging or even prevent normal charging, resulting in a "charging interruption risk" charging issue. Since User B urgently needs to charge before their road trip, a charging interruption would severely impact their travel plans, so the severity of this issue is assessed as "serious."

[0035] Based on the charging problem type determination information and charging problem severity assessment information, target data in the user's charging demand information is marked and filtered to generate abnormal charging demand data screening results. Based on the above charging problem type determination information and severity assessment information, target data in the user's charging demand information is marked and filtered. For user A, their charging time preference (6:30 PM - 8:30 PM), battery level information (current charge 30%, desired full charge), and grid access point information (a private charging station at home in the city) are marked as abnormal data, as these data are related to charging issues caused by charging power fluctuations. For user B, their charging time preference (9:00 AM - 11:00 AM on Saturdays), battery level information (remaining charge 20%, desired charge to 80%), and grid access point information (a public charging station at a highway service area) are marked as abnormal data, as these data are closely related to charging issues that may be caused by grid overload. These marked abnormal data are aggregated to generate the abnormal charging demand data screening results.

[0036] The results of the abnormal charging demand data screening are integrated and quantified to generate reference information for charging strategy adjustments. This information characterizes the correlation between user charging demand and historical charging and grid load issues. For user A's abnormal data, considering the impact of power fluctuations on charging time, the calculation shows that under the current power fluctuations, charging time is expected to be extended by 1.5 hours. Based on this, user A is advised to start charging half an hour earlier or charge during a period with less power fluctuations, such as the nighttime off-peak period (11:00 PM - 7:00 AM), to ensure a full charge within the expected time. For user B's abnormal data, due to the high risk of grid overload in highway service areas, the calculated probability of normal charging during this period is 60%. Therefore, user B is advised to check the real-time usage of charging stations in advance. If the grid load in the service area is too high, user B can choose a charging station in another nearby service area or charge at home the night before departure to reduce the risk of charging interruption.

[0037] These charging strategy recommendations for different users are integrated to generate reference information for charging strategy adjustments. This information clearly demonstrates the correlation between user charging needs and historical charging and grid load issues, providing an important basis for the subsequent formulation of reasonable charging scheduling strategies. For example, for users like User A who charge during periods of power fluctuations, the scheduling system can guide them to charge at more appropriate times to balance the grid load. For users like User B who are at risk of grid overload, the scheduling system can provide more reliable charging location recommendations to ensure their charging needs.

[0038] S104: Processing charging strategy adjustment reference information based on the federated learning privacy protection parameter information and the power grid information of different regions to generate a cross-regional collaborative optimization factor.

[0039] In one embodiment, charging strategy adjustment reference information is subjected to target element extraction and data classification to generate charging power adjustment information, charging time adjustment information, and charging area adjustment information. By analyzing the charging strategy adjustment reference information, extracting target elements, and then classifying them, it is assumed that when analyzing the correlation between user charging demand and historical charging and grid load issues, it is discovered that when user A charges between 7:00 PM and 8:00 PM, the grid in his area faces a high risk of overload, and charging power fluctuates significantly, impacting charging efficiency. To address this issue, charging power adjustment information is generated to reduce charging power by 30% during this period to alleviate grid load pressure. Charging time adjustment information recommends that user A postpone charging to 10:00 PM to 11:00 PM to avoid peak electricity demand. Charging area adjustment information is targeted at users who experience high charging costs or inconvenience in specific areas. For example, user B, who frequently charges at public charging stations in the city center (Area A), is advised to charge using the suburban grid (where electricity prices are relatively low) when traveling to the suburbs for errands to save costs.

[0040] Based on the privacy-preserving parameters of federated learning and information about regional power grids, a comprehensive analysis of charging power adjustment information, charging time adjustment information, and charging area adjustment information is performed to generate adjustment strategy importance assessment information and adjustment strategy feasibility assessment information. This adjustment information is comprehensively analyzed based on the privacy-preserving parameters of federated learning and information about regional power grids (such as power supply stability, electricity price differences, and load conditions). In terms of importance, adjusting charging power and time is crucial for alleviating grid overload, as it directly affects the stable operation of the grid and the normal electricity consumption of other users. The importance assessment score is set at 8 out of 10. Adjusting charging areas is also crucial for reducing user charging costs, especially for users who frequently travel across regions. The importance assessment score is set at 7. From a feasibility perspective, adjusting charging power is technically feasible with smart charging stations and has minimal impact on users, so the feasibility score was set at 9. Adjusting charging time requires users to change their charging habits, but considering that electricity prices are lower during the nighttime off-peak period, this is attractive to users, so the feasibility score was set at 7. Adjusting charging areas is highly feasible for users with cross-regional travel needs, but less feasible for users with a fixed range of activities, so the feasibility score was set at 6. This generates information on the importance and feasibility of the adjustment strategy.

[0041] Based on the adjustment strategy importance assessment information and adjustment strategy feasibility assessment information, key data in the charging strategy adjustment reference information is screened and correlated to generate the charging strategy target data screening results. For charging power adjustment information, data from users who charge during peak grid load periods and have large power fluctuations is screened, such as the charging data of User A. For charging time adjustment information, data from users who charge during peak hours and have the possibility of adjusting their charging times is correlated, such as User A and some users with flexible charging times. For charging area adjustment information, data from users who frequently travel across regions and charge in high-cost areas is selected, such as User B. This filtered data is integrated to form the charging strategy target data screening results.

[0042] The results of the charging strategy target data screening are integrated and processed, combining the federated learning privacy protection parameter information (such as data encryption method and data sharing scope) and the weighting information of different regional power grids (for example, the importance weights of different regional power grids: the weight of the downtown power grid is set to 0.4 due to its high load, the weight of the suburban power grid is set to 0.3, and the weight of the industrial park power grid is set to 0.3). For example, a cross-regional collaborative optimization factor is calculated to be 0.65. This value indicates that in the current situation, the direction of cross-regional collaborative optimization is to encourage users to rationally distribute charging behavior across different regions, prioritizing charging in areas with lower load and lower costs. In terms of degree, this means that there is a 65% probability of achieving more efficient charging scheduling through cross-regional collaborative optimization. Using this cross-regional collaborative optimization factor, the system can guide new energy vehicle users in different regions to rationally adjust their charging strategies, achieve optimal cross-regional power resource allocation, balance the load of different regional power grids, reduce user charging costs, and improve overall charging efficiency and grid stability.

[0043] S105 : Processing the user charging demand information and the regional information corresponding to the user charging demand information based on the cross-region collaborative optimization factor to generate charging demand classification information and priority information.

[0044] In one implementation, data extraction and processing is performed on user charging demand information to generate information about the user's vehicle type, charging time preference, battery charge level, and grid access point. Assume there are three new energy vehicle users in the area. User A's vehicle is a pure electric sedan with a 60 kWh battery and a current charge of 30%. He typically starts charging at 6:30 PM after get off work, often at a private charging station at his home in the city. User B's vehicle is an electric SUV with an 80 kWh battery and a remaining charge of 20%. He plans to charge before his weekend road trip (9:00 AM on Saturday) at a public charging station in a highway service area. User C's vehicle is an electric minivan with a 35 kWh battery and a current charge of 40%. He typically charges around 10:00 PM at night, often at a charging station in his company's parking lot. By extracting and processing data on these users' charging demand information, we can obtain user vehicle type information (User A: pure electric sedan; User B: electric SUV; User C: electric microcar), charging time preference information (User A: 18:30; User B: Saturday morning 9:00; User C: around 22:00), battery power information (User A: 30%; User B: 20%; User C: 40%), and grid access point information (User A: private charging station at home in the city; User B: public charging station at the highway service area; User C: charging station in the company parking lot).

[0045] Based on the regional information corresponding to the user's charging needs, the user's vehicle type, charging time preference, battery level, and grid access point information are standardized and integrated, converting all types of information into a unified, processable format to generate a standardized charging demand information set. This region is divided into different subgrids: the city center grid, the suburban grid, and the industrial park grid. The regional information corresponding to User A shows that the urban grid in which they are located has stable voltage and high power supply reliability. The highway service area corresponding to User B is located at the junction of the suburban grid and the city center grid, and power supply may be tight during peak hours. User C's company is located in an industrial park, and the regional grid mainly provides industrial power, with certain preferential policies for residential charging at night.

[0046] Based on this regional information, users' vehicle type, charging time preference, battery level, and grid access point information are standardized and integrated. For example, vehicle type information is uniformly coded: pure electric sedans are coded 01, electric SUVs are coded 02, and electric minicars are coded 03. Charging time preference information is uniformly converted to a numerical value in hours (18:30 is converted to 18.5, 9:00 is converted to 9, and around 22:00 is converted to 22). Battery level information is uniformly converted to a numerical value expressed as a percentage. Grid access point information is categorized and coded based on the regional power grid, such as a private charging station at home in the city is coded A01, a public charging station at a highway service area is coded B02, and a charging station in a company parking lot is coded C03. This integration generates a standardized set of charging demand information. For example, user A's standardized information is [01, 18.5, 30, A01].

[0047] Based on the cross-regional collaborative optimization factor, the standardized charging demand information set is evaluated and filtered for importance. Based on the optimization direction and degree reflected by the cross-regional collaborative optimization factor, the importance weights of each piece of information in charging demand classification and prioritization are generated. Assuming the cross-regional collaborative optimization factor is 0.7, as calculated previously, this means that cross-regional collaborative optimization has a significant positive impact on charging scheduling in the current situation, tending to guide users to rationally distribute their charging behavior across different regions, prioritizing charging in areas with lower load and lower costs. Based on this, the standardized charging demand information set is evaluated and filtered for importance. Regarding charging time preference information, since cross-regional collaborative optimization emphasizes avoiding peak hours, the importance weight of charging time preference during peak hours (e.g., 6:00 PM to 8:00 PM) decreases, while the importance weight of charging time preference during off-peak hours (e.g., 10:00 PM to 6:00 AM) increases. For example, user A's original time preference weight for charging at 6:30 PM might be set at 0.3, but this weight is adjusted to 0.2 after evaluation. User C's time preference weight for charging at 10:00 PM increases from 0.3 to 0.4.

[0048] For battery level information, the importance of this information in prioritizing users with low battery levels who urgently need to charge will be increased. For example, if user B has 20% remaining battery and needs to charge before a road trip, the weight of their battery level will be increased from 0.2 to 0.3. For user C, whose battery level is currently at 40%, the weight will be adjusted from 0.2 to 0.15. For grid access point information, the weight of locations in high-load areas (such as some charging stations in the city center) will be reduced, while the weight of locations in low-load and low-cost areas (such as some charging stations in the suburbs) will be increased. For example, the weight of user A's private charging station in the city center will be reduced from 0.3 to 0.25, while if there is a charging station in the suburbs, its weight may be increased from 0.2 to 0.3. This generates the importance weights of each piece of information in charging demand classification and prioritization.

[0049] The importance weight of each piece of information in charging demand classification and priority determination is processed to generate charging demand classification and priority information. A comprehensive score is calculated based on the weights. User B's battery level is low and charging time falls during the weekend peak period. Although the weights of public charging piles in highway service areas are adjusted due to regional factors, the comprehensive score shows that their charging needs are relatively urgent. Therefore, they can be classified as "urgent charging needs" with a high priority. User A's current battery level is not very low, but charging time falls during the evening peak electricity consumption period. The comprehensive score results in a "general charging need" category with a medium priority. User C's battery level is relatively high and charging time falls during the night off-peak period. Therefore, they are classified as "delayable charging needs" with a low priority. This method clarifies the charging demand classification and priority of different users, providing an important basis for subsequent charging scheduling.

[0050] S106 , inputting the benefit evaluation parameter information, charging demand classification information and priority information, and historical charging and grid load problem information into a scheduling model based on the target Actor-Critic algorithm for processing to generate charging scheduling allocation result information.

[0051] In one implementation, training data includes dynamic grid electricity price information, carbon emission signal information, user charging demand information, historical charging and grid load issue information, regional grid information, federated learning privacy protection parameter information, and corresponding charging scheduling and allocation result samples. Similar information and samples are also obtained for verification. An initial scheduling model for the pre-set Actor-Critic algorithm is also obtained. Dynamic grid electricity price information for the past year is collected from the regional grid operation system, including price data for different time periods on weekdays and weekends. For example, the average price for weekday peak hours (8:00 AM - 6:00 PM) is 0.8 yuan / kWh, and the average price for nighttime off-peak hours (11:00 PM - 7:00 AM) is 0.3 yuan / kWh. Carbon emission signal information is obtained from environmental protection and energy monitoring departments, such as the carbon emission intensity of regional thermal power generation of 0.8 kg / kWh, carbon emission intensity data for hydropower and wind power, as well as total carbon emission limits and real-time surpluses for different time periods.

[0052] Through mobile phone charging apps and in-vehicle systems, users of new energy vehicles can collect information such as: User A's vehicle type is a pure electric sedan with a 60 kWh battery capacity, currently at 30% charge. They typically start charging at 6:30 PM after get off work, expecting a full charge in two hours, and typically charge at a private charging station at home in the city. User B's vehicle is an electric SUV with an 80 kWh battery capacity, currently at 20% charge, and they plan to charge to 80% at a public charging station in a highway service area at 9:00 AM on weekends. Historical charging and grid load issue information is also collected from the power grid history database. For example, during the past month, concentrated charging in some areas between 7:00 PM and 10:00 PM on weekdays caused grid overload and transformer overheating alarms; and charging power fluctuated significantly during certain periods. Regional grid information is also collected, such as differences in power source composition, load characteristics, and power supply capacity between the city center, suburban areas, and industrial parks. In multi-institutional collaboration, privacy protection parameters for federated learning are set, such as using homomorphic encryption algorithms for data encryption and computation, sharing only aggregated and anonymized statistical information, allowing grid operators to view only grid load-related data, and automakers to access only aggregated data on vehicle charging behavior. The charging scheduling and allocation results for this information at different time periods are collected as samples. For example, if user A follows the system's recommended charging strategy during a certain time period, the actual charging power, duration, and cost data are collected. Similar verification information and samples are obtained in the same manner for subsequent model validation.

[0053] A preset Actor-Critic algorithm initial scheduling model is selected. This model is a two-layer neural network structure. Both the actor and critic networks contain one hidden layer. The number of nodes in the actor network input layer is determined by the number of input data features. Assuming 10 features are considered (such as electricity price fluctuations, carbon emission intensity, vehicle type, etc.), the number of input layer nodes is 10. The number of hidden layer nodes is set to 30, determined through experience and experiments, and is used for feature extraction and nonlinear transformation of the input data. The number of output layer nodes is 3, corresponding to the charging power adjustment, charging time adjustment, and charging area selection (assuming three areas). The critic network also has 10 input layer nodes, 30 hidden layer nodes, and 1 output layer node, which is used to evaluate state value.

[0054] Data cleaning, feature extraction, and normalization are performed on the training grid's dynamic electricity price information, carbon emission signal information, user charging demand information, historical charging and grid load issue information, regional grid information, and federated learning privacy protection parameter information to generate preprocessed training feature data. This preprocessed training feature data includes characteristics of electricity price fluctuations, peak and valley periods, regional differences in electricity prices, changes in carbon emission intensity, peak periods, regional carbon emission trends, user vehicle type information, charging time preferences, battery level information, and grid access point information. The training grid's dynamic electricity price information is checked for outliers. If the electricity price at a specific moment is significantly incorrect (e.g., several times exceeding the normal range), it is removed or corrected. Incomplete or erroneous data in user charging demand information is processed, such as data with missing battery level information or incorrect charging time format. Extract price fluctuation characteristics from dynamic grid electricity price information. For example, calculate the difference between peak and off-peak electricity prices during a given week and then average it. Extract peak and off-peak price characteristics to identify weekday and weekend peak and off-peak periods. Extract regional price differences to compare price differences across regional grids. For carbon emission signal information, extract carbon emission intensity variation characteristics to calculate the rate of change in thermal power carbon emission intensity over the past year. Extract peak emission time characteristics to identify peak carbon emissions during the winter heating season and during peak electricity consumption hours each day. Extract regional carbon emission trend characteristics to analyze carbon emission trends across different regions (e.g., city centers, suburbs, and industrial parks). Extract user charging demand information to extract information about vehicle type, charging time preferences, battery charge level, and grid access points.

[0055] Numerical features such as electricity price fluctuations and carbon emission intensity change rates are normalized and mapped to the [0,1] interval to make different features comparable. For example, the maximum and minimum values ​​of electricity price fluctuations are normalized. Assuming that the maximum price fluctuation in a certain region is 0.6 yuan / kWh and the minimum is 0.1 yuan / kWh, and the fluctuation at a certain moment is 0.3 yuan / kWh, the normalized value is (0.3-0.1) / (0.6-0.1)=0.4. User vehicle type information is encoded, such as 01 for a pure electric sedan and 02 for an electric SUV. Charging time preference information is converted to a numerical value in hours and then normalized. Battery charge information is directly expressed as a percentage, without further normalization. Grid access point information is encoded according to the region to which it belongs and also normalized accordingly. The final preprocessed training feature data is generated, including electricity price fluctuation amplitude characteristics, electricity price peak and valley period characteristics, electricity price regional difference characteristics, carbon emission intensity change characteristics, carbon emission peak period characteristics, carbon emission trend characteristics in different regions, user vehicle type information, charging time preference information, battery power information, and grid access point information.

[0056] The preprocessed training feature data is processed to generate scheduling model prediction parameters. The scheduling model prediction parameter vector represents the predicted information about the charging schedule allocation results, key factors affecting charging scheduling, and optimization strategies. The actor-critic algorithm model is trained using the backpropagation algorithm using the preprocessed training feature data. During training, the model continuously adjusts the connection weights between neurons to minimize the error between the predicted charging schedule allocation results and the actual results. For example, in one training session, the model predicts a charging schedule strategy (charging power, time, and region) based on current input features (such as a user's vehicle type, charging time, current electricity price, and carbon emissions). The predicted results are compared with actual charging schedule allocation results and the error is calculated. Based on the error, the backpropagation algorithm adjusts the weights of neurons in each layer of the model to make subsequent predictions more accurate. After multiple training iterations, the model learns the characteristic patterns in the data and generates the scheduling model prediction parameter vector. This vector contains prediction information on the charging scheduling allocation results (such as expected charging power, charging time, and charging area), as well as prediction information on key factors affecting charging scheduling (such as the impact of electricity price fluctuations on charging costs, the impact of grid load in different regions on charging feasibility, etc.) and optimization strategies (such as recommendations to increase charging power during periods of low electricity prices).

[0057] Based on the scheduling model's predicted parameter vector, the preset Actor-Critic algorithm's initial scheduling model is optimized and trained to generate a trained Actor-Critic algorithm scheduling model. The weight information in the predicted parameter vector is applied to the connection weights of the initial model to further adjust the model parameters. During training, a learning rate of 0.01 is set, which determines the step size for parameter updates at each iteration. The number of iterations is set to 500, allowing the model to continuously adapt to the training data. For example, in each iteration, the model calculates output based on the new weights, compares the output with the actual results, and adjusts the weights again to make the model's predictions on the training data more accurate. After this optimization training, a trained Actor-Critic algorithm scheduling model is generated. At this point, the model can more accurately predict charging scheduling strategies based on various input information and better account for the impact of various factors on charging scheduling.

[0058] Based on validation information for dynamic grid electricity prices, carbon emission signals, user charging demand, historical charging and grid load issues, regional grid information, federated learning privacy protection parameters, and corresponding charging schedule allocation results, the trained Actor-Critic algorithm scheduling model is simulated for charging scheduling to generate validation results. The validation samples are fed into the trained model, which outputs predictions for these users, such as the charging power, charging time, and charging area for user A during the current validation period. The predictions are compared with the actual charging schedule allocation results, and relevant metrics, such as the mean squared error between the predicted and actual charging power and the error between the predicted and actual charging time, are calculated to generate validation results. These metrics are used to evaluate the model's performance on the validation samples and determine whether the model can accurately predict charging schedules.

[0059] Based on the validation results, the trained Actor-Critic algorithm scheduling model is evaluated and adjusted to generate a target Actor-Critic algorithm scheduling model. If the validation results indicate large model errors or low accuracy, such as a mean squared error exceeding a set threshold, this indicates overfitting or underfitting. If the model is found to be underfitting (large error and low accuracy) on the validation samples, the number of hidden layer neurons can be increased and retraining can be performed. If overfitting is detected (performing well on the training samples but poorly on the validation samples), the learning rate or number of iterations can be appropriately reduced. After multiple evaluations and adjustments, the model achieves optimal performance on the validation samples, ultimately generating a target Actor-Critic algorithm scheduling model. This target model enables more accurate evaluation and decision-making for charging scheduling of new energy vehicles, providing strong support for the subsequent input of benefit evaluation parameters, charging demand classification and priority information, and historical charging and grid load information into the model to generate accurate charging scheduling and allocation results.

[0060] In another embodiment, a scheduling model based on a targeted actor-critic algorithm analyzes and processes benefit evaluation parameter information, charging demand classification and priority information, and historical charging and grid load information to generate a charging resource demand correlation assessment value. Using the aforementioned method, the charging demand classification and priority information, benefit evaluation parameter information, and historical charging and grid load information for users A, B, and C are obtained. User A is a pure electric sedan with a current battery level of 30% and plans to charge at a private charging station at home at 6:30 PM, representing a general charging demand. User B is an electric SUV with a remaining battery level of 20% and plans to charge at a public charging station in a highway service area at 9:00 AM on Saturday, representing an urgent charging demand. User C is an electric mini car with a remaining battery level of 40% and plans to charge at a charging station in the company parking lot at 10:00 PM, representing a deferrable charging demand. Benefit evaluation parameter information indicates that electricity prices fluctuate moderately during the current period, and changes in carbon emission intensity have a minimal impact on costs. Historical charging and grid load information indicates that grid loads are high in some areas between 6:00 PM and 8:00 PM.

[0061] This information is input into the target actor-critic algorithm's scheduling model. The model comprehensively considers these factors and analyzes each user's charging resource needs. For user A, the model considers that their charging time occurs during a period of high grid load and their battery level is not extremely low, and calculates a charging resource demand correlation assessment value of 0.5 (ranging from 0 to 1, with higher values ​​indicating more urgent resource needs). For user B, since they have an urgent charging need and charge during peak weekend hours, the model assigns a charging resource demand correlation assessment value of 0.8. User C, whose battery level is relatively high and who charges during a period of low load, receives a charging resource demand correlation assessment value of 0.3. These assessment values ​​reflect each user's current demand for charging resources, taking into account factors such as user needs, grid conditions, electricity prices, and carbon emissions.

[0062] Based on the charging resource demand correlation evaluation value, the charging strategy decision vector within the target Actor-Critic algorithm scheduling model is optimized and adjusted to generate a charging schedule deviation vector. Assuming the model's original charging strategy decision vector is [charging power adjustment, charging time adjustment, charging zone selection], for user A, the original decision vector might be [maintain normal power, charge on schedule, current zone]. Because their charging resource demand correlation evaluation value is 0.5, the model determines that the current strategy may need to be adjusted to balance grid load and user demand. After optimization and adjustment, the resulting charging schedule deviation vector is [reduce charging power by 10%, delay charging by 0.5 hours, maintain current zone].

[0063] For user B, due to their urgent demand, the model will increase their allocation of charging resources. The original decision vector might be [normal power, charging as planned, current zone], but the adjusted charging scheduling deviation vector becomes [increase charging power by 20%, prioritize charging, maintain current zone]. For user C, given their less urgent demand and their advantage during off-peak hours, the original decision vector was [normal power, charging on schedule, current zone], but the adjusted deviation vector becomes [maintain normal power, delay charging by 0.5-1 hour, maintain current zone]. These deviation vectors reflect the direction and extent of the model's adjustment to the original charging strategy based on the associated assessment of the user's charging resource demand.

[0064] The charging schedule deviation vector is parsed and converted to generate charging schedule allocation results. This information represents the grid resources allocated to new energy vehicle charging and the target quantitative indicators for charging resource allocation. For example, parsing the charging schedule deviation vector for user A reveals an adjusted charging strategy: charging begins at 7:00 PM, with a 10% reduction in charging power, and charging occurs at the user's home charging station (in the current area). Combined with grid resource information, the grid resources allocated to user A are determined to be part of the electricity resources in their home's electricity consumption area during that time period. While the reduced power may extend charging time, this will not affect the overall grid load balance. Based on the target quantitative indicators for charging resource allocation, user A's charging costs may be slightly reduced due to the power adjustment and time delay, and carbon emissions will also be reduced. This will not place excessive pressure on the grid, ensuring normal power supply for other users.

[0065] For user B, after analyzing the deviation vector, the charging strategy is to immediately charge at a public charging station in a highway service area at a 20% higher power level. The grid resources allocated to user B are sufficient power, prioritized by the service area, ensuring that charging can be completed as quickly as possible to meet the urgent needs of self-driving tours. In terms of target quantitative indicators, while charging costs may increase due to the increased power level, this meets the user's urgent needs, and through reasonable scheduling, the impact on the overall grid is within an acceptable range. User C's charging schedule is to begin charging between 10:30 PM and 11:00 PM, maintaining normal power, and charging at a charging station in the company parking lot. The allocated grid resources are surplus power from the industrial park's nighttime off-peak hours, ensuring optimal utilization of power resources. From the perspective of target quantitative indicators, user C's charging costs are low, helping to reduce peak loads and fill valleys on the grid, while also meeting their need for delayed charging. In this way, the generated charging scheduling and allocation result information clarifies the grid resource information allocated to each new energy vehicle user for charging, as well as the target quantitative indicators of charging resource allocation considering multiple aspects such as cost, efficiency, and grid load balance, providing a basis for achieving efficient and reasonable charging scheduling.

[0066] The server obtains information such as dynamic grid electricity prices, carbon emission signals, user charging demands, historical charging and grid load issues, regional grids, and federated learning privacy-preserving parameters. Next, it extracts and quantifies the dynamic grid electricity prices and carbon emission signals, and combines them with privacy-preserving parameters to generate benefit evaluation parameters. Simultaneously, it processes historical charging and grid load issue information, along with user charging demand information, to generate reference information for adjusting charging strategies. This information is then processed based on the federated learning privacy-preserving parameters and regional grid information to generate cross-regional collaborative optimization factors, which are used to generate charging demand classification and priority information.

[0067] Afterwards, by acquiring training and validation data, the preset Actor-Critic algorithm initial scheduling model undergoes data cleaning, feature extraction, normalization, optimization training, and validation adjustments to obtain the target scheduling model. Finally, the target model is fed with revenue assessment parameters, charging demand classification and priority information, and historical charging and grid load problem information. After analysis and processing, strategy vector optimization and adjustment, and deviation vector analytical conversion, the charging scheduling allocation results are generated, covering the allocated grid resources and target quantitative indicators. This achieves charging scheduling optimization that comprehensively considers cost, efficiency, and grid stability, thereby improving the charging management level of new energy vehicles.

[0068] In one embodiment, Figure 2 As shown, the present application also provides a charging scheduling device for automobile electricity prices and carbon emissions based on Actor-Critic, including:

[0069] Acquisition module 201, used to obtain dynamic grid electricity price information, carbon emission signal information, user charging demand information, historical charging and grid load problem information, grid information in different regions, and federated learning privacy protection parameter information;

[0070] Processing module 202 is used to process the dynamic electricity price information and carbon emission signal information of the power grid based on the federated learning privacy protection parameter information to generate benefit evaluation parameter information; process the user charging demand information based on the historical charging and power grid load problem information to generate charging strategy adjustment reference information; process the charging strategy adjustment reference information based on the federated learning privacy protection parameter information and the power grid information of different regions to generate a cross-regional collaborative optimization factor; process the user charging demand information and the regional information corresponding to the user charging demand information based on the cross-regional collaborative optimization factor to generate charging demand classification information and priority information, wherein the user charging demand information is used to characterize the user's vehicle type, charging time preference, battery power information and power grid access point information; input the benefit evaluation parameter information, charging demand classification information and priority information and the historical charging and power grid load problem information into the scheduling model based on the target Actor-Critic algorithm for processing to generate charging scheduling allocation result information.

[0071] Each embodiment in this application is described in a related manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the embodiment of the charging scheduling method, electronic device, electronic device, and readable storage medium for evaluating the Actor-Critic-based automobile electricity price and carbon emissions, since they are basically similar to the embodiment of the charging scheduling method for automobile electricity price and carbon emissions based on Actor-Critic, the description is relatively simple, and the relevant parts can be referred to the partial description of the embodiment of the charging scheduling method for automobile electricity price and carbon emissions based on Actor-Critic.

Claims

1. A charging scheduling method based on automobile electricity price and carbon emissions based on Actor-Critic, characterized by: include: Obtain dynamic grid electricity price information, carbon emission signal information, user charging demand information, historical charging and grid load problem information, grid information in different regions, and federated learning privacy protection parameter information; Based on the privacy-preserving parameter information of federated learning, the dynamic electricity price information and carbon emission signal information of the power grid are processed to generate the benefit evaluation parameter information; Process user charging demand information based on historical charging and grid load problem information to generate reference information for charging strategy adjustment; Based on the privacy-preserving parameter information of federated learning and the information of power grids in different regions, the charging strategy adjustment reference information is processed to generate a cross-regional collaborative optimization factor. The cross-regional collaborative optimization factor is used to characterize the direction and degree of cross-regional collaborative optimization of new energy vehicle charging strategies in different regions. Based on the cross-regional collaborative optimization factor, user charging demand information and regional information corresponding to the user charging demand information are processed to generate charging demand classification information and priority information. The user charging demand information is used to characterize the user's vehicle type, charging time preference, battery power information, and grid access point information. The benefit evaluation parameter information, charging demand classification information and priority information, and historical charging and grid load problem information are input into the scheduling model based on the target Actor-Critic algorithm for processing to generate charging scheduling allocation result information.

2. The method according to claim 1, wherein Based on the privacy-preserving parameter information of federated learning, the dynamic electricity price information and carbon emission signal information of the power grid are processed to generate the benefit evaluation parameter information, including: Perform feature extraction and processing on the dynamic electricity price information of the power grid to generate electricity price fluctuation amplitude characteristics, electricity price peak and valley period characteristics, and electricity price regional difference characteristics; Perform feature extraction and processing on carbon emission signal information to generate carbon emission intensity change characteristics, carbon emission peak period characteristics, and carbon emission trend characteristics in different regions; Quantitative analysis is performed on the characteristics of electricity price fluctuation amplitude, peak and valley period characteristics, and regional differences in electricity prices to generate quantitative characteristic values ​​of electricity prices; Quantitatively analyze and process the changing characteristics of carbon emission intensity, the characteristics of carbon emission peak periods, and the characteristics of carbon emission trends in different regions to generate quantitative characteristic values ​​of carbon emissions; Based on the privacy protection parameter information of federated learning, the quantitative characteristic values ​​of electricity prices and carbon emissions are processed to generate comprehensive benefit evaluation characteristics and corresponding weight calculation result information; The comprehensive benefit evaluation characteristics and the corresponding weight calculation result information are processed to generate benefit evaluation parameter information, wherein the benefit evaluation parameter information is used to characterize the impact of the dynamic electricity price of the power grid and the carbon emission signal on the charging benefits of new energy vehicles.

3. The method according to claim 1, wherein Based on historical charging and grid load problem information, user charging demand information is processed to generate reference information for charging strategy adjustment, including: Extract and classify historical charging and grid load problem information to generate historical charging power fluctuation information, grid load overload information, and abnormal charging time information; Based on the user's charging demand information, including the user's vehicle type, charging time preference, battery level information, and grid access point information, the system processes historical charging power fluctuation information, grid load overload information, and abnormal charging time information to generate charging problem type judgment information and charging problem severity assessment information; Mark and filter target data in the user's charging demand information based on the charging problem type judgment information and the charging problem severity assessment information, and generate charging demand abnormal data screening results; The results of the abnormal charging demand data screening are integrated and quantified to generate charging strategy adjustment reference information, where the charging strategy adjustment reference information is used to characterize the correlation between user charging demand and historical charging and grid load issues.

4. The method according to claim 1, wherein Based on the privacy-preserving parameter information of federated learning and the information of power grids in different regions, the charging strategy adjustment reference information is processed to generate cross-regional collaborative optimization factors, including: Perform target element extraction and data classification processing on the charging strategy adjustment reference information to generate charging power adjustment information, charging time adjustment information, and charging area adjustment information; Based on the privacy-preserving parameter information of federated learning and the information of power grids in different regions, a comprehensive analysis is performed on the charging power adjustment information, charging time adjustment information, and charging area adjustment information to generate adjustment strategy importance assessment information and adjustment strategy feasibility assessment information; Based on the adjustment strategy importance assessment information and the adjustment strategy feasibility assessment information, the key data in the charging strategy adjustment reference information is screened and correlated to generate a charging strategy target data screening result; The charging strategy target data screening results are integrated and processed in combination with the privacy protection parameter information of federated learning and the weight information in the power grid information of different regions to generate a cross-regional collaborative optimization factor.

5. The method according to claim 1, wherein Based on the cross-region collaborative optimization factor, the user charging demand information and the regional information corresponding to the user charging demand information are processed to generate charging demand classification information and priority information, including: Extract and process user charging demand information to generate user vehicle type information, charging time preference information, battery power information, and grid access point information; Based on the regional information corresponding to the user's charging demand information, the user's vehicle type information, charging time preference information, battery power information, and grid access point information are standardized and integrated, and various types of information are converted into a unified processable format to generate a standardized charging demand information set; Based on the cross-regional collaborative optimization factors, the importance of the standardized charging demand information set is evaluated and screened. According to the optimization direction and degree reflected by the cross-regional collaborative optimization factors, the importance weight of each information in the charging demand classification and priority judgment is generated; The importance weight of each information in charging demand classification and priority judgment is processed to generate charging demand classification information and priority information.

6. The method according to claim 1, wherein It also includes the scheduling model for generating the target Actor-Critic algorithm, specifically: Obtain dynamic grid electricity price information, carbon emission signal information, user charging demand information, historical charging and grid load problem information, grid information in different regions, federated learning privacy protection parameter information, and corresponding charging scheduling allocation result samples for training; obtain similar information and samples for verification; and obtain the initial scheduling model of the preset Actor-Critic algorithm. Perform data cleaning, feature extraction, and normalization on the training grid dynamic electricity price information, carbon emission signal information, user charging demand information, historical charging and grid load problem information, regional grid information, and federated learning privacy protection parameter information to generate preprocessed training feature data. The preprocessed training feature data includes electricity price fluctuation amplitude characteristics, electricity price peak and valley period characteristics, electricity price regional difference characteristics, carbon emission intensity change characteristics, carbon emission peak period characteristics, carbon emission trend characteristics in different regions, user vehicle type information, charging time preference information, battery power information, and grid access point information. Processing the preprocessed training feature data to generate a scheduling model prediction parameter vector, wherein the scheduling model prediction parameter vector is used to represent the prediction information of the charging scheduling allocation results, key factors affecting the charging scheduling, and optimization strategy; Based on the scheduling model prediction parameter vector, the preset Actor-Critic algorithm initial scheduling model is optimized and trained to generate a trained Actor-Critic algorithm scheduling model; Based on the verification information of dynamic grid electricity prices, carbon emission signals, user charging demand information, historical charging and grid load problem information, grid information in different regions, federated learning privacy protection parameter information, and corresponding charging scheduling allocation result samples, the trained Actor-Critic algorithm scheduling model is simulated for charging scheduling to generate verification results. Based on the verification results, the trained Actor-Critic algorithm scheduling model is evaluated and adjusted to generate the scheduling model of the target Actor-Critic algorithm.

7. The method according to claim 6, wherein The benefit evaluation parameter information, charging demand classification information and priority information, and historical charging and grid load problem information are input into the scheduling model based on the target Actor-Critic algorithm for processing to generate charging scheduling allocation result information, including: The scheduling model based on the target actor-critic algorithm analyzes and processes the benefit evaluation parameter information, charging demand classification information and priority information, and historical charging and grid load problem information to generate the charging resource demand correlation evaluation value; Based on the charging resource demand correlation evaluation value, the charging strategy decision vector within the target Actor-Critic algorithm scheduling model is optimized and adjusted to generate a charging scheduling deviation vector. The charging scheduling deviation vector is parsed and converted to generate charging scheduling allocation result information, wherein the charging scheduling allocation result information is used to represent the grid resource information allocated to the charging of new energy vehicles and the target quantitative indicators of charging resource allocation.

8. A charging scheduling device for automobile electricity prices and carbon emissions based on Actor-Critic, characterized by: The device comprises: The acquisition module is used to obtain dynamic grid electricity price information, carbon emission signal information, user charging demand information, historical charging and grid load problem information, grid information in different regions, and federated learning privacy protection parameter information; A processing module is used to process the dynamic electricity price information and carbon emission signal information of the power grid based on the privacy protection parameter information of federated learning to generate benefit evaluation parameter information; process the user charging demand information based on the historical charging and power grid load problem information to generate charging strategy adjustment reference information; process the charging strategy adjustment reference information based on the privacy protection parameter information of federated learning and the power grid information of different regions to generate a cross-regional collaborative optimization factor, wherein the cross-regional collaborative optimization factor is used to characterize the direction and degree of cross-regional collaborative optimization of the charging strategy of new energy vehicles in different regions; process the user charging demand information and the regional information corresponding to the user charging demand information based on the cross-regional collaborative optimization factor to generate charging demand classification information and priority information, wherein the user charging demand information is used to characterize the user's vehicle type, charging time preference, battery power information and power grid access point information; input the benefit evaluation parameter information, charging demand classification information and priority information and the historical charging and power grid load problem information into the scheduling model based on the target Actor-Critic algorithm for processing to generate charging scheduling allocation result information.

9. An electronic device, characterized in that: include: a first processor; and a memory for storing executable instructions of the first processor; The first processor is configured to execute the Actor-Critic based vehicle electricity price and carbon emission charging scheduling method according to any one of claims 1 to 7 by executing the executable instructions.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the second processor, the method for charging scheduling of automobile electricity price and carbon emissions based on Actor-Critic is implemented.

Images