Grid-connected load intelligent simulation and regulation method for multi-model electric vehicles

Abstract

Embodiments of the present disclosure disclose a grid-connected load intelligent simulation and regulation method for multi-model electric vehicles. A specific embodiment of the method includes: for each charging station, generating local load regulation information; generating global load regulation information according to each regulation information; for the global load regulation information, performing the processing steps: sending each load regulation information to the corresponding charging station; for each charging station, performing the sending step: performing charging simulation simulation; sending the corresponding load problem set to the regional agent; in response to determining not to perform load regulation, sending the global load regulation information to the corresponding power control intelligent agent. Based on the charging conditions of multi-model electric vehicles in the target peak time period, the embodiment utilizes the overall load scheduling of the regional agent and the load simulation simulation of each edge intelligent agent to achieve precise load scheduling for each charging station, avoiding the problem of insufficient power supply.

Description

Technical Field

[0001] Embodiments of this disclosure relate to the fields of intelligent agents and electric vehicle technology, and more specifically to a method, apparatus, electronic device, and computer-readable medium for intelligent simulation and control of grid-connected load for multi-model electric vehicles. Background Technology

[0002] Currently, electric vehicle charging stations are a core infrastructure supporting the widespread adoption of new energy vehicles. With the advancement of the national "dual-carbon" strategy and the explosive growth in new energy vehicle sales, charging demand has surged. As a new type of urban infrastructure, charging stations not only solve car owners' "range anxiety" and "charging difficulties," but are also a crucial link in promoting the green and low-carbon transformation of transportation and optimizing the energy structure. For the management of power load on substations within a target commercial area, the common approach is as follows: For each charging station, firstly, obtain the historical charging information corresponding to the charging station. Then, input the historical charging information into the load forecasting model to obtain the load control information corresponding to the charging station.

[0003] However, when using the above method, the following technical problems often arise: Load forecasting based on load forecasting models suffers from poor initiative and, based on full data, cannot effectively address the charging congestion problem of electric vehicles during peak hours. Furthermore, during peak hours, it cannot effectively manage the power load of charging stations based on the overall regional electricity consumption, potentially leading to insufficient power supply for other electrical equipment.

[0004] The information disclosed in this background section is only intended to enhance the understanding of the background of the inventive concept, and therefore may contain information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The summary portion of this disclosure is intended to provide a brief overview of the concepts, which will be described in detail in the detailed description portion. This summary portion is not intended to identify key or essential features of the claimed technical solutions, nor is it intended to limit the scope of the claimed technical solutions.

[0006] Some embodiments of this disclosure provide a method, apparatus, electronic device, and computer-readable medium for intelligent simulation and control of grid-connected load for multi-model electric vehicles to solve one or more of the technical problems mentioned in the background section above.

[0007] In a first aspect, some embodiments of this disclosure provide a method for intelligent simulation and control of grid-connected load for multi-model electric vehicles, comprising: for each charging station in a target commercial area, generating local load control information for the charging station in future peak periods based on charging information during a target peak time period, using an edge agent deployed on a corresponding edge terminal, wherein the charging information includes vehicle model information; generating global load control information based on each control information using a regional agent corresponding to the target commercial area, wherein the control information includes charging information and local load control information; for the global load control information, performing a processing step using the regional agent: sending each load control information in the global load control information to the corresponding charging station; for each charging station, performing a sending step using a corresponding edge agent: performing a charging simulation during a target off-peak time period based on the corresponding load control information, obtaining simulation results; sending the load problem set corresponding to the simulation results to the regional agent; determining whether to perform load control based on the obtained load problem sets and simulation results; and in response to determining not to perform, sending the global load control information to the power management agent corresponding to the target commercial area to control the load corresponding to each charging station.

[0008] Secondly, some embodiments of this disclosure provide an intelligent simulation and control device for grid-connected load of multi-model electric vehicles, comprising: a first generation unit configured to, for each charging station in a target commercial area, generate local load control information for the charging station in a future peak period based on charging information during a target peak period, using an edge agent correspondingly deployed on an edge terminal; the charging information includes: vehicle model information; a second generation unit configured to, based on each control information, generate global load control information using an area agent corresponding to the target commercial area; the control information includes: charging information and local load control information; and an execution unit configured to, for the entire... The load control information is processed using the aforementioned regional agent, following these steps: Each load control information in the global load control information is sent to the corresponding charging station; for each charging station, the corresponding edge agent performs the following sending steps: based on the corresponding load control information, a charging simulation is performed during the target off-peak period to obtain the simulation results; the load problem set corresponding to the simulation results is sent to the regional agent; based on the obtained load problem sets and simulation results, it is determined whether to perform load control; in response to the determination not to perform load control, the global load control information is sent to the power management agent corresponding to the target commercial area to regulate the load corresponding to each charging station.

[0009] Thirdly, some embodiments of this disclosure provide an electronic device, including: one or more processors; and a storage device having one or more programs stored thereon, such that when the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any implementation of the first aspect.

[0010] Fourthly, some embodiments of this disclosure provide a computer-readable medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method as described in any implementation of the first aspect.

[0011] The above-described embodiments of this disclosure have the following beneficial effects: Through the intelligent simulation and control method for grid-connected load of multi-model electric vehicles according to some embodiments of this disclosure, based on the charging status of multi-model electric vehicles during the target peak time period, and utilizing the overall load scheduling of regional intelligent agents and the load simulation of each edge intelligent agent, precise load scheduling of each charging station can be achieved, avoiding the problem of insufficient power supply. Specifically, the reason for the inaccurate load control of charging stations during peak periods is that load prediction based on load prediction models has poor initiative, and load prediction based on full data cannot effectively solve the problem of electric vehicle charging congestion during peak periods. Furthermore, during peak periods, effective control of the power load corresponding to charging stations cannot be achieved based on the overall regional power consumption, which may lead to insufficient power supply for other electrical equipment during peak periods. Therefore, the intelligent simulation and control method for grid-connected load of multi-model electric vehicles firstly, for each charging station in the target commercial area, based on the charging information during the target peak time period, utilizes the corresponding edge intelligent agents deployed on the edge terminals to generate local load control information for the charging station during future peak periods. The charging information includes: vehicle model information. Here, by deploying corresponding edge agents for each charging station, accurate generation of local load control information can be achieved based on charging information during the target peak time period. Given the issue of peak-hour congestion at charging stations, load control is only performed using charging information from the peak time period, avoiding the influence of off-peak charging information on peak-hour load management predictions. Furthermore, by deploying edge agents, demand processing for load control during peak hours and subsequent charging simulations can be implemented to strive to meet the maximum load demand for charging under peak charging conditions at charging stations. In addition, by considering vehicle model information, edge agents can accurately generate local load control information. Then, based on the various control information, using the corresponding regional agents for the target commercial area, global load control information can be accurately generated. This control information includes both charging information and local load control information. Here, by setting up regional intelligent agents, based on the control information sent by each edge intelligent agent and combined with the power consumption status of the target commercial area corresponding to the entire area, adaptive control of the load demand of each charging station can be achieved from the perspective of the power consumption status of the entire area and combined with the load demand of each charging station. This ensures that the load demand of each charging station is handled only after meeting the power consumption of the entire area, avoiding the occurrence of insufficient power supply during peak periods due to the high load demand of individual charging stations.Next, for the global load control information, the processing steps are executed using the aforementioned regional agent: First, each load control information in the global load control information is sent to the corresponding charging station to inform each charging station of the load information that each charging station can control based on the regional global electricity consumption perspective. Second, for each charging station, the sending step is executed using the corresponding edge agent: Sub-step 1, based on the corresponding load control information, a charging simulation is performed during the target off-peak period to obtain the simulation results. Here, when the edge agent obtains the corresponding load control information, it simulates whether the basic charging needs during future peak periods can be met by performing a charging simulation during the target off-peak period, in order to avoid charging congestion for electric vehicles during future peak periods as much as possible. Sub-step 2, the load problem set corresponding to the above simulation results is sent to the regional agent so that the regional agent can assess the potential problems of charging processing under the load control information. Third, based on the obtained load problem sets and simulation results, it is determined whether to implement load control. Here, the regional agent, based on the load problem sets and simulation results sent by various edge agents, checks whether load regulation of charging stations needs to be re-implemented from the perspective of overall regional electricity consumption during peak hours. This ensures that each charging station can basically perform charging operations without excessive congestion, while maintaining overall regional electricity balance. In the fourth step, in response to the decision not to implement load regulation, the aforementioned global load regulation information is sent to the power management agent corresponding to the target commercial area to regulate the load of each charging station. Here, by sending global load regulation information to the power management agent, effective management of the power equipment at each charging station is achieved to meet load demands. In summary, based on the charging status of multiple electric vehicle models during the target peak hours, and utilizing the regional agent's overall load scheduling and the load simulation of each edge agent, precise load scheduling of each charging station can be achieved, avoiding power supply shortages. Attached Figure Description

[0012] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and elements are not necessarily drawn to scale.

[0013] Figure 1 This is a flowchart of some embodiments of the intelligent simulation and control method for grid-connected load of multi-model electric vehicles according to the present disclosure; Figure 2 This is a schematic diagram of the structure of some embodiments of the grid-connected load intelligent simulation and control device for multi-model electric vehicles according to the present disclosure; Figure 3 This is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure.

[0014] Wherein, 201-first generation unit; 202-second generation unit; 203-execution unit; 301-processing device; 302-read-only memory (ROM); 303-random access memory (RAM); 304-bus; 305-input / output (I / O) interface; 306-input device; 307-output device; 308-storage device; 309-communication device. Detailed Implementation

[0015] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0016] It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other.

[0017] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0018] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0019] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0020] This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0021] refer to Figure 1 The flowchart 100 illustrates some embodiments of the intelligent simulation and control method for grid-connected load of multi-model electric vehicles according to the present disclosure. The intelligent simulation and control method for grid-connected load of multi-model electric vehicles includes the following steps: Step 101: For each charging station in the target commercial area, based on the charging information during the target peak time period, use the corresponding edge agent deployed on the edge terminal to generate local load control information for the charging station during future peak time periods.

[0022] In some embodiments, for each charging station in the target commercial area, the executing entity (e.g., an intelligent agent system) of the above-described intelligent simulation and control method for grid-connected load of multi-model electric vehicles can generate local load control information for the charging station in future peak periods based on the charging information during the target peak time period and using the corresponding edge intelligent agents deployed on the edge terminals. The charging information includes vehicle model information. The target commercial area can be an area where charging station load control is to be performed. That is, load control is performed on each charging station in the target commercial area to meet charging demand during peak periods and solve the problem of charging congestion. In practice, there are multiple charging stations in the target commercial area. For example, the target commercial area can be a district in the target city. For example, the target commercial area can be Haidian District. A charging station can be a site that supports charging for multiple vehicle models and includes multiple charging stations. The target peak time period can be a period of high traffic congestion on roads or public transportation before the current time. During this period, the charging station also experiences high-volume charging congestion. The target peak time period can be various peak time segments within a month before the current time. For example, if the peak period is 17:00-20:00 on weekdays, the corresponding target peak period is the peak periods on weekdays within the month preceding the current time. In practice, in the transportation sector, public transportation typically experiences heavy congestion between 17:00 and 20:00 on weekdays, while in the power sector, high electricity load and higher electricity prices can occur between 18:00 and 21:00. Therefore, during peak periods, there is a situation of high public transportation flow, high charging flow, and high electricity load. Charging information can be the specific charging status of charging stations during various peak time segments within the month preceding the current time. In practice, charging information includes: vehicle model information, charging flow information, vehicle flow information, and the charging power corresponding to each vehicle model. Vehicle model information can be the vehicle model status of each electric vehicle charging during the target peak period. For example, vehicle model information can include the various car brands being charged, the number of charging sessions for each car brand, and the total charging time. Charging flow information can be the charging status of each charging pile in the charging station during the target peak period. In practice, charging status can include: number of charging sessions, charging power, and charging time. Traffic flow information refers to the traffic volume of electric vehicles traveling in the vicinity of a charging station during the target peak time period. Each charging station has a corresponding deployed edge terminal. The edge terminal supports the execution of some computing tasks to achieve intelligent processing. Each edge terminal has a deployed edge agent. The edge agent deployed at the charging station is a lightweight AI system deployed locally at the charging station, with real-time data processing and autonomous decision-making capabilities. By completing key computing tasks at the edge, it achieves millisecond-level response, reduces cloud dependence, and improves security and operational efficiency.Primarily used to transform the traditional "passive response" model into a closed loop of "local real-time perception-analysis-decision-execution," it is particularly suitable for scenarios requiring rapid handling of safety incidents and optimization of charging strategies. In practice, edge agents can have the following functions: real-time risk prevention and control (e.g., completing anomaly detection and localized handling within 100 milliseconds (such as over-temperature, over-current, and smoke recognition), avoiding the communication delay risks caused by relying on the cloud), lightweight AI inference (e.g., deploying pruned / quantized models for predicting battery status, identifying device faults, and generating local load control), and charging simulation. Specifically, edge agents can connect to charging piles, meters, and other devices through multi-protocol adapters, compatible with national standards and manufacturer-specific protocols. Edge agents can also execute local rule engines (e.g., determining offline status after 10 consecutive minutes without a heartbeat) or lightweight AI models (e.g., charging load prediction). The models deployed by the edge agent can be deployed for each function. For example, the model deployed by the edge agent can be a multimodal large model, which can be a conventional architecture model supporting multimodal inputs and outputs. The multimodal large model is pre-trained on training datasets corresponding to each function of the charging station. Future peak time periods can be any peak time period within a time span following the current time. For example, future peak time periods can be any peak time period on any weekday within a month following the current time. For instance, for the current time being April 30th, the corresponding future peak time periods could be the peak time periods on any weekday from May 1st to May 31st, such as 17:00-20:00 on any weekday. Local load control information can be the load control content within the corresponding area of ​​the charging station. In practice, local load control information can be the load control content under each peak time period within the future peak time period. In practice, local load control information is generated by the edge agent based on the charging information it acquires, representing the charge demand during the future peak time period.

[0023] As an example, firstly, prompts are generated based on charging information to predict local load regulation information during future peak periods. Then, the generated prompts are input into a multimodal large model in the edge agent to obtain local load regulation information.

[0024] In some optional implementations of certain embodiments, the aforementioned execution entity can generate local load regulation information for the charging station during future peak periods based on charging information within the target peak time period, utilizing edge agents deployed on corresponding edge terminals, including the following steps: The first step is to obtain the charging station influence area division information sent by the aforementioned regional intelligent agent. This information characterizes the charging influence area corresponding to each charging station. Here, the charging influence area can be the area centered on the actual location of the charging station that may affect electric vehicles coming to charge at the station. For example, for charging station A, electric vehicles within a 1-kilometer radius of charging station A frequently use charging station A for charging.

[0025] It should be noted that the charging station impact area division information here includes key-value pairs for each charging station. These key-value pairs include the charging station identifier and the corresponding area range information. In practice, the charging station impact area division information can be set based on historical experience. For example, the division could be a 1-kilometer circular area centered on the charging station as its charging impact area. Alternatively, the impact area can be determined by a regional intelligent agent based on the charging status of vehicles at each charging station, or it can be manually divided by relevant personnel on the substation processing page.

[0026] The second step involves retrieving the corresponding charging sub-information from the charging information, based on the aforementioned charging station influence area division information. This charging sub-information includes: charging vehicle model ratio information, traffic flow information within the corresponding influence area, and charging power consumption for different vehicle models. The traffic flow information within the corresponding influence area can be the traffic flow status within the charging station's influence area at various time periods. For example, traffic flow information could include: traffic flow over 30 days. The charging vehicle model ratio information can be the ratio of different vehicle models undergoing charging at various time periods. For example, the charging vehicle model ratio information could include: the ratio of different vehicle models undergoing charging over 30 days. In practice, different vehicle models can include: private passenger cars, ride-hailing vehicles, and logistics vehicles. Charging power consumption can be the charging power of the electric vehicle corresponding to the vehicle model.

[0027] As an example, firstly, determine the influence area information corresponding to the charging station in the above-mentioned influence area division information. Then, extract the charging sub-information corresponding to the influence area information from the charging information.

[0028] The third step, based on the aforementioned charging electronic information, utilizes an edge intelligence agent to execute the first generation step: Sub-step 1: Based on the aforementioned charging information, generate the first future charging information for the aforementioned future peak time period. The first future charging information includes: the future charging vehicle model ratio information and future traffic flow. Specifically, the first future charging information can be the predicted charging situation during the future peak time period. The future charging vehicle model ratio can be the proportion of each electric vehicle model charging during the future peak time period.

[0029] As an example, firstly, predictive prompts are generated to forecast future charging information during peak periods based on charging information. Then, the predicted prompts are input into a multimodal large model deployed by an edge agent to obtain the first future charging information.

[0030] Sub-step 2 involves determining whether the target holiday period exists within the aforementioned future peak timeframes. In practice, the target peak timeframes are defined as the peak hours during the weekdays within the month following the current date. For example, if the current date is April 30th, the corresponding future peak timeframes could be the peak hours during the weekdays of May. The current and future peak timeframes include the peak hours corresponding to the Labor Day holiday. Because charging demand during holidays is often higher than daily charging demand, it is necessary to determine whether the credit limit exists during the target holiday period.

[0031] Sub-step 3: In response to the confirmation of existence, obtain the historical charging station information sequence corresponding to the target holiday period. The historical charging station information sequence can be charging station information within various historical time periods. The historical charging station information sequence is sorted in ascending order of the corresponding historical time. Each historical time period refers to the holiday periods of all years preceding the current year. For example, for the target holiday period of May 1st-May 5th, 2026, the corresponding historical times are May 1st-May 5th, 2023, May 1st-May 5th, 2024, and May 1st-May 5th, 2025. The corresponding historical charging station information sequences can be the charging station information for May 1st-May 5th, 2023, May 1st-May 5th, 2024, and May 1st-May 5th, 2025.

[0032] Sub-step 4: Based on the aforementioned historical charging information sequence, the first future charging information is weighted and adjusted to obtain the second future charging information. The weighting adjustment can be a weighted average of the charging information predicted for holidays in historical years and the charging information predicted for holidays in the current year.

[0033] As an example, firstly, the aforementioned executing entity can predict the first future holiday charging information (i.e., the charging situation during the target holiday period in the future) based on the historical charging sub-information sequence and using a multimodal large model deployed by the edge agent. Then, it filters out the second future holiday charging information corresponding to the target holiday period from the first future holiday sub-information. Next, it performs information weighting on the first and second future holiday charging information to obtain future holiday weighted charging information. Finally, it replaces the second future holiday charging information in the first future holiday sub-information with the future holiday weighted charging information to obtain the second future holiday sub-information.

[0034] As another example, the aforementioned executing entity can use prompt word technology to weight and adjust the aforementioned first future charging electronic information based on the aforementioned historical charging electronic information sequence to obtain the second future charging electronic information.

[0035] The fourth step is to generate local load control information based on the aforementioned second future charging electronic information.

[0036] As an example, the aforementioned execution entity can generate local load control information based on the aforementioned second future charging electronic information and by utilizing the load control and conversion function corresponding to the edge agent.

[0037] In some optional implementations of certain embodiments, generating the first future charging information for the future peak time period based on the charging information includes the following steps: The first step involves preprocessing the aforementioned charging vehicle ratio information, traffic flow information, and various charging power consumption mapping information to obtain first preprocessed information, second preprocessed information, and various third preprocessed information. The charging power consumption mapping information represents the mapping relationship between charging power consumption and vehicle type. Preprocessing may include: unit conversion, information supplementation, standardization, and vectorization. The first preprocessed information can be the result of preprocessing the charging vehicle ratio information. The second preprocessed information can be the result of preprocessing the traffic flow information. The third preprocessed information can be the result of preprocessing the charging power consumption mapping information.

[0038] The second step involves inputting the aforementioned first preprocessed information, second preprocessed information, and each of the third preprocessed information into their respective encoding layers to obtain the first encoding result, second encoding result, and third encoding result. Specifically, the charging vehicle ratio information has a corresponding encoding layer, a network layer specifically used for encoding and dimensionality reduction of this information. Similarly, traffic flow information has a corresponding encoding layer, a network layer specifically used for encoding and dimensionality reduction of this information. Charging power consumption mapping information also has a corresponding encoding layer, a network layer specifically used for encoding and dimensionality reduction of this information. In practice, each encoding layer can be a Transformer encoding layer. The first, second, and third encoding results can all be information in vector form, and the corresponding vector dimensions are the same.

[0039] The third step involves time-aligning and feature-fusion of the first, second, and third encoding results to obtain an encoding fusion information sequence. This encoding fusion information is the fusion result of each encoded content corresponding to the charging vehicle ratio, vehicle flow, and charging power consumption mapping information at the same time.

[0040] As an example, firstly, the target peak time period is divided into days, resulting in a time period sequence. Each time period corresponds to one day. Then, for each time period in the time period sequence, the first step is to extract the corresponding coded sub-results from the first, second, and third coded results. The second step is to concatenate the coded sub-results to obtain the coded fusion information.

[0041] The fourth step involves dividing the target peak time period into simultaneous periods to obtain a set of peak time segment sequences. The peak time segments within this sequence are all time segments within the same period. This simultaneous period division can be achieved by grouping time segments within the same week into one category, resulting in a peak time segment sequence. Each time segment within this sequence is a time segment within the same week and is ordered chronologically by date. For example, if the target peak time period is April, and the peak time segment sequence corresponds to Monday, then the peak time segment sequence could be "April 4th, April 11th, April 18th, April 25th". The set of peak time segment sequences can include: the peak time segment sequence corresponding to Monday, Tuesday, Wednesday, Thursday, Friday, Saturday, and Sunday.

[0042] Fifth, for each peak time segment sequence, perform the second generation step: Sub-step 1 involves selecting the coded fusion information sub-sequences corresponding to the peak time segment sequences from the aforementioned coded fusion information sequences. There is a one-to-one correspondence between the peak time frequency bands in the peak time segment sequences and the coded fusion information in the coded fusion information sub-sequences.

[0043] Sub-step 2 involves inputting the aforementioned encoded fusion information sub-sequence into the temporal decoding layer to obtain the initial future charging vehicle ratio information. The temporal decoding layer can be a decoding layer that decodes the temporal input data. In practice, the temporal decoding layer can be a decoding layer constructed from multiple cascaded deconvolutional layers. The initial future charging vehicle ratio information can be the charging vehicle ratio status under the initially generated peak time segment sequence. That is, the initial future charging vehicle ratio information here is the vehicle ratio information to be further determined.

[0044] Sub-step 3: Based on the initial future charging vehicle ratio information mentioned above, generate future charging vehicle ratio sub-information under the peak time segment sequence mentioned above.

[0045] As an example, the initial future charging vehicle ratio information can be checked for a ratio range to determine whether it falls within a normal ratio range. If it does, the initial future charging vehicle ratio information is determined as the future charging vehicle ratio sub-information. If it does not, the initial future charging vehicle ratio information is adjusted according to the normal ratio range so that the final adjusted result falls within the normal ratio range, thus obtaining the future charging vehicle ratio sub-information.

[0046] The sixth step is to generate future charging vehicle proportion information based on the obtained future charging vehicle proportion sub-information sequence.

[0047] As an example, the aforementioned implementing entity can directly determine the future charging vehicle proportion sub-information sequence as the future charging vehicle proportion information.

[0048] In some optional implementations of certain embodiments, after generating the future charging vehicle ratio sub-information under the peak time segment sequence based on the initial future charging vehicle ratio information, the method further includes: The first step is to filter out the encoded information subsequences corresponding to the peak time segment sequences from the second encoding results. There is a one-to-one correspondence between the peak time segments in the peak time segment sequence and the encoded information in the encoded information subsequences. That is, the encoded information subsequences here represent the traffic flow information for each peak time segment.

[0049] The second step involves inputting the aforementioned encoded information subsequence into the time-series decoding layer to obtain the initial future sub-vehicle traffic flow. This initial future sub-vehicle traffic flow can be the predicted traffic flow situation for the future peak period corresponding to the peak time segment sequence.

[0050] The third step is to obtain the previous peak time segment sequence and the next peak time segment sequence for the previous period, corresponding to the above peak time segment sequence. Details will not be elaborated further.

[0051] The fourth step involves generating a first-period proximity relationship based on the previous peak time segment sequence corresponding to the previous coded information subsequence, the next peak time segment sequence corresponding to the next coded information subsequence, and the aforementioned coded information subsequences. This first-period proximity relationship can be the association between various peak time segment sequences within the same period (e.g., various time segment sequences within the same peak time period on different weeks). Based on this first-period proximity relationship, traffic flow changes in future peak time segments within the same period can be determined.

[0052] Fifth, based on the proximity relationship of the first period mentioned above, adjust the initial future sub-vehicle flow to obtain the future sub-vehicle flow.

[0053] As an example, firstly, based on the proximity relationship of the first period, the estimated range of future sub-vehicle traffic flow under the peak time segment sequence is determined. Then, based on the estimated range of future sub-vehicle traffic flow, the initial future sub-vehicle traffic flow is adjusted (i.e., if the initial future sub-vehicle traffic flow is higher than the maximum value corresponding to the estimated range of future sub-vehicle traffic flow, the maximum value corresponding to the estimated range of future sub-vehicle traffic flow is determined as the future sub-vehicle traffic flow; if the initial future sub-vehicle traffic flow is lower than the minimum value corresponding to the estimated range of future sub-vehicle traffic flow, the minimum value corresponding to the estimated range of future sub-vehicle traffic flow is determined as the future sub-vehicle traffic flow), to obtain the future sub-vehicle traffic flow.

[0054] Optionally, the steps also include: Based on the obtained future sub-vehicle flow sequence, generate future vehicle flow.

[0055] As an example, the aforementioned implementing entity can directly determine the future sub-vehicle flow sequence as the future vehicle flow.

[0056] In some optional implementations of certain embodiments, generating the future charging vehicle ratio sub-information under the peak time segment sequence based on the initial future charging vehicle ratio information includes the following steps: The first step is to obtain the previous peak time segment sequence and the next peak time segment sequence for the previous period, corresponding to the peak time segment sequence mentioned above. Details will not be elaborated further.

[0057] The second step involves generating a second-period proximity relationship based on the previous peak time segment sequence corresponding to the previous coded fusion information sub-sequence, the next peak time segment sequence corresponding to the next coded fusion information sub-sequence, and the aforementioned coded fusion information sub-sequences. This second-period proximity relationship can represent the correlation between various coded fusion information sub-sequences, indirectly characterizing the changing relationship of charging power consumption mapping information across different time periods within the same period.

[0058] The third step involves adjusting the initial future charging vehicle ratio information based on the proximity relationships of the second period, thereby obtaining the future charging vehicle ratio sub-information. The specific implementation method will not be elaborated further; please refer to the generation of future sub-vehicle traffic flow.

[0059] Step 102: Based on the various control information, generate global load control information using the regional intelligent agents corresponding to the target commercial areas mentioned above.

[0060] In some embodiments, the aforementioned executing entity can generate global load control information based on various control information and utilizing the corresponding regional intelligent agent of the target commercial area. The control information includes charging information and local load control information. The regional intelligent agent can be an intelligent agent that processes various matters within the target commercial area. The regional intelligent agent is a distributed AI decision-making hub deployed for a specific geographical or administrative region (e.g., urban area, transportation network, industrial park). Through cross-system data fusion and strategy-level optimization, the regional intelligent agent achieves dynamic scheduling and collaborative governance of regional resources. Unlike edge intelligent agents (focusing on real-time control of single-point devices), regional intelligent agents emphasize multi-source data integration, long-term strategy generation, and cross-departmental collaboration. Application scenarios for regional intelligent agents include traffic flow control, energy load allocation, and emergency response. In the charging field, the regional intelligent agent can make decisions based on various control information sent by edge intelligent agents to generate global load control information from the perspective of regional peak electricity consumption. In practice, regional intelligent agents can also be deployed with large models to support various intelligent decision-making processes. The corresponding large-scale model not only supports intelligent processing of various aspects within the charging field but also supports intelligent processing in other fields. Therefore, the large-scale model can be trained using task training datasets from various tasks within different fields. The large-scale model can be a heavyweight multimodal model. A multimodal large-scale model can be a large model with a conventional architecture deployed on a private server, supporting multimodal content input and output. Among these, the global load regulation information can be the overall charge regulation content corresponding to each charging station.

[0061] As an example, firstly, based on various charging information, a large model deployed by regional agents is used to generate initial global control information. Then, the initial global control information is compared with the local load control information to obtain the comparison results. If the differences in the comparison results are small, the local load control information is combined to obtain the global load control information. If the differences in the comparison results are large, the initial global load control information is determined as the global load control information.

[0062] In some optional implementations of certain embodiments, the above-mentioned generation of global load control information based on various control information and utilizing the regional intelligent agent corresponding to the target commercial area includes the following steps: The first step is to generate global charging information for the target commercial area during the target peak time period based on the charging information from the aforementioned control information. This global charging information can represent the global charging status of the target commercial area.

[0063] As an example, the various charging information can be combined to obtain global charging information.

[0064] The second step involves generating initial global load control information for the target commercial area based on the aforementioned global charging information and historical global charging information sequence. The historical global charging information sequence can represent the global charging status at various historical time points. The historical global charging information entries are ordered chronologically. The initial global load control information can be the preliminary control measures for regulating the overall power supply load within the target commercial area.

[0065] As an example, based on the aforementioned global charging information and historical global charging information sequences, the temporal prediction capabilities of the regional agent can be used to generate future global charging information sequences. Then, based on the sequence differences between the current global charging information sequence and the future global charging information sequence, initial global load control information can be generated.

[0066] The third step involves determining the initial global load control information as the global load control information if the control difference between the initial global load control information and the control information of each local load is less than the target difference value. The target difference can be a pre-determined threshold used to measure whether the control difference is significant. The target difference value can be determined based on historical experience.

[0067] Fourth, in response to the existence of at least one local load control information corresponding to a control difference not less than the aforementioned target difference value, the following verification steps are performed for each local load control information: Sub-step 1: Select global load control sub-information from the above global load control information that exists in the same sub-region as the above local load control information.

[0068] Sub-step 2 involves performing multi-faceted verification of the aforementioned global load control sub-information to obtain verification results. This multi-faceted verification can determine whether the control scheme corresponding to the global load control sub-information meets the power supply requirements of the target commercial area. For example, multi-faceted verification may include: whether the load supply is sufficient during peak hours, whether it affects electricity consumption in residential areas, and whether it affects electricity consumption in areas with critical equipment. The verification result can be one of the following: indicating a problem with the global load control sub-information, or indicating that the global load control sub-information does not have a problem. Here, the multi-faceted verification can be set and executed by the regional intelligent agent based on the load requirements of the target commercial area.

[0069] Sub-step 3: In response to the above verification result indicating that there is no problem, the above global load control sub-information is determined as the target global load control sub-information.

[0070] Sub-step 4: In response to the problem indicated by the above verification results, the adjustment information corresponding to the above local load control information is determined as the target global load control sub-information.

[0071] The fifth step is to adjust the initial global load control information based on at least one target global load control sub-information to obtain the global load control information.

[0072] As an example, the aforementioned executing entity can replace the corresponding adjustment sub-information in the initial global load control information with at least one target global load control sub-information to obtain global load control information.

[0073] Step 103: For global load control information, utilize the aforementioned regional agent to execute the following processing steps: Step 1031: Send each load control information in the global load control information to the corresponding charging station.

[0074] In some embodiments, the aforementioned executing entity can send each load control information in the global load control information to the corresponding charging station. There is a one-to-one correspondence between the load control information in each load control information and the charging station in each charging station. The load control information can be the prediction result of load control prediction for the charging station from the global load control information.

[0075] Step 1032: For each charging station, execute the sending step using the corresponding edge agent: Step 10321: Based on the corresponding load control information, perform charging simulation during the target off-peak period to obtain simulation results.

[0076] In some embodiments, the aforementioned execution entity can perform charging simulation during a target off-peak period based on corresponding load control information to obtain simulation results. The corresponding load control information can be the load control information for the charging station. The target off-peak period can be the period during which the charging flow at the charging station is lower than a target flow value in the current time. Specifically, this applies to a day in general. The target off-peak period can be the early morning hours (e.g., 1 AM to 4 AM) within the current date. The charging simulation can simulate a scenario where the charging station charges under the load supply corresponding to the load control information during future peak periods. The simulation results can be the charging status of the charging station during the simulation process. For example, the simulation results may include: mild charging congestion during time period A and moderate charging congestion during time period B.

[0077] As an example, firstly, the aforementioned execution entity can enhance the charging information corresponding to the target peak time period to obtain enhanced charging information. Then, based on the existing charging information and enhanced charging information, a charging simulation scenario for future peak time periods is deployed. Next, based on the corresponding load control information, the supplied charge in the charging simulation scenario is adaptively adjusted to obtain a charging simulation scenario under charge supply conditions. Finally, using the large model corresponding to the edge agent, the charging flow process corresponding to the charging simulation scenario is executed, and the charging state sequence is recorded to obtain the simulation results. The charging state can be the charging status of the charging station at various time points within the future peak time period.

[0078] In addressing the technical problems mentioned in the background section, the application scenario of charging simulation often presents the following challenge: insufficient simulation resources may arise during the simulation process. To meet the specific requirements of this application scenario—achieving comprehensive charging simulation—we have decided to adopt the following solution: In some optional implementations of certain embodiments, the above-mentioned charging simulation is performed during the target off-peak period based on the corresponding load control information to obtain simulation results, including: The first step is to obtain the charging settings information for the future peak time period set on the charging simulation page corresponding to the aforementioned edge agent. The charging simulation page can be a simulation page for setting charging simulation-related parameters. The charging settings information can be the simulation settings information for executing the charging simulation during the future peak time period. In practice, the charging settings information may include: time period, execution terminal information, and data description information.

[0079] The second step involves generating spatiotemporal simulation data for charging and traffic flow for multiple electric vehicle models, based on the aforementioned charging settings and charging information. The charging spatiotemporal simulation data can be charging-related, while the traffic flow spatiotemporal simulation data can be traffic flow-related. Multiple electric vehicle models can refer to electric vehicles of various types.

[0080] As an example, reinforcement learning can be used to generate spatiotemporal simulation data of charging and spatiotemporal simulation data of traffic flow for electric vehicles of various models, based on the aforementioned charging settings and charging information.

[0081] The third step is to determine the off-peak charging information of the aforementioned edge intelligence agents within the target off-peak time period. This off-peak charging information can be the charging status during the target off-peak time period. In other words, the off-peak charging information can be the load supply status during the target off-peak time period.

[0082] The fourth step involves determining multiple neighboring agents corresponding to the aforementioned edge agent, based on the charging station occupancy status indicated by the aforementioned low-peak charging information meeting the target occupancy condition. Simulation resource consumption information is then determined according to the aforementioned spatiotemporal simulation data of charging, the aforementioned spatiotemporal simulation data of traffic flow, and the aforementioned load control information. Specifically, each neighboring agent corresponds to a charging station occupancy status indicated by the low-peak charging information that does not meet the aforementioned target occupancy condition. Charging station occupancy status can be the computing resources used by the charging station during off-peak charging periods. The target occupancy condition can be a charging station occupancy status not less than a target percentage (a percentage value used to measure whether the charging station has sufficient computing resources). Multiple neighboring agents can be multiple agents belonging to multiple neighboring charging stations that are adjacent in terms of charging station spacing. Simulation resource consumption information can be the estimated amount of resources to be consumed during the simulation.

[0083] As an example, the simulation resource consumption information corresponding to the aforementioned charging spatiotemporal simulation data, traffic flow spatiotemporal simulation data, and load control information can be determined based on the mapping relationship. The mapping relationship can characterize the mapping relationship between the data volume corresponding to the charging spatiotemporal simulation data, the data volume corresponding to the traffic flow spatiotemporal simulation data, the data volume corresponding to the load control information, and the simulation resource consumption.

[0084] Fifth, based on the simulated resource consumption information and the edge terminal layout distance between each neighboring agent and the edge agent, select the target neighboring agent from among the multiple neighboring agents to perform the simulation. The edge terminal layout distance can be the distance between the charging station corresponding to the neighboring agent and the charging station corresponding to the edge agent. The target neighboring agent can be the neighboring agent with the highest weighted value corresponding to distance and remaining resources.

[0085] Step 6: Package the above-mentioned charging spatiotemporal simulation data, traffic flow spatiotemporal simulation data, load control information, and simulation resource consumption information and send them to the edge terminal corresponding to the target neighboring intelligent agent to execute the following steps: Sub-step 1 involves simulating the charging spatiotemporal simulation data, traffic flow spatiotemporal simulation data, and the corresponding load of the load control information, within the target low-peak time period, to obtain an initial charging state sequence at each future time within the future peak time period. This initial charging state sequence can be the supply state after charging is initiated at each future time.

[0086] Step 7: Obtain the initial charging state sequence sent by the target's neighboring agents.

[0087] Step 8: In response to the initial charging state sequence being verified as correct, the initial charging state sequence is determined as the charging state sequence. The verification of the initial charging state sequence can be to check whether there are any abnormalities in the charging state.

[0088] Step nine involves generating a 3D charging simulation model and a summary of the charging simulation results based on the aforementioned charging state sequence. The 3D charging simulation model can be a digital twin model built upon the corresponding 3D model of the charging station, reflecting the load supply status. The summary of the charging simulation results is a compilation of the various states corresponding to the charging state sequence.

[0089] Step 10: Determine the above 3D charging simulation model and the above charging simulation summary results as the simulation results.

[0090] The aforementioned "steps one through ten" serve as another inventive point. During the charging simulation process, the resource consumption information can be determined through the charging settings and charging information. Secondly, the resource status and layout distance of each neighboring agent favored by low-peak charging information are determined to select neighboring agents that support resource supply, thereby ensuring sufficient resources during the charging simulation. Finally, a 3D charging simulation model and a summary of the charging simulation results are generated to achieve diverse displays of the charging simulation.

[0091] Step 10322: Send the load problem set corresponding to the above simulation results to the regional agent.

[0092] In some embodiments, the executing entity may send the load problem set corresponding to the simulation results to the regional agent. The load problem may be a charging problem arising from load supply issues in the simulation results. The load problem can characterize various charging problems that may occur during charging processing using load control information. For example, a load problem may be moderate charging congestion due to limited charge supply during time period B, or severe charging congestion due to limited charge supply during time period C. The load problem set may be an empty set or at least one load problem.

[0093] Step 1033: Based on the obtained load problem sets and simulation results, determine whether to perform load regulation.

[0094] In some embodiments, the aforementioned executing entity can determine whether to perform load regulation based on the obtained sets of load problems and simulation results. There is a one-to-one correspondence between the load problem sets in each load problem set and the charging stations in each charging station. Similarly, there is a one-to-one correspondence between the simulation results in each simulation result and the charging stations in each charging station.

[0095] As an example, if no severe charging congestion event is found among the charging congestion events corresponding to each load problem in each load problem set, it is determined that load regulation will not be performed. If severe charging congestion events are found among the charging congestion events corresponding to each load problem in each load problem set, it is determined that load regulation will be performed.

[0096] Step 1034: In response to the determination not to execute, the above-mentioned global load control information is sent to the power management and control agent corresponding to the target commercial area to control the load corresponding to each of the above-mentioned charging stations.

[0097] In some embodiments, in response to a decision not to execute, the aforementioned executing entity can send the aforementioned global load control information to the power management intelligent agent corresponding to the target commercial area, so as to regulate the load corresponding to each of the aforementioned charging stations. The power management intelligent agent corresponding to the target commercial area can be an intelligent agent that performs comprehensive power management of the city corresponding to the target commercial area. That is, based on the power management intelligent agent, it can assist the city corresponding to the target commercial area in power management. The power management intelligent agent can be deployed in the core area of ​​the city corresponding to the target commercial area to facilitate the acquisition of power load control information in various areas and reduce communication transmission. The city power management intelligent agent is an AI decision-making hub covering the entire city's power grid and integrating data from the entire chain of power generation, grid, load, and storage. Through integrated scheduling of the main and distribution networks, second-level fault self-healing, and cross-departmental resource collaboration, it realizes a paradigm upgrade of the city's power grid from "passive response" to "active regulation." The large model corresponding to the power management intelligent agent can be a large model used to handle various power tasks in the power field.

[0098] As an example, the power management and control intelligent agent can generate power control instructions for each charging station within a target commercial area based on global load control information. By executing these instructions, the load corresponding to each charging station can be controlled.

[0099] In some optional implementations of certain embodiments, after step 1034, the steps further include: The first step, in response to the determination execution, is to perform the following determination steps for each charging station: Sub-step 1: Based on the simulation results corresponding to the charging stations mentioned above, perform problem verification on the load problem set to obtain the problem verification results. The problem verification results can be one of the following: a result indicating failure to pass verification, or a result indicating successful verification.

[0100] As an example, the aforementioned execution entity can utilize a regional agent to identify the set of state anomaly information corresponding to each execution state in the simulation results. Each state anomaly information corresponds to at least one execution state and a specific time point. Then, the set of state anomaly information and the set of load problems are compared to determine if their occurrence times are the same and if there is a correlation (i.e., whether there is a causal relationship between the state anomaly information and the load problems), thus obtaining the problem verification results.

[0101] Sub-step 2: In response to the above problem verification result indicating that the verification failed, the local load control information corresponding to the above charging station is adjusted according to the above load problem set to obtain the target local load control information.

[0102] As an example, firstly, target load problems that do not correspond to state anomaly information in the state anomaly information set are selected from the load problem set, resulting in at least one load problem. Then, the at least one load problem is removed from the load problem set, resulting in a remaining load problem set. This at least one load problem can be a problem that the edge agent considers potentially existing, but the regional agent considers not to exist. Finally, based on the time period set in the remaining load problem set, the local load control information corresponding to the charging station is adjusted to obtain target local load control information. Here, the load supply set corresponding to the time period set in the local load control information can be boosted to solve the charging congestion problem occurring in each time period of the time period set. It should be noted that the boost amount corresponding to this load supply can be determined based on the severity of the charging congestion corresponding to the load problem.

[0103] As another example, firstly, the aforementioned implementing entity can filter out a set of remaining load problems and at least one load problem from the set of load problems. The remaining load problem set is tagged with "problematic," and the at least one load problem is tagged with "suspected problem." Then, based on the tagged remaining load problem set and the tagged at least one load problem, the local load control information corresponding to the aforementioned charging station is adjusted using prompt word technology to obtain the target local load control information.

[0104] Sub-step 3: In response to the above problem verification result characterization passing the verification, the local load control information is determined as the target local load control information.

[0105] The second step is to combine the obtained local load control information of each target to obtain global load control information.

[0106] The third step is to use the global load adjustment information as the global load adjustment information and continue to execute the above processing steps.

[0107] In addressing the technical problems mentioned in the background section, and considering the application scenario of load regulation for a target commercial area, it is crucial to fully consider the load supply of each area during peak hours to prevent the predicted power load of the target commercial area from exceeding the overall load supply requirements. This often leads to the following technical problem: the predicted global load regulation information for the target commercial area may exceed the load supply at the regional level. Given the following requirements for this application scenario: achieving reasonable handling of load supply, we have decided to adopt the following solution: In some optional implementations of certain embodiments, after step 103, the steps further include: The first step is to use the power management intelligent agent to perform the following information sending steps: Sub-step 1 involves obtaining global load control information for each region and peak electricity load information for each region. The peak electricity load information can be the electricity load status during peak hours. In practice, peak electricity load information can include both peak hour information and usage load information. Each region can be any area set up within the city. That is, each region can include target commercial areas.

[0108] Sub-step 2 involves determining the electricity load ratio for each region within the target peak time period based on the peak load information for each region. The electricity load ratio can be the supply ratio for electricity load in each region, representing the specific power supply situation in each region.

[0109] Sub-step 3: Based on the aforementioned electricity load ratio, determine the electricity load information corresponding to the target commercial area. Here, the user load information can be the load status supplied to the target commercial area. That is, the user load information here is determined from the perspective of each area, representing the load status supported by the target commercial area.

[0110] As an example, electricity load information can be determined by multiplying the total load supply and the user load ratio.

[0111] Sub-step 4: Based on the above electricity load information, determine whether to adjust the global load control information corresponding to the above target commercial area.

[0112] As an example, the load magnitude is compared between the electricity load information and the total load corresponding to the global load control information to determine whether the global load control information corresponding to the target commercial area needs to be adjusted. That is, if the total load is determined to be greater than the electricity load information, it indicates that adjustment is required.

[0113] Sub-step 5: In response to the confirmation to make adjustments, the available total global load information for each charging station within the target commercial area is sent to the regional agent. The total global load information can be the total load status supported within the target commercial area.

[0114] The second step involves using the aforementioned regional intelligent agent to perform the following steps: Sub-step 1: Based on the aforementioned total global load information, the total global load control information corresponding to the target commercial area is further processed to obtain the total global load re-control information. The total load corresponding to the total global load re-control information is no greater than the electricity load information.

[0115] Sub-step 2: Use the global load re-regulation information as the global load regulation information and continue to execute the above processing steps.

[0116] Optionally, the above determination of whether to perform load regulation based on the obtained sets of load problems and simulation results includes: Based on the aforementioned load problem sets, simulation results, and global total load information, it is determined whether to perform load regulation. The specific implementation details will not be elaborated further.

[0117] The above-mentioned content, as another inventive point, solves another technical problem: "The predicted global load control information for the target commercial area does not meet the load supply at the regional level, resulting in invalid predictions." Based on this, firstly, the electricity load ratio of each region corresponding to the target commercial area is determined. Then, based on the electricity load ratio, the electricity load information corresponding to the target commercial area is determined. Next, based on the electricity load information and the global load control information, it is determined whether to adjust the load for the target commercial area. Finally, the global total load information is sent to the regional agent to further adjust the global load control information, ensuring that the load control for the target commercial area is less than the global total load information. Secondly, it is possible to determine whether to perform load control based on the global total load information. In summary, accurate generation of load control for the target commercial area can be achieved.

[0118] The above-described embodiments of this disclosure have the following beneficial effects: Through the intelligent simulation and control method for grid-connected load of multi-model electric vehicles in some embodiments, based on the charging status of multi-model electric vehicles during the target peak time period, and utilizing the overall load scheduling of regional intelligent agents and the load simulation of each edge intelligent agent, the load of each charging station can be accurately scheduled, avoiding the problem of insufficient power supply. Specifically, the reason for the inaccurate load control of charging stations during the relevant peak time period is that: load prediction based on load prediction models has poor initiative, and load prediction based on full data cannot effectively solve the problem of electric vehicle charging congestion during peak time periods. Moreover, during peak time periods, it is not possible to effectively control the power load of charging stations based on the overall power consumption of the region, which may lead to insufficient power supply for other electrical equipment during peak time periods. Based on this, the intelligent simulation and control method for grid-connected load of multi-model electric vehicles firstly, for each charging station in the target commercial area, based on the charging information during the target peak time period, uses the corresponding edge intelligent agents deployed on the edge terminals to generate local load control information for the charging station during future peak time periods. The charging information includes: vehicle model information. Here, by deploying corresponding edge agents for each charging station, accurate generation of local load control information can be achieved based on charging information during the target peak time period. Given the issue of peak-hour congestion at charging stations, load control is only performed using charging information from the peak time period, avoiding the influence of off-peak charging information on peak-hour load management predictions. Furthermore, by deploying edge agents, demand processing for load control during peak hours and subsequent charging simulations can be implemented to strive to meet the maximum load demand for charging under peak charging conditions at charging stations. In addition, by considering vehicle model information, edge agents can accurately generate local load control information. Then, based on the various control information, using the corresponding regional agents for the target commercial area, global load control information can be accurately generated. This control information includes both charging information and local load control information. Here, by setting up regional intelligent agents, based on the control information sent by each edge intelligent agent and combined with the power consumption status of the target commercial area corresponding to the entire area, adaptive control of the load demand of each charging station can be achieved from the perspective of the power consumption status of the entire area and combined with the load demand of each charging station. This ensures that the load demand of each charging station is handled only after meeting the power consumption of the entire area, avoiding the occurrence of insufficient power supply during peak periods due to the high load demand of individual charging stations.Next, for the global load control information, the processing steps are executed using the aforementioned regional agent: First, each load control information in the global load control information is sent to the corresponding charging station to inform each charging station of the load information that each charging station can control based on the regional global electricity consumption perspective. Second, for each charging station, the sending step is executed using the corresponding edge agent: Sub-step 1, based on the corresponding load control information, a charging simulation is performed during the target off-peak period to obtain the simulation results. Here, when the edge agent obtains the corresponding load control information, it simulates whether the basic charging needs during future peak periods can be met by performing a charging simulation during the target off-peak period, in order to avoid charging congestion for electric vehicles during future peak periods as much as possible. Sub-step 2, the load problem set corresponding to the above simulation results is sent to the regional agent so that the regional agent can assess the potential problems of charging processing under the load control information. Third, based on the obtained load problem sets and simulation results, it is determined whether to implement load control. Here, the regional agent, based on the load problem sets and simulation results sent by various edge agents, checks whether load regulation of charging stations needs to be re-implemented from the perspective of overall regional electricity consumption during peak hours. This ensures that each charging station can basically perform charging operations without excessive congestion, while maintaining overall regional electricity balance. In the fourth step, in response to the decision not to implement load regulation, the aforementioned global load regulation information is sent to the power management agent corresponding to the target commercial area to regulate the load of each charging station. Here, by sending global load regulation information to the power management agent, effective management of the power equipment at each charging station is achieved to meet load demands. In summary, based on the charging status of multiple electric vehicle models during the target peak hours, and utilizing the regional agent's overall load scheduling and the load simulation of each edge agent, precise load scheduling of each charging station can be achieved, avoiding power supply shortages.

[0119] Further reference Figure 2 As an implementation of the methods shown in the above figures, this disclosure provides some embodiments of a grid-connected load intelligent simulation and control device for multi-model electric vehicles. These device embodiments are similar to... Figure 1 Corresponding to the method embodiments shown, the intelligent simulation and control device for grid-connected load of multi-model electric vehicles can be specifically applied to various electronic devices.

[0120] like Figure 2As shown, a grid-connected load intelligent simulation and control device 200 for multi-model electric vehicles includes: a first generation unit 201, a second generation unit 202, and an execution unit 203. The first generation unit 201 is configured to generate local load control information for each charging station in the target commercial area during future peak hours, based on charging information during the target peak time period and utilizing the corresponding edge agent deployed on the edge terminal. The charging information includes vehicle model information. The second generation unit 202 is configured to generate global load control information based on the control information and utilizing the corresponding regional agent of the target commercial area. The control information includes charging information and local load control information. The execution unit 203 is configured to, for the global load control information, utilize the regional agent to perform the following processing steps: sending each load control information in the global load control information to the corresponding charging station; for each charging station, utilizing the corresponding edge agent to perform the following sending steps: performing charging simulation during the target off-peak time period based on the corresponding load control information to obtain simulation results; sending the load problem set corresponding to the simulation results to the regional agent; determining whether to perform load control based on the obtained load problem sets and simulation results; and in response to determining not to perform, sending the global load control information to the power management agent corresponding to the target commercial area to control the load corresponding to each charging station.

[0121] It is understandable that the units and references described in the intelligent simulation and control device 200 for grid-connected load of multi-model electric vehicles are... Figure 1 The steps in the described method correspond accordingly. Therefore, the operations, features, and beneficial effects described above for the method are also applicable to the grid-connected load intelligent simulation and control device 200 for multi-model electric vehicles and the units contained therein, and will not be repeated here.

[0122] The following is for reference. Figure 3 It shows a schematic diagram of the structure of an electronic device (e.g., an electronic device) 300 suitable for implementing some embodiments of the present disclosure. Figure 3 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of this disclosure.

[0123] like Figure 3As shown, the electronic device 300 may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 301, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 302 or a program loaded from a storage device 308 into a random access memory (RAM) 303. The RAM 303 also stores various programs and data required for the operation of the electronic device 300. The processing unit 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0124] Typically, the following devices can be connected to I / O interface 305: input devices 306 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 307 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 308 including, for example, magnetic tapes, hard disks, etc.; and communication devices 309. Communication device 309 allows electronic device 300 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 3 An electronic device 300 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively. Figure 3 Each box shown can represent a device or multiple devices as needed.

[0125] In particular, according to some embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, some embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication device 309, or installed from storage device 308, or installed from ROM 302. When the computer program is executed by processing device 301, it performs the functions defined in the methods of some embodiments of this disclosure.

[0126] It should be noted that, in some embodiments of this disclosure, the computer-readable medium described above may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In some embodiments of this disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In some embodiments of this disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0127] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol) and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future-developed networks.

[0128] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable medium carries one or more programs. When the aforementioned one or more programs are executed by the electronic device, the electronic device causes the following: For each charging station in the target commercial area, based on the charging information during the target peak time period, using an edge agent deployed on the corresponding edge terminal, to generate local load control information for the charging station during future peak time periods. The charging information includes vehicle model information. Based on each control information, using a regional agent corresponding to the target commercial area, to generate global load control information. The control information includes charging information and local load control information. For the global load control information, using the regional agent, to perform the following processing steps: sending each load control information in the global load control information to the corresponding charging station; For each charging station, using the corresponding edge agent, to perform the following sending steps: performing charging simulation during the target off-peak time period based on the corresponding load control information, and obtaining simulation results; sending the load problem set corresponding to the simulation results to the regional agent; determining whether to perform load control based on the obtained load problem sets and simulation results; and in response to determining not to perform, sending the global load control information to the power management agent corresponding to the target commercial area to control the load corresponding to each charging station.

[0129] Computer program code for performing operations of some embodiments of this disclosure can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0130] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0131] The units described in some embodiments of this disclosure can be implemented in software or hardware. The described units can also be housed in a processor; for example, a processor may be described as including a first generation unit, a second generation unit, and an execution unit. The names of these units do not necessarily limit the specific unit; for example, the second generation unit may be described as "a unit that generates global load control information based on various control information and utilizing the regional intelligent agent corresponding to the target commercial area."

[0132] The functions described above in this document can be performed at least in part by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system-on-a-chip (SoCs), complex programmable logic devices (CPLDs), and so on.

[0133] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A method for intelligent simulation and control of grid-connected load for multi-model electric vehicles, comprising: For each charging station in the target commercial area, based on the charging information during the target peak time period, the corresponding edge agent deployed on the edge terminal is used to generate local load control information for the charging station during the future peak time period. The charging information includes: vehicle model information. Based on the various control information, global load control information is generated using the regional intelligent agent corresponding to the target commercial area. The control information includes charging information and local load control information. For global load regulation information, the processing steps are performed using the regional agent: Send each load control information in the global load control information to the corresponding charging station; For each charging station, the sending step is executed using the corresponding edge agent: Based on the corresponding load control information, a charging simulation was conducted during the target off-peak period to obtain the simulation results. The set of load problems corresponding to the simulation results is sent to the regional agent; Based on the obtained load problem sets and simulation results, determine whether to implement load regulation; In response to the determination not to execute, the global load control information is sent to the power management and control agent corresponding to the target commercial area to regulate the load corresponding to each charging station.

2. The method according to claim 1, wherein, The method further includes: In response to the determination process, the following determination steps are performed for each charging station: Based on the simulation results corresponding to the charging station, the load problem set is verified to obtain the problem verification results; In response to the failure of the verification result of the problem, the local load control information corresponding to the charging station is adjusted according to the load problem set to obtain the target local load control information; In response to the verification results indicating that the problem has been verified, the local load control information is determined as the target local load control information; The obtained local load control information of each target is combined to obtain global load control information; The global load readjustment information is used as the global load readjustment information, and the processing steps are continued.

3. The method according to claim 1, wherein, The step of generating local load regulation information for the charging station during future peak periods based on charging information within the target peak time period, using edge agents deployed on corresponding edge terminals, includes: Obtain the charging station influence area division information sent by the regional intelligent agent; Based on the charging station's influence area division information, the charging sub-information corresponding to the charging station is obtained from the charging information. The charging sub-information includes: charging vehicle model ratio information, traffic flow information within the corresponding influence area, and charging power consumption corresponding to different vehicle models. Based on the charging electronic information, the first generation step is performed using an edge intelligence agent: Based on the charging device information, first future charging device information for the future peak time period is generated, the first future charging device information including: future charging vehicle type ratio information and future traffic flow; Determine whether the target holiday period exists within the future peak time period; In response to the determination of existence, the historical charging sub-information sequence corresponding to the charging station is obtained according to the target holiday period; Based on the historical charging electron information sequence, the first future charging electron information is weighted and adjusted to obtain the second future charging electron information; Based on the second future charging electronic information, local load regulation information is generated.

4. The method according to claim 3, wherein, The step of generating the first future charging information for the future peak time period based on the charging information includes: The charging vehicle ratio information, the traffic flow information, and each charging power consumption mapping information are preprocessed to obtain first preprocessed information, second preprocessed information, and each third preprocessed information, wherein the charging power consumption mapping information represents the mapping relationship between charging power consumption and vehicle type; The first preprocessing information, the second preprocessing information, and each of the third preprocessing information are respectively input into the corresponding encoding layer to obtain the first encoding result, the second encoding result, and the third encoding result; The first encoding result, the second encoding result, and the third encoding result are time-aligned and feature-fused to obtain an encoding fusion information sequence. The target peak time period is divided into simultaneous periods to obtain a peak time segment sequence set, wherein the peak time segments in the peak time segment sequence are the various time segments of the same period. For each peak time segment sequence, perform the second generation step: Select the coded fusion information subsequence corresponding to the peak time segment sequence from the coded fusion information sequence; The encoded fusion information subsequence is input into the time-series decoding layer to obtain the initial future charging vehicle ratio information; Based on the initial future charging vehicle ratio information, generate future charging vehicle ratio sub-information under the peak time segment sequence; Based on the obtained sequence of future charging vehicle proportion sub-information, the future charging vehicle proportion information is generated.

5. The method according to claim 4, wherein, After generating the future charging vehicle ratio sub-information under the peak time segment sequence based on the initial future charging vehicle ratio information, the method further includes: The encoded information subsequence corresponding to the peak time segment sequence is selected from the second encoding result; The encoded information subsequence is input into the time-series decoding layer to obtain the initial future sub-vehicle traffic flow; Obtain the previous peak time segment sequence and the next peak time segment sequence corresponding to the previous period of the peak time segment sequence; Based on the previous peak time segment sequence corresponding to the previous coded information subsequence, the next peak time segment sequence corresponding to the next coded information subsequence, and the coded information subsequence, a first period proximity relationship is generated; Based on the proximity relationship of the first period, the initial future sub-vehicle flow is adjusted to obtain the future sub-vehicle flow; and The method further includes: Based on the obtained future sub-vehicle flow sequence, generate future vehicle flow.

6. The method according to claim 4, wherein generating the future charging vehicle ratio sub-information under the peak time segment sequence based on the initial future charging vehicle ratio information comprises: Obtain the previous peak time segment sequence of the previous period and the next peak time segment sequence of the next period corresponding to the peak time segment sequence; Based on the previous peak time segment sequence corresponding to the previous coding fusion information subsequence, the next peak time segment sequence corresponding to the next coding fusion information subsequence, and the coding fusion information subsequence, a second period proximity relationship is generated; Based on the proximity relationship of the second period, the initial future charging vehicle ratio information is adjusted to obtain future charging vehicle ratio sub-information.

7. The method according to claim 1, wherein generating global load control information based on various control information and utilizing the regional agent corresponding to the target commercial area includes: Based on the charging information in each of the control information, global charging information for the target commercial area during the target peak time period is generated. Based on the global charging information and the historical global charging information sequence, the initial global load control information corresponding to the target commercial area is generated; In response to the fact that the control difference between the initial global load control information and each local load control information is less than the target difference value, the initial global load control information is determined as the global load control information; In response to the existence of at least one local load control information corresponding to a control difference not less than the target difference value, the following verification steps are performed for each local load control information: Filter out global load control sub-information that exists in the same sub-region as the local load control information from the global load control information; Perform multi-faceted verification of the global load control sub-information to obtain the verification results; In response to the verification result indicating that there is no problem, the global load control sub-information is determined as the target global load control sub-information; In response to the problem identified in the verification result, the adjustment information corresponding to the local load control information is determined as the target global load control sub-information. Based on at least one target global load control sub-information obtained, the initial global load control information is adjusted to obtain global load control information.

8. A grid-connected load intelligent simulation and control device for multi-model electric vehicles, comprising: The first generation unit is configured to generate local load control information for each charging station in the target commercial area during future peak periods, based on the charging information during the target peak time period and using the corresponding edge agent deployed on the edge terminal. The charging information includes: vehicle model information. The second generation unit is configured to generate global load control information based on various control information and using the regional intelligent agent corresponding to the target commercial area. The control information includes charging information and local load control information. The execution unit is configured to, for global load control information, utilize the regional agent to perform the following processing steps: sending each load control information in the global load control information to the corresponding charging station; for each charging station, utilizing the corresponding edge agent, performing the following sending steps: based on the corresponding load control information, performing charging simulation during the target off-peak time period to obtain simulation results; sending the load problem set corresponding to the simulation results to the regional agent; determining whether to perform load control based on the obtained load problem sets and simulation results; in response to determining not to perform, sending the global load control information to the power management agent corresponding to the target commercial area to regulate the load corresponding to each charging station.

Images