Electric vehicle charging planning method and device based on artificial intelligence, medium and product

Abstract

The invention discloses an electric vehicle charging planning method and device based on artificial intelligence, a medium and a product, and relates to the technical field of electric vehicle charging management.The method comprises the steps that charging related characteristics at the current moment are input into a user demand prediction model, and user demand data are obtained, and determining a planned charging pile according to the user demand data, the charging pile state analysis model and the power grid load prediction model, and finally planning a charging plan according to the data of the planned charging pile. According to the method, the charging related characteristics at the current moment are input into the user demand prediction model to obtain the user demand data, so that the dynamic change of the charging behavior of the user is accurately captured, and the prediction precision is greatly improved. The prediction considers the multi-dimensional factor at the current moment, so that the system can more comprehensively understand and predict the real-time charging demand of the user, and the technical problem of how to plan a reasonable charging plan is solved.

Description

Technical Field

[0001] The present application relates to the technical field of electric vehicle charging management, and in particular to an electric vehicle charging planning method, equipment, medium and product based on artificial intelligence. Background Art

[0002] With the rapid popularity of electric vehicles around the world, the demand for charging infrastructure is also growing. In order to meet the increasing demand for charging while ensuring the stable operation of the power grid and the efficient use of charging resources, the importance of electric vehicle charging management systems has become increasingly prominent. Effective charging management can not only improve user experience, but also optimize grid resource allocation and reduce operating costs.

[0003] Existing EV charging management systems typically use fixed scheduling algorithms to allocate charging resources. These systems allocate charging stations and time slots to users based on preset rules and parameters, such as the geographical location of the charging stations, current availability, and electricity prices. Some systems also consider historical data, such as the average charging demand over a specific time period, to predict future charging demand and allocate resources accordingly.

[0004] However, this charging management method based on a fixed algorithm has limitations when dealing with complex and changing actual situations. First, it cannot accurately capture and respond to the dynamic changes in user needs, which may lead to uneven or untimely distribution of charging resources. Secondly, due to the lack of comprehensive analysis of the real-time status of charging piles, the system may assign users to unsuitable charging piles, affecting charging efficiency. Finally, this method is difficult to effectively predict and balance the load on the power grid, and may cause local power grid overload during peak power consumption, affecting power supply stability. Therefore, how to plan a reasonable charging plan has become a technical problem that needs to be solved urgently. Summary of the invention

[0005] The main purpose of this application is to provide an electric vehicle charging planning method, device, medium and product based on artificial intelligence, aiming to solve the technical problem of how to plan a reasonable charging plan.

[0006] To achieve the above objectives, the present application proposes an electric vehicle charging planning method based on artificial intelligence, and the electric vehicle charging planning method based on artificial intelligence includes:

[0007] Inputting the current charging-related features into the user demand prediction model to obtain user demand data;

[0008] Determine the planned charging piles based on user demand data, charging pile status analysis model and grid load forecasting model;

[0009] A charging plan is planned based on the data of the planned charging piles.

[0010] In one embodiment, the step of determining the planned charging pile according to the user demand data, the charging pile state analysis model and the power grid load prediction model specifically includes:

[0011] Determine the required charging power based on the charging amount and charging duration in the user demand data;

[0012] Input the required charging power and charging pile data into the charging pile state analysis model to determine whether the output power of the charging pile is greater than the required charging power, and mark all charging piles whose output power is greater than or equal to the required power as target charging piles;

[0013] Based on the location of the target charging pile, obtaining power grid data of the corresponding location;

[0014] Inputting the power grid data of the corresponding location into a power grid load prediction model to determine whether the power grid at the corresponding location will be overloaded;

[0015] When the power grid at the corresponding location is not overloaded, the target charging pile at the corresponding location is used as the planned charging pile.

[0016] In one embodiment, the step of determining the planned charging pile according to the user demand data, the charging pile state analysis model and the power grid load prediction model further includes:

[0017] In the case where the power grid at the corresponding position is overloaded, the charging time is adjusted and / or the target charging pile at a different position is replaced, and the steps of determining the planned charging piles are performed according to the user demand data, the charging pile state analysis model and the power grid load prediction model, until the output result of the power grid load prediction model is not overloaded;

[0018] When the power grid at the corresponding location is not overloaded, the target charging pile at the corresponding location is used as the planned charging pile.

[0019] In one embodiment, the step of planning a charging plan according to the data of the planned charging piles specifically includes:

[0020] According to the data of the planned charging piles, obtaining the arrival route of each of the planned charging piles;

[0021] According to the charging time when not overloaded and the data of the planned charging piles, the available charging time and charging cost corresponding to each of the planned charging piles are obtained;

[0022] The charging plan for each of the planned charging piles is determined according to the arrival route of each of the planned charging piles and the charging time and charging fee corresponding to each of the planned charging piles.

[0023] In one embodiment, the step of inputting the current charging-related features into the user demand prediction model to obtain the user demand data includes:

[0024] Collect users' charging record samples and charging-related feature samples;

[0025] The initial machine learning model is supervisedly trained using the charging record samples and charging-related feature samples to obtain the user demand prediction model.

[0026] In one embodiment, the step of determining the planned charging pile according to the user demand data, the charging pile state analysis model and the grid load prediction model comprises:

[0027] Collect historical charging power, historical charging pile status, historical charging pile location and charging pile operation data under historical charging power;

[0028] The initial machine learning model is trained according to the historical charging power, the historical charging pile state, the historical charging pile position and the charging pile operation data under the historical charging power to obtain the charging pile state analysis model.

[0029] In one embodiment, the step of determining the planned charging pile according to the user demand data, the charging pile state analysis model and the power grid load prediction model further includes:

[0030] Collect historical load data, historical meteorological data and historical event data of the power grid;

[0031] The historical meteorological data and the historical event data are used as inputs, and high load is used as a label to train an initial machine model to obtain the power grid load prediction model.

[0032] In addition, to achieve the above-mentioned purpose, the present application also proposes an artificial intelligence-based electric vehicle charging planning device, the device comprising: a memory, a processor, and a computer program stored on the memory and executable on the processor, the computer program being configured to implement the steps of the artificial intelligence-based electric vehicle charging planning method as described above.

[0033] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, which is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by the processor, the steps of the electric vehicle charging planning method based on artificial intelligence as described above are implemented.

[0034] In addition, to achieve the above-mentioned purpose, the present application also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, it implements the steps of the electric vehicle charging planning method based on artificial intelligence as described above.

[0035] One or more technical solutions proposed in this application have at least the following technical effects:

[0036] The present application inputs the current charging-related characteristics into the user demand prediction model to obtain user demand data, and determines the planned charging piles based on the user demand data, the charging pile status analysis model and the power grid load prediction model, and finally plans the charging plan based on the data of the planned charging piles.

[0037] Specifically, the present invention obtains user demand data by inputting the current charging-related features into the user demand prediction model, thereby accurately capturing the dynamic changes of user charging behavior and greatly improving the prediction accuracy. This prediction takes into account the multi-dimensional factors at the current moment, enabling the system to more comprehensively understand and predict the user's real-time charging needs.

[0038] Combining user demand data, charging pile status analysis model and grid load forecasting model to determine the planned charging piles, this comprehensive analysis method enables the system to comprehensively consider multiple factors. The charging pile status analysis model can evaluate the real-time availability and performance of charging piles to avoid assigning users to charging piles that may have problems. The grid load forecasting model helps to optimize the allocation of charging resources while ensuring the stability of the grid.

[0039] At the same time, the charging plan is planned based on the determined planned charging pile data, which ensures the executability and efficiency of the charging plan. The system can dynamically adjust the charging plan according to the real-time situation, which not only optimizes the allocation of charging pile resources, but also significantly improves the overall charging efficiency. For example, the system can formulate the optimal charging time and location arrangement according to the predicted user needs and charging pile status, avoiding the situation where some charging piles are overcrowded while other charging piles are idle. Compared with the traditional fixed algorithm charging management system, the scheme of the present invention shows significant progress and substantial advantages. It can better cope with complex and changeable charging needs, and realize the efficient use of charging resources while ensuring user experience. Through intelligent demand forecasting and resource allocation, the system effectively alleviates the problems of uneven distribution of charging piles and long waiting time for users, and provides electric vehicle users with more convenient and reliable charging services. This artificial intelligence-based charging planning method not only improves user satisfaction, but also optimizes the utilization efficiency of charging resources, and promotes the popularization and sustainable development of electric vehicle use. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0041] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0042] Figure 1 A flowchart diagram of the first embodiment of the electric vehicle charging planning method based on artificial intelligence of the present application is provided;

[0043] Figure 2 A flow chart of the second embodiment of the electric vehicle charging planning method based on artificial intelligence provided by the present application;

[0044] Figure 3 A flowchart diagram of the third embodiment of the electric vehicle charging planning method based on artificial intelligence of the present application is provided;

[0045] Figure 4 This is a schematic diagram of the device structure of the hardware operating environment involved in the electric vehicle charging planning method based on artificial intelligence in the embodiment of the present application.

[0046] The purpose, features and advantages of this application will be further described in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0047] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0048] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0049] It should be noted that the execution subject of this embodiment can be a computing service device with data processing, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device capable of realizing the above functions, etc. The following takes the electric vehicle charging planning system based on artificial intelligence as an example to illustrate this embodiment and the following embodiments.

[0050] Based on this, the embodiment of the present application provides an electric vehicle charging planning method based on artificial intelligence, referring to Figure 1 , Figure 1 This is a flow chart of the first embodiment of the electric vehicle charging planning method based on artificial intelligence of the present application.

[0051] In this embodiment, the electric vehicle charging planning method based on artificial intelligence includes steps S10 to S30:

[0052] Step S10, inputting the current charging-related characteristics into the user demand prediction model to obtain user demand data;

[0053] Among them, charging-related features refer to various factors related to the charging demand of electric vehicles, including but not limited to time, weather, geographic location, user historical behavior, etc. The user demand prediction model refers to a model trained based on a machine learning algorithm to predict the user's charging demand. User demand data refers to the user's charging demand information obtained through model prediction, such as the expected charging time, charging amount, etc.

[0054] This step is performed when the electric vehicle user initiates a charging request or the system actively plans charging. Specifically, the system first collects various charging-related features at the current moment, such as the current time, user location, historical charging records, etc. Then, these feature data are used as input to a pre-trained user demand prediction model. Based on the input features, the model uses machine learning algorithms to analyze and predict the user's charging needs and outputs the prediction results, i.e., user demand data. These data may include information such as the expected charging time, required charging amount, and charging location preference.

[0055] In some embodiments, user demand prediction can be achieved in multiple ways:

[0056] Optionally, use a deep learning model such as a long short-term memory network for prediction. First, collect historical charging data and related features, preprocess and standardize the data. Then, build a long short-term memory network structure, including an input layer, a long short-term memory layer, and an output layer. Next, use the processed historical data to train the model and optimize the model parameters through a back-propagation algorithm. Finally, input the feature data at the current moment to obtain the prediction result.

[0057] Optionally, an ensemble learning method is used for prediction. First, select multiple different types of basic models, such as decision trees, random forests, and gradient boosting trees. Then, use historical data to train these models separately. Next, design an ensemble strategy, such as voting or weighted average. Finally, input the current features into each basic model, and synthesize the prediction results of each model according to the ensemble strategy to obtain the final user demand forecast.

[0058] It is understandable that other machine learning algorithms or statistical methods may also be used to predict user demand, which is not limited here.

[0059] Step S20, determining the planned charging piles according to the user demand data, the charging pile state analysis model and the grid load prediction model;

[0060] Among them, the charging pile status analysis model refers to a model used to evaluate and predict the operating status of the charging pile, which can provide information such as the availability and charging efficiency of the charging pile. The power grid load prediction model refers to a model used to predict the power grid load situation, which can predict the power grid load level in different time periods. The planned charging pile refers to the charging pile that is determined to be most suitable for the user's charging needs after comprehensive analysis.

[0061] This step is performed after obtaining the user demand data, and is used to select the most suitable charging pile. Specifically, the system first comprehensively considers the user demand data obtained in step S10, optionally, such as the expected charging time and charging amount. Then, the charging pile status analysis model is used to evaluate the current status and expected availability of each charging pile, optionally including charging power, usage rate, failure probability, etc. At the same time, the grid load situation in different regions and time periods is analyzed through the grid load prediction model. The system comprehensively analyzes and weighs these three aspects of information, considering factors such as the distance between the charging pile and the user, the availability of the charging pile, the grid load level, etc., and finally selects one or more most suitable charging piles as the planned charging piles.

[0062] In some embodiments, the determination of the planned charging pile can be achieved in multiple ways:

[0063] Optionally, a multi-objective optimization algorithm is used. First, multiple optimization objectives are defined, such as minimizing user waiting time, maximizing charging efficiency, balancing grid load, etc. Then, a mathematical model is established to transform factors such as user demand, charging pile status, and grid load into constraints and objective functions. Next, a multi-objective optimization method such as a genetic algorithm or a particle swarm optimization algorithm is applied to solve the model. Finally, the best solution is selected from the Pareto optimal solution set to determine the planned charging piles.

[0064] Alternatively, a rule-based expert system approach can be used. First, a series of rules are designed, such as "If the user demand is greater than X and the distance is less than Y, then select the Z type of charging pile." Then, a knowledge base is established to store these rules and information related to charging piles and power grids. Next, an inference engine is developed to apply rules for inference based on the currently input user demand data, charging pile status, and power grid load. Finally, based on the inference results, the charging pile that best meets the conditions is selected as the planned charging pile.

[0065] It is understandable that other decision algorithms or hybrid methods may also be used to determine the planned charging piles, which are not limited here.

[0066] Step S30, planning a charging plan according to the data of the planned charging pile;

[0067] The data of planned charging piles refers to the relevant information of the confirmed charging piles, including location, available time period, charging power, etc. The charging plan represents the specific charging plan arranged for the user, including detailed information such as charging time, location, duration, etc.

[0068] This step is performed after the planned charging pile is determined, and is used to formulate the final charging plan. Specifically, the system first analyzes the various data of the planned charging pile, such as the geographical location of the charging pile, the current queue situation, the expected available time, etc. Then, combined with the user's charging demand data, optionally, such as the required charging amount, the expected charging completion time, etc., a detailed charging plan is formulated. This plan needs to take into account multiple factors, optionally, such as the user's schedule, the efficiency of the charging pile, electricity price fluctuations, etc., mainly to ensure that the charging plan not only meets the user's needs, but also can achieve efficient use of resources.

[0069] In some embodiments, the charging plan can be planned in a variety of ways:

[0070] Optionally, use heuristic algorithms for planning. First, define evaluation indicators, such as shortest charging time, minimum waiting time, and lowest cost. Then, generate an initial charging plan based on the data of the planned charging piles. Next, continuously adjust and optimize the charging plan through heuristic algorithms such as simulated annealing or taboo search. In each iteration, evaluate the performance of the current plan and decide whether to accept the new solution based on the evaluation indicators. Finally, output the optimal charging plan when the termination condition is reached.

[0071] Optionally, a dynamic programming method is used to make a plan. First, the charging plan problem is decomposed into a series of sub-problems, such as charging decisions for each time period. Then, the state space and transfer function are defined, and the state may include the current charge amount, remaining time, etc. Next, starting from the initial state, the optimal decision and the corresponding value of each state are calculated step by step. In this process, factors such as the availability of charging piles and changes in electricity prices are considered. Finally, a complete charging plan is obtained by backtracking the optimal path. It can be understood that other optimization algorithms or decision-making methods can also be used to plan the charging plan, which is not limited here.

[0072] In summary, the present invention obtains user demand data by inputting the current charging-related features into the user demand prediction model, thereby accurately capturing the dynamic changes of user charging behavior and greatly improving the prediction accuracy. This prediction takes into account the multi-dimensional factors of the current moment, enabling the system to more comprehensively understand and predict the user's real-time charging needs.

[0073] Combining user demand data, charging pile status analysis model and grid load forecasting model to determine the planned charging piles, this comprehensive analysis method enables the system to comprehensively consider multiple factors. The charging pile status analysis model can evaluate the real-time availability and performance of charging piles to avoid assigning users to charging piles that may have problems. The grid load forecasting model helps to optimize the allocation of charging resources while ensuring the stability of the grid.

[0074] At the same time, the charging plan is planned based on the determined planned charging pile data, which ensures the executability and efficiency of the charging plan. The system can dynamically adjust the charging plan according to the real-time situation, which not only optimizes the allocation of charging pile resources, but also significantly improves the overall charging efficiency. For example, the system can formulate the optimal charging time and location arrangement according to the predicted user needs and charging pile status, avoiding the situation where some charging piles are overcrowded while other charging piles are idle. Compared with the traditional fixed algorithm charging management system, the scheme of the present invention shows significant progress and substantial advantages. It can better cope with complex and changeable charging needs, and realize the efficient use of charging resources while ensuring user experience. Through intelligent demand forecasting and resource allocation, the system effectively alleviates the problems of uneven distribution of charging piles and long waiting time for users, and provides electric vehicle users with more convenient and reliable charging services. This artificial intelligence-based charging planning method not only improves user satisfaction, but also optimizes the utilization efficiency of charging resources, and promotes the popularization and sustainable development of electric vehicle use.

[0075] Based on the first embodiment of the present application, in the second embodiment of the present application, the same or similar contents as those in the above-mentioned embodiment 1 can be referred to the above introduction, and will not be repeated in the following. Figure 2 Based on the first embodiment, this embodiment provides a more specific electric vehicle charging planning method based on artificial intelligence, as follows:

[0076] Step S201, inputting the current charging-related features into a user demand prediction model to obtain user demand data;

[0077] Refer to step S10, which will not be described in detail here.

[0078] Step S202, determining the required charging power based on the charging amount and charging duration in the user demand data;

[0079] Among them, user demand data represents the specific requirements of users for charging electric vehicles, including information such as charging capacity and charging time. Charging capacity refers to the amount of electricity that users need to charge into the battery, usually in kilowatt-hours (kWh). Charging time refers to the time that users expect to complete charging, usually in hours. The required charging power refers to the power required to complete the charging capacity required by the user within a specified time, usually in kilowatts (kW).

[0080] Specifically, this step is performed after obtaining the user demand data to calculate the charging power required to meet the user's charging needs. First, the charging amount and charging time information are extracted from the user demand data, and then the required charging power is calculated according to the physical formula power = energy / time. For example, if the user needs to charge 40kWh of electricity within 2 hours, the required charging power is 40kWh / 2h=20kW.

[0081] In some embodiments, the calculation of the required charging power can be achieved in multiple ways:

[0082] Optionally, a simple division operation can be used to directly divide the charge amount by the charging time to obtain the average power; then the average power is corrected according to the characteristics of the charging curve to obtain the peak power; finally, the actual required input power is calculated in combination with the charging efficiency.

[0083] Optionally, a table lookup method can also be used to pre-establish a corresponding relationship table between the charging amount, charging time and required power, and obtain the required power by looking up the table according to the input charging amount and charging time; then dynamically adjust the table lookup results according to factors such as battery status and ambient temperature to obtain a more accurate required power.

[0084] It is understandable that other methods can also be used to calculate the required charging power, such as using machine learning model prediction, which is not limited here.

[0085] Step S203, inputting the required charging power and charging pile data into a charging pile state analysis model, determining whether the output power of the charging pile is greater than the required charging power, and marking all charging piles whose output power is greater than or equal to the required power as target charging piles;

[0086] Among them, charging pile data refers to information describing the current status and performance of the charging pile, including the location, type, maximum output power, current available status, etc. of the charging pile. The charging pile status analysis model is a model used to analyze whether the charging pile can meet the user's charging needs. It considers various parameters and current status of the charging pile. The target charging pile refers to those charging piles that can meet the user's charging needs.

[0087] Specifically, this step is performed after the required charging power is determined, and is used to screen out charging piles that can meet the user's charging needs. First, the calculated required charging power and the data of all charging piles in the system are input into the charging pile status analysis model. The model analyzes factors such as the maximum output power and current load status of each charging pile to determine whether it can provide sufficient charging power. For those charging piles whose output power is greater than or equal to the required charging power, the system will mark them as target charging piles as candidates for subsequent charging planning.

[0088] In some embodiments, the screening and marking of charging piles can be achieved in a variety of ways:

[0089] Optionally, a simple comparison method can be used to directly compare the maximum output power of each charging pile with the required charging power, and mark the charging piles that meet the conditions as target charging piles; then consider the current usage status of the charging piles. If a charging pile meets the requirements although its maximum power meets the requirements, but is currently charging other vehicles, its remaining available power needs to be calculated to re-judge whether it meets the requirements.

[0090] Optionally, a more complex scoring system can be used to calculate a comprehensive score for each charging pile by taking into account multiple factors such as the output power of the charging pile, the distance from the user, and the historical reliability; then a score threshold is set and the charging piles with scores higher than the threshold are marked as target charging piles.

[0091] It is understandable that other methods may be used to implement the screening and marking of charging piles, such as using heuristic algorithms or machine learning methods, which are not limited here.

[0092] Step S204, based on the location of the target charging pile, obtaining power grid data of the corresponding location;

[0093] Among them, the target charging pile refers to the charging pile marked in the previous steps as being able to meet the user's charging needs. The location of the charging pile represents the geographical coordinates or area where the charging pile is located. Grid data refers to information that describes the operating status of the power system, including but not limited to voltage level, current load, power factor, line capacity, etc. The grid data at the corresponding location refers to the local grid operation data related to the location of the target charging pile, which is used to evaluate the power supply capacity and load conditions in the area.

[0094] Specifically, this step is performed after the target charging pile is determined, and is used to collect grid operation data related to the target charging pile. First, the system obtains the geographic location information of each target charging pile. Then, based on this location information, the grid operation data of the corresponding area is extracted from the grid management system or related database. This data may include information such as the substation capacity, distribution line load, and peak power consumption period in the area. The purpose of collecting this data is to evaluate the impact of increasing the charging load on the local power grid in subsequent steps.

[0095] In some embodiments, the acquisition of the corresponding location power grid data can be achieved in a variety of ways:

[0096] Alternatively, data can be obtained through historical data analysis and model estimation. First, a comprehensive database containing historical power consumption patterns, weather conditions, holidays, and other factors is established; then, relevant historical data is extracted from the database based on the location of the target charging pile and the current time; finally, data mining and machine learning algorithms are used to estimate the current grid status data based on historical data.

[0097] It is understandable that other methods may be used to acquire power grid data, such as using smart meter data to infer regional power grid status, etc., which is not limited here.

[0098] Step S205, inputting the power grid data of the corresponding location into a power grid load prediction model to determine whether the power grid at the corresponding location will be overloaded;

[0099] Among them, the power grid load forecasting model is an algorithm model used to predict the changes in power grid load. It predicts the load of the power grid in the future based on historical data and current status. Power grid overload refers to the state where the actual load of the power grid exceeds its design capacity or safe operation threshold, which may lead to a decline in power supply quality or safety hazards.

[0100] Specifically, this step is performed after obtaining the grid data of the corresponding location, and is used to evaluate whether the grid will be overloaded after adding the electric vehicle charging load. First, the grid data obtained in the previous step is input into the pre-trained grid load prediction model. The model considers a variety of factors, such as historical load patterns, weather conditions, special events, etc., to predict grid load changes in the future (usually the time period when the user plans to charge). Then, the model adds the predicted load to the expected charging load and compares it with the safety capacity of the grid to determine whether an overload will occur.

[0101] In some embodiments, grid load prediction and overload judgment can be achieved in a variety of ways:

[0102] Alternatively, a time series analysis method can be used. First, the historical grid load data is decomposed into time series to separate the trend, seasonality and random components; then, a prediction model is established based on the decomposed components; finally, the prediction results are compared with the grid capacity to determine whether there will be overload.

[0103] Optionally, machine learning methods can also be used. First, collect multi-dimensional feature data including historical load, weather, holidays, etc.; then, use algorithms such as random forest or deep neural network to train the load prediction model; finally, input the current feature data for prediction, and judge whether overload will occur based on the prediction results and safety thresholds.

[0104] It is understandable that other methods may be used to implement grid load prediction and overload judgment, such as using fuzzy logic or expert systems, which are not limited here.

[0105] Step S206, when the power grid at the corresponding position is overloaded, adjusting the charging time and / or replacing the target charging pile at a different position, and executing the step of determining the planned charging pile according to the user demand data, the charging pile state analysis model and the power grid load prediction model, until the output result of the power grid load prediction model is not overloaded;

[0106] Among them, grid overload refers to the state where the actual load of the grid exceeds its design capacity or safe operation threshold. Charging time refers to the duration of charging planned by the user. The target charging pile refers to the candidate charging pile that meets the user's charging needs. The planned charging pile refers to the charging pile that is finally determined for the user's charging. User demand data includes information such as the user's charging amount and charging time. The charging pile state analysis model is used to evaluate whether the charging pile can meet user needs. The grid load prediction model is used to predict the load of the grid after the charging load is increased.

[0107] Specifically, this step is performed when it is found that the power grid may be overloaded, and it aims to avoid power grid overload by adjusting charging parameters or selecting new charging piles. First, the system checks the output results of the power grid load forecasting model to confirm the severity and duration of the overload. Then, the system will try to adjust the charging parameters, mainly to extend the charging time. For example, if the original plan is to charge 40kWh in 2 hours, the system may try to extend the charging time to 3 hours or longer to reduce the charging power per unit time.

[0108] When adjusting the charging time, the system will take into account the time limit set by the user in the user demand data to ensure that the adjusted charging time does not exceed the user's acceptable range. At the same time, the system will recalculate the required charging power and input the new parameters into the charging pile state analysis model to ensure that the selected charging pile can still meet the adjusted demand.

[0109] If adjusting the charging time alone cannot solve the overload problem, or the adjusted charging time exceeds the user's acceptable range, the system will consider changing the target charging pile in a different location. During this process, the system will re-evaluate other candidate charging piles previously marked as target charging piles, giving priority to those located in areas with lighter grid loads.

[0110] For each new charging station considered, the system will re-execute the complete evaluation process: first use the charging station status analysis model to check whether it meets user needs, then obtain the power grid data at the location of the charging station, and use the power grid load prediction model to evaluate its impact on the local power grid. This process will be repeated until a charging station is found that can meet user needs without causing grid overload.

[0111] During the whole process, the system will continuously update and display the current optimization progress, including the number of solutions tried, details of the current best solution, etc., so that operators or users can understand the real-time status of the optimization. In addition, the system will set a threshold for the maximum number of attempts or the maximum running time to prevent an infinite loop in extreme cases.

[0112] In some embodiments, the adjustment of charging parameters and the selection of charging piles can be achieved in a variety of ways:

[0113] Optionally, an iterative optimization method can be used. First, the system calculates the amount of load that needs to be reduced based on the degree of grid overload; then, the charging time is gradually increased by a fixed time unit (such as 15 minutes) each time, and the charging power is recalculated; finally, the grid load forecast is re-executed using the updated parameters to check whether there is still overload. If increasing the charging time to the maximum value acceptable to the user still does not solve the problem, the system will turn to look for other charging piles.

[0114] Optionally, a heuristic search algorithm can also be used. First, define a comprehensive scoring function that takes into account charging time, charging pile location, and grid load; then, use methods such as simulated annealing or genetic algorithms to search for the optimal solution in the feasible solution space; finally, verify the optimal solution obtained by the search to ensure that it will not cause grid overload.

[0115] It is understandable that other methods may be used to adjust charging parameters and select charging piles, such as using a reinforcement learning algorithm to dynamically adjust the charging strategy, which is not limited here.

[0116] Step S207, when the power grid at the corresponding location is not overloaded, the target charging pile at the corresponding location is used as a planned charging pile;

[0117] Among them, the planned charging pile refers to the charging pile that is finally determined and recommended to users. It is the best choice after multiple conditions are screened and optimized. The situation where the power grid will not be overloaded means that after adding new charging loads, the predicted power grid load is still within its safe operating range.

[0118] Specifically, this step is performed after confirming that the target charging pile will not cause grid overload, and it is the last step of the entire charging pile selection process. First, the system will reconfirm the output results of the grid load forecasting model to ensure that the grid can still operate safely after the charging load is increased at the current target charging pile location. Then, the system will perform a final suitability check on the target charging pile, including verifying its current status (such as whether it is online or whether there is a fault), checking the reservation status (ensuring that it is available within the time period required by the user), and verifying whether the charging interface type matches the user's vehicle. Finally, the system will mark this target charging pile as a "planned charging pile" and send the relevant information to the user, while updating the reservation status in the charging pile management system.

[0119] Step S208, obtaining the arrival route of each of the planned charging piles according to the data of the planned charging piles;

[0120] The planned charging pile refers to the charging pile that is finally determined and recommended to the user. The data of the planned charging pile includes the geographical coordinates and address of the charging pile. The arrival route refers to the driving path from the user's current location or the specified starting point to the planned charging pile. The route may include multiple modes of transportation, such as driving, public transportation or walking.

[0121] Specifically, this step is performed after the planned charging pile is determined, aiming to provide users with convenient navigation services. First, the system obtains the user's current location or the specified departure point. Then, the system combines real-time traffic data, road conditions, and the user's possible preferences (such as avoiding toll roads, choosing the shortest time route, etc.) to calculate multiple possible arrival routes. For each route, the system estimates the driving time, distance, and possible costs (such as tolls, parking fees, etc.).

[0122] In some embodiments, the route can be obtained in a variety of ways: Optionally, an integrated map service API can be used. First, the system calls the API of a third-party map service and inputs the coordinates of the starting point and the end point; then, the multiple route information returned by the API is sorted in combination with the user's historical selection preferences; finally, the sorted route information is integrated into the charging plan, and detailed navigation instructions for each route are provided.

[0123] It is understandable that other methods can also be used to obtain the route to the destination, such as using a machine learning model to predict the route that best suits the user, taking into account the user's historical driving data and current environmental factors, which are not limited here.

[0124] Step S209, obtaining the available charging time and charging fee corresponding to each of the planned charging piles according to the charging time when not overloaded and the data of the planned charging piles;

[0125] Among them, the charging time when not overloaded refers to the duration of time that users can charge on the premise that the power grid will not be overloaded. The data of the planned charging pile includes information such as the power of the charging pile, electricity price strategy, current reservation status, etc. The charging time refers to the specific time or time period when the user can start charging. The charging fee refers to the estimated fee calculated based on the charging time, electricity price and other possible additional fees.

[0126] Specifically, this step is performed after the planned charging pile and safe charging time are determined, with the aim of providing users with specific charging time options and cost estimates. First, the system will determine the possible charging start time based on the reservation status and operating hours of the planned charging pile. Then, the system will consider the intraday changes in the grid load and find out the time periods when the grid load is lower, which usually correspond to lower electricity prices. The system will try to arrange charging during these time periods to reduce charging costs and relieve pressure on the grid.

[0127] When calculating the charging cost, the system will take into account multiple factors: basic electricity fee (calculated based on the charging amount and electricity price), service fee, time difference electricity price (peak and valley electricity price), possible preferential activities, etc. Finally, the system will generate multiple charging time options, each of which contains information such as the start time, end time, and estimated cost.

[0128] In some embodiments, the charging time and charging cost can be obtained in a variety of ways:

[0129] Optionally, a dynamic programming algorithm can be used. First, a day is divided into multiple time periods, each corresponding to a different electricity price and grid load; then, a state transition matrix is ​​constructed to represent the cost and benefit of starting charging at different time periods; finally, a dynamic programming algorithm is used to find the charging solution with the lowest total cost while ensuring that the user's charging needs and the load limit of the grid are met.

[0130] Alternatively, a heuristic search algorithm can be used. First, a scoring function is defined that takes into account factors such as charging cost, user convenience, and grid load; then, simulated annealing or genetic algorithms are used to search in the space of possible charging times and durations; finally, the top-scoring options are selected as recommended options.

[0131] It is understandable that other methods can also be used to obtain the charging time and charging cost, such as using a reinforcement learning algorithm to learn the optimal charging strategy by interacting with the environment, which is not limited here.

[0132] Step S210, determining a charging plan for each of the planned charging piles according to the arrival route of each of the planned charging piles and the charging time and charging cost corresponding to each of the planned charging piles;

[0133] Among them, the charging plan refers to a complete charging plan, which includes information about the entire process from the user's departure to the completion of charging. The arrival route refers to the driving path of the user from the current location to the charging pile. The charging time refers to the time point or time period when the user can start charging. The charging cost is the estimated total cost of charging.

[0134] Specifically, this step is performed after obtaining the arrival route, available charging time, and charging cost, aiming to provide users with a comprehensive and personalized charging plan. First, the system integrates the information obtained in the previous steps, including route information, charging time options, cost estimates, etc. Then, the system considers the user's personal preferences and constraints, such as the expected arrival time, the maximum acceptable waiting time, budget constraints, etc. Based on this information, the system generates multiple possible charging plans.

[0135] Each charging plan usually includes the following: departure time recommendation, estimated arrival time at the charging station, recommended charging start time, estimated charging end time, detailed navigation route, estimated total travel time, charging fee details, and possible additional services (such as nearby restaurant recommendations, activities that can be performed during charging, etc.). The system also takes into account possible unexpected situations, such as traffic congestion or temporary failure of the charging station, and includes alternative plans in the plan.

[0136] In some embodiments, the charging plan can be determined in a variety of ways:

[0137] Optionally, a scoring ranking method can be used. First, design a comprehensive scoring standard for each possible charging plan, including total time score, total cost score, convenience score, etc. For example, the time score can be based on the sum of travel time and charging time, the cost score can be based on the sum of road costs and charging costs, and the convenience score can take into account the richness of facilities around the charging pile. Then, set weights for each scoring factor based on user preferences. Next, calculate the weighted total score of each possible solution. Finally, select the solutions with the highest total scores as recommended charging plans and present them to the user in order of scores.

[0138] It is understandable that other methods can also be used to determine the charging plan, such as using a simple decision tree algorithm to quickly screen and generate a charging plan based on the user's main needs (such as time priority or cost priority), which is not limited here. These methods are all aimed at balancing computing efficiency and solution quality to provide users with practical and personalized charging plan recommendations.

[0139] In summary, in the embodiments of the present application, due to the use of a multi-dimensional prediction method that comprehensively considers user personal data, real-time environmental information, and historical patterns, a more accurate and personalized charging plan can be generated, which effectively solves the problem that traditional charging plans are difficult to adapt to user dynamic needs and complex urban environments, thereby realizing the intelligence and convenience of the electric vehicle charging process. This innovative method not only significantly improves the user's charging experience and reduces charging anxiety, but also helps users better plan their itineraries and charging times. By providing accurate charging recommendations, the system helps reduce user waiting time, improve charging efficiency, and optimize personal time and resource utilization. This intelligent charging plan provides strong support for the promotion of the use of electric vehicles and helps to improve the practicality and attractiveness of electric vehicles.

[0140] Based on the first and second embodiments of the present application, in the third embodiment of the present application, the same or similar contents as those of the above-mentioned first and second embodiments can be referred to the above introduction, and will not be described in detail later. Figure 3 This embodiment provides the corresponding training methods for the above-mentioned user demand prediction model, charging pile status analysis model and power grid load prediction model as follows:

[0141] Step S301, collecting charging record samples and charging-related feature samples of users;

[0142] Among them, the charging record sample refers to a detailed data set of the user's past charging behavior, including but not limited to information such as charging time, charging location, charging duration, charging amount, charging cost, etc. The charging-related feature sample refers to a data set of various external and internal factors that may affect the user's charging behavior, such as weather conditions, electricity price fluctuations, user travel plans, vehicle battery status, etc.

[0143] Specifically, this step is performed when the system is initialized or the database is updated regularly. The collection process needs to consider the comprehensiveness, authenticity and timeliness of the data. The system will obtain a large number of users' historical charging records and related feature data from multiple channels to form a rich training data set. These data will provide a basis for subsequent model training.

[0144] Step S302, performing supervised training on the initial machine learning model through the charging record samples and the charging-related feature samples to obtain the user demand prediction model;

[0145] The initial machine learning model refers to the original model structure that has not been trained, which can be an algorithm framework such as a neural network, decision tree, support vector machine, etc. Supervised training refers to the process of training a model using labeled data (i.e. known inputs and corresponding correct outputs). The user demand prediction model refers to the final model that can accurately predict the user's charging needs based on the current features after training.

[0146] Specifically, this step is performed after sufficient data is collected. It uses the charging record samples collected in step S301 as labels (output) and charging-related feature samples as features (input) to train the selected machine learning model. During the training process, the model continuously adjusts internal parameters to learn the complex relationship between charging behavior and various influencing factors, and finally forms a model that can accurately predict future charging needs.

[0147] In some embodiments, this step can be implemented in a variety of ways:

[0148] Optionally, the following process can be used: 1) Data preprocessing: clean the collected data, handle missing values ​​and outliers, and perform feature engineering (such as feature selection, feature combination, etc.); 2) Model selection: select appropriate machine learning algorithms such as random forests, deep neural networks, gradient boosting trees, etc. according to the characteristics of the problem and the size of the data; 3) Data set division: randomly divide the preprocessed data set into training sets, validation sets, and test sets, usually in a ratio of 6:2:2; 4) Model training: use the training set to iteratively train the model and continuously adjust the parameters. 5) Model evaluation and optimization: use the validation set to evaluate the model performance, adjust the model structure or hyperparameters based on the evaluation results, and may need to repeat the training and evaluation process multiple times; 6) Final test: when the model performance meets the requirements on the validation set, use the test set that has not been involved in the training before for the final evaluation to ensure the actual prediction ability of the model; 7) Model deployment: save and deploy the trained model to the production environment, ready to receive real-time data for prediction.

[0149] It is understandable that other methods can be used to implement model training, such as using advanced technologies such as transfer learning and ensemble learning, or using online learning to continuously update the model, which is not limited here. The key is to ensure the accuracy, robustness and interpretability of the model.

[0150] Step S303, inputting the current charging-related features into the user demand prediction model to obtain user demand data;

[0151] Refer to step S10, which will not be described in detail here.

[0152] Step S304, collecting historical charging power, historical charging pile status, historical charging pile location and charging pile operation data under historical charging power;

[0153] Among them, historical charging power refers to the power output level provided by the charging pile over a period of time in the past, usually in kilowatts (kW). The historical charging pile status indicates the working status of the charging pile at different points in time, such as normal operation, failure or maintenance. The historical charging pile location is used to record the geographic coordinates and surrounding environment information of the charging pile. The charging pile operation data under historical charging power refers to the performance and operating parameters of the charging pile under different charging power conditions, such as temperature, voltage fluctuations and charging efficiency.

[0154] Specifically, this step is usually performed when the system is initialized or the database is updated regularly. It involves collecting historical operation data of charging piles from multiple sources to build a comprehensive data set to provide a basis for subsequent state analysis model training. This step needs to consider the time span, sampling frequency and data quality of the data to ensure that the collected information can accurately reflect the operating status of the charging pile under various conditions.

[0155] In some embodiments, the data collection step can be implemented in a variety of ways:

[0156] Optionally, you can work with charging pile operators to directly obtain their historical operating data; install smart monitoring equipment on the charging piles to record data in real time; and use geographic information systems to update location information.

[0157] Alternatively, another approach is to obtain user charging experience through an application; using machine learning technology to identify abnormal conditions from power consumption patterns;

[0158] It is understandable that data collection may also be achieved in other ways, such as using satellite images to analyze changes in the surrounding environment, which is not limited here.

[0159] Step S305, training an initial machine learning model according to the historical charging power, the historical charging pile state, the historical charging pile position and the charging pile operation data under the historical charging power to obtain the charging pile state analysis model;

[0160] The initial machine learning model refers to the original algorithm structure that has not been trained, which can be a neural network, decision tree, etc. The training process refers to the use of collected historical data to optimize the model parameters so that the model can accurately capture the relationship between the charging pile status and various factors. The charging pile status analysis model refers to a model that can predict the future status and performance of the charging pile based on the current input conditions after training.

[0161] Specifically, this step is performed after sufficient historical data has been collected. It uses the comprehensive data set collected in step S304 to train the machine learning model so that it can understand how factors such as charging power, environmental conditions, and usage patterns affect the status and performance of charging piles. The training process involves multiple stages such as data preprocessing, feature engineering, model parameter optimization, and performance evaluation. The goal is to build an intelligent system that can accurately predict the status of charging piles and identify potential problems.

[0162] In some embodiments, the model training step can be implemented in a variety of ways:

[0163] Optionally, you can perform data preprocessing and feature engineering first, then select an appropriate machine learning algorithm such as LSTM or random forest, divide the dataset into training, validation, and test sets, and perform model training and evaluation.

[0164] Alternatively, another approach is to adopt an ensemble learning method, train multiple models of different types, implement an online learning mechanism, introduce uncertainty quantification, and establish a model monitoring system.

[0165] It is understandable that other methods can also be used to implement model training, such as using transfer learning or federated learning technology, which is not limited here.

[0166] Step S306, collecting historical load data, historical meteorological data and historical event data of the power grid;

[0167] Among them, historical load data refers to the past electricity consumption records of the power grid, usually measured per hour or per day. Historical meteorological data represents weather factors that affect electricity consumption, such as temperature, humidity, etc. Historical event data refers to special event records that may affect the load of the power grid, such as holidays or large-scale events.

[0168] Specifically, this step is performed before building a grid load forecasting model. It involves collecting comprehensive historical data from multiple sources to provide a basis for subsequent load forecasting model training. This step needs to consider the time span, sampling frequency, and quality of the data to ensure that the collected information accurately reflects the relationship between grid load and various factors. It is usually necessary to collect at least several years of data to capture seasonal changes and long-term trends.

[0169] In some embodiments, the data collection step can be implemented in a variety of ways:

[0170] Optionally, you can first use an open source web crawler tool to automatically collect historical meteorological data from public weather websites, including daily temperature, humidity, and precipitation information. Then, subscribe to the public event calendar API to obtain historical event data such as holidays and major events. Finally, use an open source database management system (such as MySQL or PostgreSQL) to create a local database to store and manage all the collected data.

[0171] It is understandable that data collection can also be achieved in other ways, such as using an open source IoT platform to connect a simple sensor network to collect local environmental data, which is not limited here. These methods are all aimed at achieving basic data collection functions with minimal resources and technical barriers.

[0172] Step S307, taking the historical meteorological data and the historical event data as input, taking high load as a label, and training the initial machine model to obtain the power grid load prediction model;

[0173] Among them, input refers to the independent variables used to train the model, including meteorological data and event data. High load refers to the state where the power grid load exceeds a certain threshold, which is used as the prediction target of the model. The initial machine model refers to the untrained algorithm structure. The training process refers to the optimization of model parameters using historical data. The power grid load forecasting model refers to a model that can predict future loads based on given conditions.

[0174] Specifically, this step is performed after sufficient data is collected. It uses the data set of step S306 to train the model so that it understands how various factors affect the grid load, especially the high load state. The training process includes data preprocessing, feature engineering, model optimization and evaluation. First, the "high load" standard is defined, and then the data is divided into a training set, a validation set and a test set. The model training goal is to minimize the prediction error while avoiding overfitting. Finally, the test set is used to evaluate the generalization ability of the model.

[0175] In some embodiments, the model training step can be implemented in a variety of ways:

[0176] Optionally, you can first preprocess the data, handle missing values ​​and outliers, and standardize the data. Then perform feature engineering to create new features such as the moving average of temperature. Then choose an appropriate algorithm, such as LSTM, and use cross-validation for training and parameter adjustment. Finally, perform an interpretive analysis of the model to understand the importance of each factor.

[0177] Alternatively, another approach is to adopt an ensemble learning method, train multiple models of different types, implement an online learning mechanism, introduce uncertainty quantification, and establish a model monitoring system.

[0178] It is understandable that other methods can also be used to implement model training, such as using the attention mechanism in deep learning, which is not limited here.

[0179] Step S308, determining the planned charging piles according to the user demand data, the charging pile state analysis model and the grid load prediction model;

[0180] Refer to step S20, which will not be described in detail here.

[0181] Step S309, planning a charging plan according to the data of the planned charging pile;

[0182] Refer to step S30, which will not be described in detail here.

[0183] In summary, in the embodiment of the present application, due to the innovative method of three-model collaborative prediction, the user's long-term charging habits, real-time traffic conditions and the impact of special events can be considered at the same time, which effectively solves the problem that a single model is difficult to take into account personalized needs and environmental changes, thereby realizing a smarter and more flexible electric vehicle charging plan. The basic charging demand model provides a stable prediction of personal habits, the dynamic adjustment model enhances the system's responsiveness to real-time situations, and the event impact model effectively handles the sudden impact of special situations on charging demand. This multi-model collaborative method not only improves the accuracy and practicality of charging suggestions, but also enhances the adaptability of the system. Especially in the face of complex and changing urban environments, the system can quickly adjust the charging strategy, reduce user waiting time, and improve charging efficiency. In addition, the continuous optimization mechanism of the three models ensures that the system can continuously improve performance over time and adapt to long-term changes in user behavior and urban environment. This innovative charging plan method provides electric vehicle users with a smarter and more convenient charging experience, helps to optimize the allocation of charging resources, alleviate charging pressure, and provides strong support for the promotion of electric vehicle use. In general, this embodiment builds an intelligent, precise, and adaptable electric vehicle charging planning system through the innovative method of multi-model collaboration and continuous learning, which contributes important value to the realization of intelligent transportation and sustainable urban development.

[0184] It should be noted that the above examples are only used to understand the present application and do not constitute a limitation on the artificial intelligence-based electric vehicle charging planning method of the present application. More simple transformations based on this technical concept are all within the scope of protection of the present application.

[0185] The present application provides an artificial intelligence-based electric vehicle charging planning device, which includes: at least one processor; and a memory that is communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the artificial intelligence-based electric vehicle charging planning method in the above-mentioned embodiment one.

[0186] Reference below Figure 4 , which shows a schematic diagram of the structure of an electric vehicle charging planning device based on artificial intelligence suitable for implementing the embodiment of the present application. The electric vehicle charging planning device based on artificial intelligence in the embodiment of the present application may include but is not limited to mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 4 The artificial intelligence-based electric vehicle charging planning device shown is merely an example and should not bring any limitations to the functions and scope of use of the embodiments of the present application.

[0187] like Figure 4As shown, the electric vehicle charging planning device based on artificial intelligence may include a processing device 1001 (such as a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM: Read Only Memory) 1002 or the program loaded from the storage device 1003 to the random access memory (RAM: Random Access Memory) 1004. In RAM1004, various programs and data required for the operation of the xxx device are also stored. The processing device 1001, ROM1002 and RAM1004 are connected to each other through a bus 1005. The input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems can be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touchpad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the artificial intelligence-based electric vehicle charging planning device to communicate with other devices wirelessly or by wire to exchange data. Although the figure shows an artificial intelligence-based electric vehicle charging planning device with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems can be implemented or have alternatively.

[0188] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.

[0189] The electric vehicle charging planning device based on artificial intelligence provided by the present application adopts the electric vehicle charging planning method based on artificial intelligence in the above-mentioned embodiment, which can solve the technical problem of how to plan a reasonable charging plan. Compared with the prior art, the beneficial effects of the electric vehicle charging planning device based on artificial intelligence provided by the present application are the same as the beneficial effects of the electric vehicle charging planning method based on artificial intelligence provided by the above-mentioned embodiment, and the other technical features in the electric vehicle charging planning device based on artificial intelligence are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.

[0190] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0191] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0192] The present application provides a computer-readable storage medium having computer-readable program instructions (i.e., computer programs) stored thereon, and the computer-readable program instructions are used to execute the electric vehicle charging planning method based on artificial intelligence in the above-mentioned embodiment.

[0193] The computer-readable storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination of the above.

[0194] The above-mentioned computer-readable storage medium may be included in the electric vehicle charging planning device based on artificial intelligence; or it may exist independently without being assembled into the electric vehicle charging planning device based on artificial intelligence.

[0195] The modules involved in the embodiments described in this application may be implemented by software or hardware, wherein the name of the module does not constitute a limitation on the unit itself in some cases.

[0196] The readable storage medium provided in this application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned artificial intelligence-based electric vehicle charging planning method, and can solve the technical problem of how to plan a reasonable charging plan. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as the beneficial effects of the artificial intelligence-based electric vehicle charging planning method provided in the above-mentioned embodiment, and will not be repeated here.

[0197] The present application also provides a computer program product, including a computer program, which, when executed by a processor, implements the steps of the above-mentioned artificial intelligence-based electric vehicle charging planning method.

[0198] The computer program product provided in this application can solve the technical problem of how to plan a reasonable charging plan. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as the beneficial effects of the electric vehicle charging planning method based on artificial intelligence provided in the above embodiment, which will not be repeated here.

[0199] The above descriptions are only some embodiments of the present application, and are not intended to limit the patent scope of the present application. All equivalent structural changes made using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect applications in other related technical fields are included in the patent protection scope of the present application.

Claims

1. An electric vehicle charging planning method based on artificial intelligence, characterized in that: The electric vehicle charging planning method based on artificial intelligence includes: Inputting the current charging-related features into the user demand prediction model to obtain user demand data; Determine the planned charging piles based on user demand data, charging pile status analysis model and grid load forecasting model; A charging plan is planned based on the data of the planned charging piles.

2. The method according to claim 1, characterized in that The step of determining the planned charging piles according to the user demand data, the charging pile state analysis model and the power grid load prediction model specifically includes: Determine the required charging power based on the charging amount and charging duration in the user demand data; Input the required charging power and charging pile data into the charging pile state analysis model to determine whether the output power of the charging pile is greater than the required charging power, and mark all charging piles whose output power is greater than or equal to the required power as target charging piles; Based on the location of the target charging pile, obtaining power grid data of the corresponding location; Inputting the power grid data of the corresponding location into a power grid load prediction model to determine whether the power grid at the corresponding location will be overloaded; When the power grid at the corresponding location is not overloaded, the target charging pile at the corresponding location is used as the planned charging pile.

3. The method according to claim 2, characterized in that The step of determining the planned charging piles according to the user demand data, the charging pile state analysis model and the power grid load prediction model also includes: In the case where the power grid at the corresponding position is overloaded, the charging time is adjusted and / or the target charging pile at a different position is replaced, and the steps of determining the planned charging piles are performed according to the user demand data, the charging pile state analysis model and the power grid load prediction model, until the output result of the power grid load prediction model is not overloaded; When the power grid at the corresponding location is not overloaded, the target charging pile at the corresponding location is used as the planned charging pile.

4. The method according to claim 3, characterized in that The step of planning a charging plan according to the data of the planned charging pile specifically includes: According to the data of the planned charging piles, obtaining the arrival route of each of the planned charging piles; According to the charging time when not overloaded and the data of the planned charging piles, the available charging time and charging cost corresponding to each of the planned charging piles are obtained; The charging plan for each of the planned charging piles is determined according to the arrival route of each of the planned charging piles and the charging time and charging fee corresponding to each of the planned charging piles.

5. The method according to claim 1, characterized in that The step of inputting the current charging-related features into the user demand prediction model to obtain user demand data includes: Collect users' charging record samples and charging-related feature samples; The initial machine learning model is supervisedly trained using the charging record samples and charging-related feature samples to obtain the user demand prediction model.

6. The method according to claim 1, characterized in that The step of determining the planned charging piles according to the user demand data, the charging pile state analysis model and the power grid load prediction model comprises: Collect historical charging power, historical charging pile status, historical charging pile location and charging pile operation data under historical charging power; The initial machine learning model is trained according to the historical charging power, the historical charging pile state, the historical charging pile position and the charging pile operation data under the historical charging power to obtain the charging pile state analysis model.

7. The method according to claim 1, characterized in that The step of determining the planned charging piles according to the user demand data, the charging pile state analysis model and the power grid load prediction model also includes: Collect historical load data, historical meteorological data and historical event data of the power grid; The historical meteorological data and the historical event data are used as inputs, and high load is used as a label to train an initial machine model to obtain the power grid load prediction model.

8. An electric vehicle charging planning device based on artificial intelligence, characterized in that: The artificial intelligence-based electric vehicle charging planning device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the artificial intelligence-based electric vehicle charging planning method as described in any one of claims 1 to 7.

9. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the electric vehicle charging planning method based on artificial intelligence as claimed in any one of claims 1 to 7 are implemented.

10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of the electric vehicle charging planning method based on artificial intelligence as claimed in any one of claims 1 to 7 are implemented.

Images