WO2023226368A1 - 电动汽车集群充放电控制方法、系统及相关设备 - Google Patents

电动汽车集群充放电控制方法、系统及相关设备 Download PDF

Info

Publication number
WO2023226368A1
WO2023226368A1 PCT/CN2022/137650 CN2022137650W WO2023226368A1 WO 2023226368 A1 WO2023226368 A1 WO 2023226368A1 CN 2022137650 W CN2022137650 W CN 2022137650W WO 2023226368 A1 WO2023226368 A1 WO 2023226368A1
Authority
WO
WIPO (PCT)
Prior art keywords
power consumption
electric vehicle
time period
charging
power generation
Prior art date
Application number
PCT/CN2022/137650
Other languages
English (en)
French (fr)
Inventor
杨之乐
赵世豪
郭媛君
胡天宇
刘祥飞
Original Assignee
深圳先进技术研究院
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳先进技术研究院 filed Critical 深圳先进技术研究院
Publication of WO2023226368A1 publication Critical patent/WO2023226368A1/zh

Links

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60L—PROPULSION OF ELECTRICALLY-PROPELLED VEHICLES; SUPPLYING ELECTRIC POWER FOR AUXILIARY EQUIPMENT OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRODYNAMIC BRAKE SYSTEMS FOR VEHICLES IN GENERAL; MAGNETIC SUSPENSION OR LEVITATION FOR VEHICLES; MONITORING OPERATING VARIABLES OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRIC SAFETY DEVICES FOR ELECTRICALLY-PROPELLED VEHICLES
    • B60L53/00—Methods of charging batteries, specially adapted for electric vehicles; Charging stations or on-board charging equipment therefor; Exchange of energy storage elements in electric vehicles
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/004—Artificial life, i.e. computing arrangements simulating life
    • G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Energy or water supply
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T10/00—Road transport of goods or passengers
    • Y02T10/60—Other road transportation technologies with climate change mitigation effect
    • Y02T10/70—Energy storage systems for electromobility, e.g. batteries
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T10/00—Road transport of goods or passengers
    • Y02T10/60—Other road transportation technologies with climate change mitigation effect
    • Y02T10/7072—Electromobility specific charging systems or methods for batteries, ultracapacitors, supercapacitors or double-layer capacitors
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S30/00—Systems supporting specific end-user applications in the sector of transportation
    • Y04S30/10—Systems supporting the interoperability of electric or hybrid vehicles
    • Y04S30/12—Remote or cooperative charging

Definitions

  • the present invention relates to the technical field of electric vehicle charge and discharge scheduling, and in particular to an electric vehicle cluster charge and discharge control method, system and related equipment.
  • electric vehicles can be used as a substitute for traditional fossil fuel vehicles.
  • electric vehicles can be charged using electricity generated by renewable energy sources, which is beneficial to reducing environmental pollution.
  • the main purpose of the present invention is to provide an electric vehicle cluster charging and discharging control method, system and related equipment, aiming to solve the problem of lack of reasonable arrangement and control of the charging and discharging process of electric vehicles in the prior art.
  • a first aspect of the present invention provides a method for controlling the charging and discharging of an electric vehicle cluster, wherein the above-mentioned method for controlling the charging and discharging of an electric vehicle cluster includes:
  • the power consumption prediction data of the above-mentioned electric vehicles includes the The actual power consumption of electric vehicles in each historical time segment in the previous target time period.
  • the power consumption prediction data of the above-mentioned electric vehicles includes the predicted power consumption of the electric vehicle in each prediction time segment in the above-mentioned current target time period. quantity;
  • the charging variance is the smallest, the target discharge variance is the smallest and the power consumption is the largest.
  • the above target charging variance is the variance of the charging amount of the above-mentioned electric vehicle cluster during the above-mentioned current target time period.
  • the above-mentioned target discharge variance is the above-mentioned electric vehicle cluster during the above-mentioned current target time. The variance of the discharge amount within the segment.
  • the above-mentioned consumption amount is the planned overall charging amount of the above-mentioned electric vehicle cluster in the above-mentioned current target time period.
  • the above-mentioned control constraints include cluster charging amount range constraints and cluster discharge amount range constraints.
  • the above-mentioned cluster charging The quantity range constraint is used to limit the above-mentioned consumption power to not be less than the above-mentioned target power generation, and the above-mentioned cluster discharge range constraint is used to limit the planned overall discharge of the above-mentioned electric vehicle cluster within the above-mentioned current target time period to not be less than the above-mentioned predicted overall power consumption. ;
  • Each of the above-mentioned electric vehicles in the above-mentioned electric vehicle cluster is controlled according to the above-mentioned charge and discharge control strategy.
  • the above-mentioned renewable energy includes wind energy and solar energy in the preset target area
  • the above-mentioned acquisition of the target power generation amount of renewable energy in the current target time period includes:
  • the environmental prediction information within the above-mentioned current target time period, wherein the above-mentioned environmental prediction information includes wind intensity, wind direction, light intensity and light time;
  • the above target power generation amount is obtained based on the above environmental prediction information.
  • the above-mentioned method of obtaining the above-mentioned target power generation amount based on the above-mentioned environmental prediction information includes:
  • the environmental prediction information is input into a pre-trained power generation prediction model, and the predicted power generation output from the trained power generation prediction model is used as the target power generation in the current target time period.
  • the above power generation prediction model is pre-trained according to the following steps:
  • one of the real power generation data includes actual environmental information within a power generation time period and the actual power generation amount within the power generation time period.
  • the above actual environmental information includes wind intensity, wind direction, and light intensity. and light time;
  • the above-mentioned power generation prediction model is trained according to the above-mentioned training power generation data set, and the trained power generation prediction model is tested through the above-mentioned test power generation data set until a trained power generation prediction model is obtained, wherein the above-mentioned test power generation data set is concentrated
  • the loss value between the predicted power generation output by the above-trained power generation prediction model and the actual power generation corresponding to the actual environmental information is not greater than Preset loss threshold for power generation forecast.
  • the above-mentioned time period label is any one of the pre-set label data.
  • the above-mentioned label data includes working days, weekends and special holidays.
  • the above-mentioned Transformer model is pre-trained according to the following steps:
  • one of the above real power consumption data includes a time period label corresponding to a power consumption time period, a real power consumption label and power consumption history collection data, and the above real power consumption label includes each The actual power consumption of the above-mentioned electric vehicles in each power consumption time segment of the above-mentioned power consumption time period.
  • the above-mentioned power consumption history collection data includes each power consumption history collection time of each of the above-mentioned electric vehicles in the power consumption history collection time period.
  • the actual power consumption within the segment, the above-mentioned power consumption history collection time period is the previous time period corresponding to the above-mentioned power consumption time period;
  • the above-mentioned Transformer model is trained according to the above-mentioned training power consumption data set, and the trained Transformer model is tested through the above-mentioned test power consumption data set until a trained Transformer model is obtained, in which any one of the above-mentioned test power consumption data sets is
  • the time period tags and power consumption history collection data in the real power consumption data are used as the input data of the above-trained Transformer model, the power consumption prediction data output by the above-trained Transformer model is consistent with the real power consumption in the real power consumption data.
  • the loss value between tags is not greater than the preset power consumption prediction loss threshold.
  • the above comprehensive control objective also includes minimizing the total charging distance of the above electric vehicle cluster.
  • the above method also includes:
  • the preset particle swarm algorithm is used to optimize the solution and obtain the charge and discharge control strategy of each of the above electric vehicles, including:
  • An objective function is constructed according to the above comprehensive control objective, wherein the above objective function is the sum of the inverse number of the above consumed power, the above target charging variance, the above target discharge variance and the above total charging distance;
  • the above particle swarm algorithm is used to optimize the solution and obtain the charge and discharge control strategy of each of the above electric vehicles.
  • a second aspect of the present invention provides an electric vehicle cluster charge and discharge control system, wherein the above-mentioned electric vehicle cluster charge and discharge control system includes:
  • the power generation acquisition module is used to obtain the target power generation of renewable energy within the current target time period
  • the power consumption prediction module is used to obtain the time period label corresponding to the current target time period and the power consumption history data of each electric vehicle in the electric vehicle cluster, and input the above time period label and the above power consumption history data into the pre-trained Transformer model,
  • the power consumption prediction data corresponding to each of the above-mentioned electric vehicles in the above-mentioned current target time period is obtained through the above-mentioned pre-trained Transformer model, and the predicted overall power consumption of the above-mentioned electric vehicle cluster in the above-mentioned current target time period is obtained, wherein the above-mentioned electric vehicles
  • the historical power consumption data of the car includes the actual power consumption of the electric car in each historical time segment in the previous target time period.
  • the power consumption forecast data of the above-mentioned electric car includes the electric car’s actual power consumption in each historical time segment in the above-mentioned current target time period. Predicted power consumption in predicted time segments;
  • the strategy acquisition module is used to construct comprehensive control objectives and control constraints. According to the above comprehensive control objectives and the above control constraints, optimization is performed through the preset particle swarm algorithm and the charging and discharging control strategies of each of the above electric vehicles are obtained, where,
  • the above-mentioned comprehensive control objectives include the minimum target charging variance, the minimum target discharge variance and the maximum power consumption.
  • the above-mentioned target charging variance is the variance of the charging amount of the above-mentioned electric vehicle cluster in the above-mentioned current target time period.
  • the above-mentioned target discharge variance is the above-mentioned electric vehicle cluster. The variance of the discharge amount of the cluster in the above-mentioned current target time period.
  • the above-mentioned consumption amount is the planned overall charging amount of the above-mentioned electric vehicle cluster in the above-mentioned current target time period.
  • the above-mentioned control constraints include the cluster charging amount range constraint and the cluster discharge amount. Range constraints.
  • the above-mentioned cluster charging range constraints are used to limit the above-mentioned consumption power to not be less than the above-mentioned target power generation.
  • the above-mentioned cluster discharge range constraints are used to limit the above-mentioned electric vehicle cluster's planned overall discharge amount within the above-mentioned current target time period to be no less than The above predicted overall power consumption;
  • a control module is used to control each of the above-mentioned electric vehicles in the above-mentioned electric vehicle cluster according to the above-mentioned charge and discharge control strategy.
  • a third aspect of the present invention provides an intelligent terminal.
  • the intelligent terminal includes a memory, a processor, and an electric vehicle cluster charging and discharging control program stored in the memory and executable on the processor.
  • the above-mentioned electric vehicle cluster charging and discharging control program When executed by the above-mentioned processor, the steps of implementing any of the above-mentioned electric vehicle cluster charging and discharging control methods are implemented.
  • a fourth aspect of the present invention provides a computer-readable storage medium.
  • the computer-readable storage medium stores an electric vehicle cluster charging and discharging control program.
  • the electric vehicle cluster charging and discharging control program is executed by a processor, any of the above-mentioned electric vehicle cluster charging and discharging control programs are implemented. Steps of the vehicle cluster charging and discharging control method.
  • the target power generation amount of renewable energy in the current target time period is obtained; the time period tag corresponding to the above-mentioned current target time period and the power consumption history data of each electric vehicle in the electric vehicle cluster are obtained, and the above-mentioned time period is The label and the above power consumption history data are input into the pre-trained Transformer model, and the power consumption prediction data corresponding to the above-mentioned electric vehicles in the above-mentioned current target time period is obtained through the above-mentioned pre-trained Transformer model, and the above-mentioned current target time period of the above-mentioned electric vehicle cluster is obtained.
  • the predicted overall power consumption within the target time period where the power consumption history data of the above-mentioned electric vehicle includes the actual power consumption of the electric vehicle in each historical time segment in the previous target time period, and the power consumption of the above-mentioned electric vehicle
  • the prediction data includes the predicted power consumption of the electric vehicle in each prediction time segment within the above-mentioned current target time period; construct a comprehensive control objective and control constraints, and use the preset particles according to the above-mentioned comprehensive control objectives and the above-mentioned control constraints.
  • the swarm algorithm is used to optimize the solution and obtain the charge and discharge control strategy of each of the above-mentioned electric vehicles.
  • the above-mentioned comprehensive control objectives include the minimum target charging variance, the minimum target discharge variance and the maximum consumption of electricity.
  • the above-mentioned target charging variance is the above-mentioned electric vehicle cluster in the above-mentioned The variance of the charging amount in the current target time period.
  • the above target discharge variance is the variance of the discharge amount of the above-mentioned electric vehicle cluster in the above-mentioned current target time period.
  • the above-mentioned consumption amount is the variance of the above-mentioned electric vehicle cluster in the above-mentioned current target time period.
  • the above control constraints include the cluster charging amount range constraint and the cluster discharge amount range constraint.
  • the above cluster charging amount range constraint is used to limit the above-mentioned consumption amount to not be less than the above-mentioned target power generation amount.
  • the above-mentioned cluster discharge amount range constraint is used to Limit the planned overall discharge amount of the above-mentioned electric vehicle cluster within the above-mentioned current target time period to not be less than the above-mentioned predicted overall power consumption; control each of the above-mentioned electric vehicles in the above-mentioned electric vehicle cluster according to the above-mentioned charge and discharge control strategy.
  • the solution of the present invention comprehensively considers the target power generation of renewable energy and the predicted overall power consumption of the electric vehicle cluster within the current target time period, and constructs comprehensive control objectives and control constraints for the electric vehicle cluster.
  • the optimization solution is performed and the optimized charging and discharging control strategy is obtained, which is conducive to better charging and discharging control of electric vehicles.
  • the minimum target charging variance is used to ensure that the fluctuations in the power grid caused by the electric vehicle cluster are small, which is beneficial to protecting the power grid.
  • the minimum target discharge variance is used to ensure that the discharge fluctuations of each electric vehicle in the electric vehicle cluster are small, which is beneficial to protecting electric vehicles and eliminating waste. Maximizing the amount of energy absorbed will help improve the utilization rate of renewable energy.
  • Figure 1 is a schematic flow chart of an electric vehicle cluster charging and discharging control method provided by an embodiment of the present invention
  • FIG. 2 is a specific flow diagram of step S100 in Figure 1 according to the embodiment of the present invention.
  • Figure 3 is a schematic structural diagram of an electric vehicle cluster charging and discharging control system provided by an embodiment of the present invention
  • Figure 4 is a functional block diagram of the internal structure of an intelligent terminal provided by an embodiment of the present invention.
  • electric vehicles can be used as a substitute for traditional fossil fuel vehicles.
  • electric vehicles can be charged using electricity generated by renewable energy sources, which is beneficial to reducing environmental pollution.
  • the user randomly selects the discharge power, charging time and charging area of the electric vehicle, which may lead to a lack of power during driving. Or it may take a long distance to reach the charging area selected by the user, resulting in a waste of time. It may even be found that there is no charging space (or charging pile) after arriving at the charging area, thus affecting the user's use.
  • excessive charging piles need to be set up in each charging area, resulting in an excess of charging piles.
  • the target power generation amount of renewable energy in the current target time period is obtained; the time period label corresponding to the above current target time period and the number of each electric vehicle in the electric vehicle cluster are obtained.
  • For power consumption history data input the above-mentioned time period label and the above-mentioned power consumption history data into the pre-trained Transformer model, and obtain the power consumption prediction data corresponding to the above-mentioned electric vehicles in the above-mentioned current target time period through the above-mentioned pre-trained Transformer model, and Obtain the predicted overall power consumption of the above-mentioned electric vehicle cluster in the above-mentioned current target time period, wherein the power consumption history data of the above-mentioned electric vehicle includes the actual power consumption of the electric vehicle in each historical time segment within the previous target time period.
  • the power consumption prediction data of the above-mentioned electric vehicle includes the predicted power consumption of the electric vehicle in each prediction time segment within the above-mentioned current target time period; construct a comprehensive control objective and control constraints, and according to the above-mentioned comprehensive control objective and the above-mentioned control Constraint conditions are optimized and solved through the preset particle swarm algorithm to obtain the charge and discharge control strategy of each of the above-mentioned electric vehicles.
  • the above-mentioned comprehensive control objectives include the minimum target charging variance, the minimum target discharge variance and the maximum power consumption.
  • the above-mentioned target charging The variance is the variance of the charging amount of the above-mentioned electric vehicle cluster in the above-mentioned current target time period
  • the above-mentioned target discharge variance is the variance of the above-mentioned discharge amount of the above-mentioned electric vehicle cluster in the above-mentioned current target time period
  • the above-mentioned consumption amount is the variance of the above-mentioned electric vehicle cluster
  • the above-mentioned control constraints include cluster charging amount range constraints and cluster discharge amount range constraints.
  • the above-mentioned cluster charging amount range constraints are used to limit the above-mentioned consumption amount to not be less than the above-mentioned target power generation amount
  • the above-mentioned cluster discharge amount range constraint is used to limit the planned overall discharge amount of the above-mentioned electric vehicle cluster within the above-mentioned current target time period to not be less than the above-mentioned predicted overall power consumption; according to the above-mentioned charge and discharge control strategy, each of the above-mentioned electric vehicles in the above-mentioned electric vehicle cluster is The car is controlled.
  • the solution of the present invention comprehensively considers the target power generation of renewable energy and the predicted overall power consumption of the electric vehicle cluster within the current target time period, and constructs comprehensive control objectives and control constraints for the electric vehicle cluster.
  • the optimization solution is performed and the optimized charging and discharging control strategy is obtained, which is conducive to better charging and discharging control of electric vehicles.
  • the minimum target charging variance is used to ensure that the fluctuations in the power grid caused by the electric vehicle cluster are small, which is beneficial to protecting the power grid.
  • the minimum target discharge variance is used to ensure that the discharge fluctuations of each electric vehicle in the electric vehicle cluster are small, which is beneficial to protecting electric vehicles and eliminating waste. Maximizing the amount of energy absorbed will help improve the utilization rate of renewable energy.
  • the distance between the electric vehicle and each charging area in the corresponding driving route can also be combined to construct a control target that minimizes the total charging distance, so as to reduce the distance (or time) required for the electric vehicle to reach the charging area, and then Reduce consumption during charging.
  • the maximum number of vehicles that each charging area can accommodate can also be used as a constraint to prevent a large number of electric vehicles from flooding into the same charging area, which is conducive to reasonable arrangement of the number of charging piles in the charging area.
  • an embodiment of the present invention provides an electric vehicle cluster charging and discharging control method. Specifically, the above method includes the following steps:
  • Step S100 Obtain the target power generation amount of renewable energy within the current target time period.
  • the above-mentioned current target time period is the time period when cluster charging and discharging control of electric vehicles is required.
  • the time length of the above-mentioned current time period is 24 hours.
  • the current time period is a time period corresponding to 24 hours starting from the current time and going forward.
  • the above-mentioned current time period is divided into multiple non-overlapping prediction time segments.
  • the above-mentioned current time period is divided into 24 prediction time segments, and the time length of each prediction time segment is 1 Hour. It should be noted that the above-mentioned current target time period and the time length of each forecast time segment can be set and adjusted according to actual needs, and are not specifically limited here.
  • the target power generation amount of renewable energy in the current target time period is the amount of power that can be obtained through renewable energy power generation in a preset target area within the current target time period.
  • the above-mentioned target power generation may be a value preset based on historical power generation (for example, the average power generation of renewable energy sources every 24 hours determined based on historical data).
  • the above-mentioned target power generation may also be a value based on the current target time.
  • the amount of electricity that can be generated by renewable energy is predicted based on the weather conditions in the next 24 hours.
  • the above-mentioned renewable energy sources include wind energy and solar energy in the preset target area.
  • the above-mentioned step S100 specifically includes the following steps:
  • Step S101 Obtain the environment prediction information within the current target time period, where the environment prediction information includes wind intensity, wind direction, light intensity and light time.
  • Step S102 Obtain the target power generation amount based on the environmental prediction information.
  • the above environmental prediction information can be obtained through meteorological forecast data. After obtaining the above environmental prediction information, the impact of the environmental prediction information on power generation can be comprehensively considered through a preset calculation formula or a pre-trained power generation prediction model to obtain the target power generation.
  • obtaining the target power generation amount based on the environment prediction information includes: inputting the environment prediction information into a pre-trained power generation prediction model, and using the predicted power generation amount output by the trained power generation prediction model as the above-mentioned The target power generation amount within the current target time period.
  • the corresponding target power generation is obtained through a pre-trained power generation prediction model, where the above power generation prediction model is pre-trained according to the following steps:
  • one of the real power generation data includes actual environmental information within a power generation time period and the actual power generation amount within the power generation time period.
  • the above actual environmental information includes wind intensity, wind direction, and light intensity. and light time;
  • the above-mentioned power generation prediction model is trained according to the above-mentioned training power generation data set, and the trained power generation prediction model is tested through the above-mentioned test power generation data set until a trained power generation prediction model is obtained, wherein the above-mentioned test power generation data set is concentrated
  • the loss value between the predicted power generation output by the above-trained power generation prediction model and the actual power generation corresponding to the actual environmental information is not greater than Preset loss threshold for power generation forecast.
  • the time length of the above-mentioned power generation time period is the same as the time length of the above-mentioned current target time period, that is, the time length of the power generation time period in this embodiment is also 24 hours.
  • any real power generation data obtained includes the wind intensity, wind direction, light intensity, light time and corresponding actual power generation in the area in any 24 hours in the past. In this way, an environment can be established based on these real power generation data. The relationship between information and power generation.
  • the training power generation data set and the test power generation data set are divided according to a preset ratio. For example, 80% of the data is used as the training power generation data set, and 2% is used as the training power generation data set. Ten data are used as test power generation data set.
  • the power generation prediction model is trained based on the training power generation data set, and the model parameters are adjusted during the training process.
  • the above power generation prediction model is set with corresponding calculation formulas for wind power generation and photovoltaic power generation, and some of the parameters are adjusted through training.
  • the power generation prediction model is tested on the test power generation data set.
  • the loss value between the predicted power generation and the actual power generation is calculated through a preset loss formula.
  • the calculated loss value is not greater than the preset Training is considered complete when the generation prediction loss threshold is reached.
  • the training is also considered completed when the number of training iterations is greater than the preset power generation training iteration threshold.
  • Step S200 obtain the time period label corresponding to the current target time period and the power consumption history data of each electric vehicle in the electric vehicle cluster, input the above time period label and the above power consumption history data into the pre-trained Transformer model, and use the above-mentioned pre-trained
  • the Transformer model obtains the power consumption prediction data corresponding to each of the above-mentioned electric vehicles in the above-mentioned current target time period, and obtains the predicted overall power consumption of the above-mentioned electric vehicle cluster in the above-mentioned current target time period.
  • the power consumption history data of the above-mentioned electric vehicle includes the actual power consumption of the electric vehicle in each historical time segment in the previous target time period
  • the power consumption prediction data of the above-mentioned electric vehicle includes the electric vehicle in the above-mentioned current target time. The predicted power consumption for each predicted time segment within the segment.
  • the above time period label is a label data determined based on the characteristics of the current target time period. Since the user's car usage habits are different on different types of days (such as working days or weekends), the corresponding power consumption habits are also different, so they can be combined Period labels predict power consumption. At the same time, considering that users’ electricity consumption habits are continuous, the electricity consumption in the next period (i.e. today) can be predicted based on the electricity consumption in the previous period (i.e., the previous day). Combining time period tags and power consumption history data can make power consumption predictions more accurate.
  • the overall power consumption of the electric vehicle cluster in the entire current target time period is not directly predicted, but is divided into small time periods. (i.e., prediction time segments, the length of each prediction time segment is 1 hour) to predict the power consumption of each electric vehicle (i.e., predicted power consumption), and then calculate the current target of the entire electric vehicle cluster The predicted overall power consumption during the time period. In this way, forecasting for individual entities and distinguishing small time periods can make the forecast results more accurate.
  • the historical power consumption data input to the Transformer model is also the actual power consumption of electric vehicles in each small historical time segment in the previous target time period.
  • the above-mentioned previous target time period is a time period before the current target time period, the two time periods are the same, and the end time of the previous target time period is not earlier than the start time of the current target time period.
  • the time length of the above historical time segment is also the same as the time length of the forecast time segment (i.e. 1 hour).
  • the time period tag is any one of preset tag data, and the tag data includes working days, weekends, and special holidays.
  • the above-mentioned special holidays may also include specific holiday names, such as Mid-Autumn Festival, Dragon Boat Festival, National Day, etc. This is because users have different travel habits and electricity consumption habits during different holidays. For example, there are fewer trips during the Mid-Autumn Festival and electric vehicles consume less electricity, while there is a greater possibility of traveling during the National Day and electric vehicles consume more electricity.
  • time period tags users' habits at different times can be considered to improve the accuracy of power consumption prediction.
  • the above Transformer model is pre-trained according to the following steps:
  • one of the above real power consumption data includes a time period label corresponding to a power consumption time period, a real power consumption label and power consumption history collection data, and the above real power consumption label includes each The actual power consumption of the above-mentioned electric vehicles in each power consumption time segment of the above-mentioned power consumption time period.
  • the above-mentioned power consumption history collection data includes each power consumption history collection time of each of the above-mentioned electric vehicles in the power consumption history collection time period.
  • the actual power consumption within the segment, the above-mentioned power consumption history collection time period is the previous time period corresponding to the above-mentioned power consumption time period;
  • the above-mentioned Transformer model is trained according to the above-mentioned training power consumption data set, and the trained Transformer model is tested through the above-mentioned test power consumption data set until a trained Transformer model is obtained, in which any one of the above-mentioned test power consumption data sets is
  • the time period tags and power consumption history collection data in the real power consumption data are used as the input data of the above-trained Transformer model, the power consumption prediction data output by the above-trained Transformer model is consistent with the real power consumption in the real power consumption data.
  • the loss value between tags is not greater than the preset power consumption prediction loss threshold.
  • the length of the above power consumption time period is the same as the current target time period, that is, it is also 24 hours, and each power consumption time segment (or power consumption history collection time segment) is also the same length as a forecast time segment, which is 1 hour.
  • the time segments corresponding to each time segment are divided in the same manner.
  • the above power consumption history collection time period is the 24 hours before the power consumption time period.
  • the training power consumption data set and the test power consumption data set are divided according to a preset ratio. For example, 80% of the data is used as the training power consumption data set. Twenty percent of the data is used as the test power consumption data set.
  • the preset Transformer model is trained based on the above training power consumption data set, and the model parameters are adjusted during the training process. And test the Transformer model through the test power consumption data set. During the test, the loss value is calculated according to the preset power consumption loss formula.
  • the power consumption training iteration threshold can also be set to determine whether the training is completed.
  • Step S300 Construct a comprehensive control objective and control constraints. According to the above comprehensive control objectives and the above control constraints, optimize and solve through the preset particle swarm algorithm and obtain the charge and discharge control strategy of each of the above electric vehicles.
  • the above-mentioned comprehensive control objectives include the minimum target charging variance, the minimum target discharge variance and the maximum power consumption.
  • the above-mentioned target charging variance is the variance of the charging amount of the above-mentioned electric vehicle cluster in the above-mentioned current target time period.
  • the above-mentioned target discharge variance is the above-mentioned The variance of the discharge amount of the electric vehicle cluster in the above-mentioned current target time period.
  • the above-mentioned consumption amount is the planned overall charging amount of the above-mentioned electric vehicle cluster in the above-mentioned current target time period.
  • the above-mentioned control constraints include cluster charging amount range constraints and cluster
  • the above-mentioned cluster charging range constraint is used to limit the above-mentioned consumption amount to not be less than the above-mentioned target power generation
  • the above-mentioned cluster discharge amount range constraint is used to limit the planned overall discharge amount of the above-mentioned electric vehicle cluster within the above-mentioned current target time period. Not less than the overall power consumption predicted above.
  • the distance between the electric vehicle and the planned charging area during charging can also be considered.
  • the above-mentioned comprehensive control target also includes minimizing the total charging distance of the above-mentioned electric vehicle cluster
  • the above-mentioned method also includes: obtaining the location of the charging area and the target driving of each of the above-mentioned electric vehicles in the above-mentioned electric vehicle cluster within the above-mentioned current target time period. Route, calculate the charging distance data of each of the above-mentioned electric vehicles in each of the above-mentioned target time segments based on the above-mentioned charging area position and the above-mentioned target driving route.
  • the above-mentioned electric vehicles are electric buses, and the electric vehicle cluster is the corresponding bus cluster. Therefore, the target driving route corresponding to each electric vehicle can be determined based on the driving route of the bus. In another application scenario, the target driving route of the electric vehicle can be directly input by the user, or the target driving route can be predicted based on the historical driving data and driving habits of each electric vehicle, which is not specifically limited here.
  • the charging distance data corresponding to an electric vehicle includes the distance of the electric vehicle from each charging area at the starting time of each target time segment.
  • the above control constraints may also include the maximum number of vehicles admitted in each charging area.
  • the number of electric vehicles planned in each charging area is not greater than the preset maximum number of vehicles in the charging area. Accept the number of vehicles to avoid congestion in a certain charging area.
  • the above-mentioned comprehensive control objectives may also include minimizing the variance of the discharge power and minimizing the variance of the charging power of the electric vehicle cluster within the above-mentioned current target time period, which are not specifically limited here.
  • the preset particle swarm algorithm is used to optimize the solution and obtain the charge and discharge control strategy of each of the above electric vehicles, including:
  • An objective function is constructed according to the above comprehensive control objective, wherein the above objective function is the sum of the inverse number of the above consumed power, the above target charging variance, the above target discharge variance and the above total charging distance;
  • the above particle swarm algorithm is used to optimize the solution and obtain the charge and discharge control strategy of each of the above electric vehicles.
  • the above-mentioned preset particle swarm algorithm may be a preset multi-objective particle swarm optimization algorithm, such as social learning particle swarm algorithm.
  • the above charging and discharging control strategy includes the charging power (or charging amount) and discharging power (and discharging amount) of each electric vehicle in each prediction time segment, where the discharging power is 0 when only charging, and the charging power is 0 when only discharging. .
  • each electric vehicle can also be set to only select one predicted time segment for charging as a constraint to reduce the number of times the user charges and avoid frequent charging.
  • the corresponding charging and discharging strategy includes the electric vehicle in a target charging time segment, the charging power in the target charging time segment, and the discharging power in other time segments.
  • one of the goals in this embodiment is to consume more new energy electricity.
  • the overall discharge amount of the plan is not less than the above predicted overall power consumption, which can consume more new energy.
  • Energy-saving electricity can ensure that electric vehicles will not be short of power as much as possible.
  • users can also adjust the discharge amount according to actual needs, and unused electricity can be stored in electric vehicles.
  • Step S400 Control each of the above-mentioned electric vehicles in the above-mentioned electric vehicle cluster according to the above-mentioned charge and discharge control strategy.
  • each electric vehicle in the above-mentioned electric vehicle cluster is controlled according to the above charge and discharge control strategy.
  • its charging power and discharge power in each predicted time segment are controlled according to the charge and discharge control strategy.
  • the above charge and discharge control strategy can also include specific charging areas, thereby controlling the electric vehicle to reach the designated charging area for charging in corresponding predicted time segments.
  • the charge and discharge control strategy solved in this embodiment is a local optimal solution, so the final control strategy may not necessarily satisfy all objectives, but only needs to satisfy the minimum function value of the objective function. Therefore, the electric energy generated by the above-mentioned renewable energy sources may not necessarily meet the demand, or may not be fully consumed.
  • supercapacitors can be used for energy storage. The stored energy can be discharged during peak power consumption periods to relieve the pressure on the power system and reduce power generation costs.
  • the electric energy generated by renewable energy cannot meet the demand, the electric energy in the supercapacitor can be released, or the electric energy can be obtained from the conventional power grid to meet the demand.
  • the target charging variance is limited to the minimum, so the electric energy can be obtained from the conventional power grid.
  • the process of electric energy is also relatively stable and will not cause violent fluctuations, which is helpful to avoid the fluctuations of traditional power grids.
  • the target power generation of renewable energy and the predicted overall power consumption of the electric vehicle cluster within the current target time period are comprehensively considered to construct a comprehensive control target and control constraints for the electric vehicle cluster, with the target Minimizing the charging variance, minimizing the target discharging variance and maximizing the power consumption are the goals to optimize the solution and obtain the optimized charging and discharging control strategy, which is conducive to better charging and discharging control of electric vehicles.
  • the minimum target charging variance is used to ensure that the fluctuations in the power grid caused by the electric vehicle cluster are small, which is beneficial to protecting the power grid.
  • the minimum target discharge variance is used to ensure that the discharge fluctuations of each electric vehicle in the electric vehicle cluster are small, which is beneficial to protecting electric vehicles and eliminating waste. Maximizing the amount of energy absorbed will help improve the utilization rate of renewable energy.
  • inventions of the present invention also provide an electric vehicle cluster charge and discharge control system.
  • the above-mentioned electric vehicle cluster charge and discharge control system includes:
  • the power generation acquisition module 510 is used to obtain the target power generation of renewable energy within the current target time period.
  • the power consumption prediction module 520 is used to obtain the time period label corresponding to the current target time period and the power consumption history data of each electric vehicle in the electric vehicle cluster, and input the above time period label and the above power consumption history data into the pre-trained Transformer model , obtain the power consumption prediction data corresponding to each of the above-mentioned electric vehicles in the above-mentioned current target time period through the above-mentioned pre-trained Transformer model, and obtain the predicted overall power consumption of the above-mentioned electric vehicle cluster in the above-mentioned current target time period, wherein, the above-mentioned
  • the historical power consumption data of electric vehicles includes the actual power consumption of the electric vehicle in each historical time segment in the previous target time period.
  • the power consumption forecast data of the above-mentioned electric vehicle includes the electric vehicle’s actual power consumption in each historical time period in the above-mentioned current target time period. Predicted power consumption for a forecast time segment.
  • the strategy acquisition module 530 is used to construct comprehensive control objectives and control constraints. According to the above comprehensive control objectives and the above control constraints, optimize and solve through the preset particle swarm algorithm and obtain the charge and discharge control strategy of each of the above electric vehicles, where , the above-mentioned comprehensive control objectives include the minimum target charging variance, the minimum target discharge variance and the maximum power consumption.
  • the above-mentioned target charging variance is the variance of the charging amount of the above-mentioned electric vehicle cluster in the above-mentioned current target time period.
  • the above-mentioned target discharge variance is the above-mentioned electric vehicle cluster. The variance of the discharge amount of the vehicle cluster in the above-mentioned current target time period.
  • the above-mentioned consumption amount is the planned overall charging amount of the above-mentioned electric vehicle cluster in the above-mentioned current target time period.
  • the above-mentioned control constraints include cluster charging range constraints and cluster discharge.
  • the above-mentioned cluster charging capacity range constraint is used to limit the above-mentioned consumption amount of electricity to not be less than the above-mentioned target power generation amount
  • the above-mentioned cluster discharge amount range constraint is used to limit the planned overall discharge amount of the above-mentioned electric vehicle cluster within the above-mentioned current target time period. Less than the overall power consumption predicted above.
  • the control module 540 is used to control each of the above-mentioned electric vehicles in the above-mentioned electric vehicle cluster according to the above-mentioned charge and discharge control strategy.
  • the present invention also provides an intelligent terminal, the functional block diagram of which can be shown in Figure 4 .
  • the above-mentioned smart terminal includes a processor and a memory.
  • the memory of the smart terminal includes an electric vehicle cluster charge and discharge control program, and the memory provides an environment for the operation of the electric vehicle cluster charge and discharge control program.
  • the electric vehicle cluster charge and discharge control program is executed by the processor, the steps of any of the above electric vehicle cluster charge and discharge control methods are implemented.
  • the above-mentioned smart terminal may also include other functional modules or units, which are not specifically limited here.
  • Embodiments of the present invention also provide a computer-readable storage medium.
  • the computer-readable storage medium stores a CNC machine tool spindle error prediction and compensation program.
  • the CNC machine tool spindle error prediction and compensation program is executed by a processor, the present invention is implemented.
  • the steps of any electric vehicle cluster charging and discharging control method are provided in the example.
  • sequence number of each step in the above embodiment does not mean the order of execution.
  • the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.
  • Module completion means dividing the internal structure of the above system into different functional units or modules to complete all or part of the functions described above.
  • Each functional unit and module in the embodiment can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit.
  • the above-mentioned integrated unit can be hardware-based. It can also be implemented in the form of software functional units.
  • the specific names of each functional unit and module are only for the convenience of distinguishing each other and are not used to limit the scope of the present invention.
  • For the specific working processes of the units and modules in the above system please refer to the corresponding processes in the foregoing method embodiments, and will not be described again here.
  • system/intelligent terminal and method can be implemented in other ways.
  • system/intelligent terminal embodiments described above are only illustrative.
  • the division of the above modules or units is only a logical function division. In actual implementation, it can be divided in other ways, such as multiple units or units. Components may be combined or may be integrated into another system, or some features may be ignored, or not implemented.
  • the above-mentioned integrated modules/units are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the present invention can implement all or part of the processes in the above-mentioned embodiment methods, and can also be completed by instructing relevant hardware through a computer program.
  • the above-mentioned computer program can be stored in a computer-readable storage medium.
  • the computer program can be stored in a computer-readable storage medium. When executed by the processor, the steps of each of the above method embodiments can be implemented.
  • the above-mentioned computer program includes computer program code, and the above-mentioned computer program code may be in the form of source code, object code, executable file or some intermediate form, etc.
  • the above-mentioned computer-readable media may include: any entity or device capable of carrying the above-mentioned computer program code, recording media, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electrical carrier signals, telecommunications signals, and software distribution media, etc. It should be noted that the content contained in the above computer-readable storage media can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Economics (AREA)
  • General Health & Medical Sciences (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • General Business, Economics & Management (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Tourism & Hospitality (AREA)
  • General Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Marketing (AREA)
  • Molecular Biology (AREA)
  • Evolutionary Computation (AREA)
  • Data Mining & Analysis (AREA)
  • Computational Linguistics (AREA)
  • Biophysics (AREA)
  • Biomedical Technology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Mechanical Engineering (AREA)
  • Quality & Reliability (AREA)
  • Operations Research (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Game Theory and Decision Science (AREA)
  • Development Economics (AREA)
  • Power Engineering (AREA)
  • Transportation (AREA)
  • Water Supply & Treatment (AREA)
  • Primary Health Care (AREA)
  • Public Health (AREA)
  • Charge And Discharge Circuits For Batteries Or The Like (AREA)

Abstract

本发明公开了一种电动汽车集群充放电控制方法、系统及相关设备，方法包括：获取当前目标时间段内可再生能源的目标发电量；获取当前目标时间段对应的时段标签和电动汽车集群中各电动汽车的耗电历史数据，将时段标签和耗电历史数据输入预先训练好的Transformer模型，获取各电动汽车在当前目标时间段对应的耗电预测数据，并获得电动汽车集群在当前目标时间段内的预测总体耗电量；构建综合控制目标和控制约束条件，根据综合控制目标和控制约束条件，通过预设的粒子群算法进行优化求解并获得各电动汽车的充放电控制策略；根据充放电控制策略对电动汽车集群中的各电动汽车进行控制。本发明方案有利于对电动汽车进行更好的充放电控制。

Description

电动汽车集群充放电控制方法、系统及相关设备 技术领域
本发明涉及电动汽车充放电调度技术领域，尤其涉及的是一种电动汽车集群充放电控制方法、系统及相关设备。
 背景技术
随着科学技术的发展，电动汽车的使用越来越广泛。一方面，电动汽车可以作为传统化石燃料汽车的替代品，另一方面，电动汽车可以使用可再生能源所发的电进行充电，有利于减少环境污染。
现有技术中，对于电动汽车缺乏合理的充放电控制，电动汽车的充放电是随意的。现有技术的问题在于，电动汽车的充放电过程缺乏合理的安排和调控，给电网带来的波动大，并且不能充分利用可再生能源所发的电，不利于提高可再生能源的利用率。
因此，现有技术还有待改进和发展。
 技术问题
本发明的主要目的在于提供一种电动汽车集群充放电控制方法、系统及相关设备，旨在解决现有技术中电动汽车的充放电过程缺乏合理的安排和调控的问题。
 技术解决方案
为了实现上述目的，本发明第一方面提供一种电动汽车集群充放电控制方法，其中，上述电动汽车集群充放电控制方法包括：
获取当前目标时间段内可再生能源的目标发电量；
获取上述当前目标时间段对应的时段标签和电动汽车集群中各电动汽车的耗电历史数据，将上述时段标签和上述耗电历史数据输入预先训练好的Transformer模型，通过上述预先训练好的Transformer模型获取各上述电动汽车在上述当前目标时间段对应的耗电预测数据，并获得上述电动汽车集群在上述当前目标时间段内的预测总体耗电量，其中，上述电动汽车的耗电历史数据包括该电动汽车在前一目标时间段内每一个历史时间分段的实际耗电量，上述电动汽车的耗电预测数据包括该电动汽车在上述当前目标时间段内每一个预测时间分段的预测耗电量；
构建综合控制目标和控制约束条件，根据上述综合控制目标和上述控制约束条件，通过预设的粒子群算法进行优化求解并获得各上述电动汽车的充放电控制策略，其中，上述综合控制目标包括目标充电方差最小、目标放电方差最小和消纳电量最大，上述目标充电方差是上述电动汽车集群在上述当前目标时间段内的充电量的方差，上述目标放电方差是上述电动汽车集群在上述当前目标时间段内的放电量的方差，上述消纳电量是上述电动汽车集群在上述当前目标时间段内的规划总体充电量，上述控制约束条件包括集群充电量范围约束和集群放电量范围约束，上述集群充电量范围约束用于限制上述消纳电量不小于上述目标发电量，上述集群放电量范围约束用于限制上述电动汽车集群在上述当前目标时间段内的规划总体放电量不小于上述预测总体耗电量；
根据上述充放电控制策略对上述电动汽车集群中的各上述电动汽车进行控制。
可选的，上述可再生能源包括预设的目标区域内的风能和太阳能，上述获取当前目标时间段内可再生能源的目标发电量，包括：
获取上述当前目标时间段内的环境预测信息，其中，上述环境预测信息包括风力强度、风向、光照强度和光照时间；
根据上述环境预测信息获取上述目标发电量。
可选的，上述根据上述环境预测信息获取上述目标发电量，包括：
将上述环境预测信息输入预先训练好的发电量预测模型，将上述训练好的发电量预测模型输出的预测发电量作为上述当前目标时间段内的目标发电量。
可选的，上述发电量预测模型根据如下步骤进行预先训练：
获取预先采集的多个真实发电数据，其中，一个上述真实发电数据包括一个发电时间段内的实际环境信息和该发电时间段内的实际发电量，上述实际环境信息包括风力强度、风向、光照强度和光照时间；
对上述真实发电数据划分获得训练发电数据集和测试发电数据集；
根据上述训练发电数据集对上述发电量预测模型进行训练，并通过上述测试发电数据集对训练的发电量预测模型进行测试，直到获得训练好的发电量预测模型，其中，将上述测试发电数据集中的任意一个实际环境信息作为上述训练好的发电量预测模型的输入数据时，上述训练好的发电量预测模型输出的预测发电量与该实际环境信息对应的实际发电量之间的损失值不大于预设的发电预测损失阈值。
可选的，上述时段标签是预先设置的标签数据中的任意一种，上述标签数据包括工作日、周末和特殊节假日，上述Transformer模型根据如下步骤进行预先训练：
获取预先采集的多个真实耗电数据，其中，一个上述真实耗电数据包括一个耗电时间段对应的时段标签、真实耗电量标签和耗电历史采集数据，上述真实耗电量标签包括各上述电动汽车在上述耗电时间段的每一个耗电时间分段内的实际耗电量，上述耗电历史采集数据包括各上述电动汽车在耗电历史采集时间段的每一个耗电历史采集时间分段内的实际耗电量，上述耗电历史采集时间段是上述耗电时间段对应的前一个时间段；
对上述真实耗电数据划分获得训练耗电数据集和测试耗电数据集；
根据上述训练耗电数据集对上述Transformer模型进行训练，并通过上述测试耗电数据集对训练的Transformer模型进行测试，直到获得训练好的Transformer模型，其中，将上述测试耗电数据集中的任意一个真实耗电数据中的时段标签和耗电历史采集数据作为上述训练好的Transformer模型的输入数据时，上述训练好的Transformer模型输出的耗电预测数据与该真实耗电数据中的真实耗电量标签的之间的损失值不大于预设的耗电预测损失阈值。
可选的，上述综合控制目标还包括上述电动汽车集群的充电总距离最小，上述方法还包括：
获取充电区域位置以及上述电动汽车集群中各上述电动汽车在上述当前目标时间段内的目标行驶路线，根据上述充电区域位置与上述目标行驶路线计算各上述电动汽车在各上述目标时间分段的充电距离数据。
可选的，上述根据上述综合控制目标和上述控制约束条件，通过预设的粒子群算法进行优化求解并获得各上述电动汽车的充放电控制策略，包括：
根据上述综合控制目标构建目标函数，其中，上述目标函数是上述消纳电量的相反数、上述目标充电方差、上述目标放电方差以及上述充电总距离之和；
以上述目标函数取得最小的函数值为目标，根据上述控制约束条件，通过上述粒子群算法进行优化求解并获得各上述电动汽车的充放电控制策略。
本发明第二方面提供一种电动汽车集群充放电控制系统，其中，上述电动汽车集群充放电控制系统包括：
发电量获取模块，用于获取当前目标时间段内可再生能源的目标发电量；
耗电量预测模块，用于获取上述当前目标时间段对应的时段标签和电动汽车集群中各电动汽车的耗电历史数据，将上述时段标签和上述耗电历史数据输入预先训练好的Transformer模型，通过上述预先训练好的Transformer模型获取各上述电动汽车在上述当前目标时间段对应的耗电预测数据，并获得上述电动汽车集群在上述当前目标时间段内的预测总体耗电量，其中，上述电动汽车的耗电历史数据包括该电动汽车在前一目标时间段内每一个历史时间分段的实际耗电量，上述电动汽车的耗电预测数据包括该电动汽车在上述当前目标时间段内每一个预测时间分段的预测耗电量；
策略获取模块，用于构建综合控制目标和控制约束条件，根据上述综合控制目标和上述控制约束条件，通过预设的粒子群算法进行优化求解并获得各上述电动汽车的充放电控制策略，其中，上述综合控制目标包括目标充电方差最小、目标放电方差最小和消纳电量最大，上述目标充电方差是上述电动汽车集群在上述当前目标时间段内的充电量的方差，上述目标放电方差是上述电动汽车集群在上述当前目标时间段内的放电量的方差，上述消纳电量是上述电动汽车集群在上述当前目标时间段内的规划总体充电量，上述控制约束条件包括集群充电量范围约束和集群放电量范围约束，上述集群充电量范围约束用于限制上述消纳电量不小于上述目标发电量，上述集群放电量范围约束用于限制上述电动汽车集群在上述当前目标时间段内的规划总体放电量不小于上述预测总体耗电量；
控制模块，用于根据上述充放电控制策略对上述电动汽车集群中的各上述电动汽车进行控制。
本发明第三方面提供一种智能终端，上述智能终端包括存储器、处理器以及存储在上述存储器上并可在上述处理器上运行的电动汽车集群充放电控制程序，上述电动汽车集群充放电控制程序被上述处理器执行时实现上述任意一种电动汽车集群充放电控制方法的步骤。
本发明第四方面提供一种计算机可读存储介质，上述计算机可读存储介质上存储有电动汽车集群充放电控制程序，上述电动汽车集群充放电控制程序被处理器执行时实现上述任意一种电动汽车集群充放电控制方法的步骤。
由上可见，本发明方案中，获取当前目标时间段内可再生能源的目标发电量；获取上述当前目标时间段对应的时段标签和电动汽车集群中各电动汽车的耗电历史数据，将上述时段标签和上述耗电历史数据输入预先训练好的Transformer模型，通过上述预先训练好的Transformer模型获取各上述电动汽车在上述当前目标时间段对应的耗电预测数据，并获得上述电动汽车集群在上述当前目标时间段内的预测总体耗电量，其中，上述电动汽车的耗电历史数据包括该电动汽车在前一目标时间段内每一个历史时间分段的实际耗电量，上述电动汽车的耗电预测数据包括该电动汽车在上述当前目标时间段内每一个预测时间分段的预测耗电量；构建综合控制目标和控制约束条件，根据上述综合控制目标和上述控制约束条件，通过预设的粒子群算法进行优化求解并获得各上述电动汽车的充放电控制策略，其中，上述综合控制目标包括目标充电方差最小、目标放电方差最小和消纳电量最大，上述目标充电方差是上述电动汽车集群在上述当前目标时间段内的充电量的方差，上述目标放电方差是上述电动汽车集群在上述当前目标时间段内的放电量的方差，上述消纳电量是上述电动汽车集群在上述当前目标时间段内的规划总体充电量，上述控制约束条件包括集群充电量范围约束和集群放电量范围约束，上述集群充电量范围约束用于限制上述消纳电量不小于上述目标发电量，上述集群放电量范围约束用于限制上述电动汽车集群在上述当前目标时间段内的规划总体放电量不小于上述预测总体耗电量；根据上述充放电控制策略对上述电动汽车集群中的各上述电动汽车进行控制。
 有益效果
与现有技术相比，本发明方案中综合考虑在当前目标时间段内可再生能源的目标发电量和电动汽车集群的预测总体耗电量，构建对于电动汽车集群的综合控制目标和控制约束条件，以目标充电方差最小、目标放电方差最小和消纳电量最大为目标进行优化求解并获得优化的充放电控制策略，有利于对电动汽车进行更好的充放电控制。其中，目标充电方差最小用于保证电动汽车集群带来的电网波动小，有利于保护电网，目标放电方差最小用于保证电动汽车集群中各个电动汽车的放电波动小，有利于保护电动汽车，消纳电量最大则有利于提高可再生能源的利用率。
 附图说明
为了更清楚地说明本发明实施例中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动性的前提下，还可以根据这些附图获得其它的附图。
图1是本发明实施例提供的一种电动汽车集群充放电控制方法的流程示意图；
图2是本发明实施例图1中步骤S100的具体流程示意图；
图3是本发明实施例提供的一种电动汽车集群充放电控制系统的结构示意图；
图4是本发明实施例提供的一种智能终端的内部结构原理框图。
 本发明的实施方式
下面结合本发明实施例的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本发明的一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其它实施例，都属于本发明保护的范围。
在下面的描述中阐述了很多具体细节以便于充分理解本发明，但是本发明还可以采用其它不同于在此描述的其它方式来实施，本领域技术人员可以在不违背本发明内涵的情况下做类似推广，因此本发明不受下面公开的具体实施例的限制。
随着科学技术的发展，电动汽车的使用越来越广泛。一方面，电动汽车可以作为传统化石燃料汽车的替代品，另一方面，电动汽车可以使用可再生能源所发的电进行充电，有利于减少环境污染。
现有技术中，对于电动汽车缺乏合理的充放电控制，电动汽车的充放电是随意的。现有技术的问题在于，电动汽车的充放电过程缺乏合理的安排和调控，给电网带来的波动大，并且不能充分利用可再生能源所发的电，不利于提高可再生能源的利用率。
在一种应用场景中，用户随意选择电动车的放电功率、充电时间和充电区域，可能导致在行驶过程中缺电。或者需要行驶较长的距离才能达到用户选的充电区域，造成时间的浪费，甚至可能在到达充电区域之后发现没有充电位（或充电桩），从而影响用户使用。在另一种应用场景中，为了保证用户在到达充电区域之后能够进行充电，需要在各个充电区域设置过量的充电桩，从而导致充电桩过剩。
为了解决上述多个问题中的至少一个问题，本发明方案中，获取当前目标时间段内可再生能源的目标发电量；获取上述当前目标时间段对应的时段标签和电动汽车集群中各电动汽车的耗电历史数据，将上述时段标签和上述耗电历史数据输入预先训练好的Transformer模型，通过上述预先训练好的Transformer模型获取各上述电动汽车在上述当前目标时间段对应的耗电预测数据，并获得上述电动汽车集群在上述当前目标时间段内的预测总体耗电量，其中，上述电动汽车的耗电历史数据包括该电动汽车在前一目标时间段内每一个历史时间分段的实际耗电量，上述电动汽车的耗电预测数据包括该电动汽车在上述当前目标时间段内每一个预测时间分段的预测耗电量；构建综合控制目标和控制约束条件，根据上述综合控制目标和上述控制约束条件，通过预设的粒子群算法进行优化求解并获得各上述电动汽车的充放电控制策略，其中，上述综合控制目标包括目标充电方差最小、目标放电方差最小和消纳电量最大，上述目标充电方差是上述电动汽车集群在上述当前目标时间段内的充电量的方差，上述目标放电方差是上述电动汽车集群在上述当前目标时间段内的放电量的方差，上述消纳电量是上述电动汽车集群在上述当前目标时间段内的规划总体充电量，上述控制约束条件包括集群充电量范围约束和集群放电量范围约束，上述集群充电量范围约束用于限制上述消纳电量不小于上述目标发电量，上述集群放电量范围约束用于限制上述电动汽车集群在上述当前目标时间段内的规划总体放电量不小于上述预测总体耗电量；根据上述充放电控制策略对上述电动汽车集群中的各上述电动汽车进行控制。
与现有技术相比，本发明方案中综合考虑在当前目标时间段内可再生能源的目标发电量和电动汽车集群的预测总体耗电量，构建对于电动汽车集群的综合控制目标和控制约束条件，以目标充电方差最小、目标放电方差最小和消纳电量最大为目标进行优化求解并获得优化的充放电控制策略，有利于对电动汽车进行更好的充放电控制。其中，目标充电方差最小用于保证电动汽车集群带来的电网波动小，有利于保护电网，目标放电方差最小用于保证电动汽车集群中各个电动汽车的放电波动小，有利于保护电动汽车，消纳电量最大则有利于提高可再生能源的利用率。
进一步的，本发明中，还可以结合电动汽车在对应的行驶路线中与各个充电区域的距离构建充电总距离最小的控制目标，以减少电动汽车到达充电区域所需要的距离（或时间），进而减少充电过程中的消耗。进一步的，也可以将各个充电区域能容纳的最大车辆数作为约束条件，避免大量的电动汽车涌入同一充电区域，有利于对充电区域内的充电桩数目进行合理安排。
示例性方法
如图1所示，本发明实施例提供一种电动汽车集群充放电控制方法，具体的，上述方法包括如下步骤：
步骤S100，获取当前目标时间段内可再生能源的目标发电量。
其中，上述当前目标时间段是需要进行电动汽车集群充放电控制的时间段。本实施例中，上述当前时间段的时间长度为24小时，具体的，当前时间段是从当前时刻开始并往后的24个小时对应的时间段。上述当前时间段被划分为多个互不重合的预测时间分段，例如，本实施例中，上述当前时间段被划分为24个预测时间分段，每一个预测时间分段的时间长度为1小时。需要说明的是，上述当期目标时间段和各个预测时间分段的时间长度可以根据实际需求进行设置和调整，在此不作具体限定。
可选的，上述当前目标时间段内可再生能源的目标发电量是在当前目标时间段内预先设置的一个目标区域内通过可再生能源发电可以获得的电量。上述目标发电量可以是根据历史发电量预先设置的一个值（例如根据历史数据确定的可再生能源每24小时的平均发电量），本实施例中，上述目标发电量还可以是根据当前目标时间段（即未来24小时）内的天气情况预测的可再生能源可发电量。
具体的，本实施例中，上述可再生能源包括预设的目标区域内的风能和太阳能，如图2所示，上述步骤S100具体包括如下步骤：
步骤S101，获取上述当前目标时间段内的环境预测信息，其中，上述环境预测信息包括风力强度、风向、光照强度和光照时间。
步骤S102，根据上述环境预测信息获取上述目标发电量。
其中，上述环境预测信息可以通过气象预报数据获取。获取到上述环境预测信息之后，可以通过预先设置的计算公式或者预先训练好的发电量预测模型综合考虑环境预测信息对发电量的影响，从而获取目标发电量。
本实施例中，上述根据上述环境预测信息获取上述目标发电量，包括：将上述环境预测信息输入预先训练好的发电量预测模型，将上述训练好的发电量预测模型输出的预测发电量作为上述当前目标时间段内的目标发电量。
即本实施例中通过预先训练好的发电量预测模型获取对应的目标发电量，其中，上述发电量预测模型根据如下步骤进行预先训练：
获取预先采集的多个真实发电数据，其中，一个上述真实发电数据包括一个发电时间段内的实际环境信息和该发电时间段内的实际发电量，上述实际环境信息包括风力强度、风向、光照强度和光照时间；
对上述真实发电数据划分获得训练发电数据集和测试发电数据集；
根据上述训练发电数据集对上述发电量预测模型进行训练，并通过上述测试发电数据集对训练的发电量预测模型进行测试，直到获得训练好的发电量预测模型，其中，将上述测试发电数据集中的任意一个实际环境信息作为上述训练好的发电量预测模型的输入数据时，上述训练好的发电量预测模型输出的预测发电量与该实际环境信息对应的实际发电量之间的损失值不大于预设的发电预测损失阈值。
其中，上述发电时间段的时间长度与上述当前目标时间段的时间长度相同，即本实施例中发电时间段的时间长度也为24小时。对于目标区域，获取的任意一个真实发电数据包括该区域在过去的任意一个24小时内的风力强度、风向、光照强度、光照时间以及对应的实际发电量，如此，可以根据这些真实发电数据建立环境信息与发电量之间的关联关系。
需要说明的是，对于获得的所有真实发电数据，按照预设比例划分获得训练发电数据集和测试发电数据集，例如，将其中百分之八十的数据作为训练发电数据集，百分之二十的数据作为测试发电数据集。
然后根据训练发电数据集对发电量预测模型进行训练，训练过程中进行模型参数的调整。在一种应用场景中，上述发电量预测模型设置有对应的风力发电和光伏发电的计算公式，且其中的部分参数通过训练进行调整。
本实施例中，在测试发电数据集上对发电量预测模型进行测试，上述预测发电量与实际发电量之间的损失值通过预设的损失公式计算，当计算出的损失值不大于预设的发电预测损失阈值时认为训练完成。在另一种应用场景中，当训练的迭代次数大于预设的发电训练迭代阈值时也认为训练完成。
步骤S200，获取上述当前目标时间段对应的时段标签和电动汽车集群中各电动汽车的耗电历史数据，将上述时段标签和上述耗电历史数据输入预先训练好的Transformer模型，通过上述预先训练好的Transformer模型获取各上述电动汽车在上述当前目标时间段对应的耗电预测数据，并获得上述电动汽车集群在上述当前目标时间段内的预测总体耗电量。
其中，上述电动汽车的耗电历史数据包括该电动汽车在前一目标时间段内每一个历史时间分段的实际耗电量，上述电动汽车的耗电预测数据包括该电动汽车在上述当前目标时间段内每一个预测时间分段的预测耗电量。
上述时段标签是根据当前目标时间段的特点确定的一个标签数据，由于在不同类型的一天（例如工作日或周末）用户的用车习惯是不同的，对应的用电习惯也不同，因此可以结合时段标签对耗电量进行预测。同时，考虑到用户的用电习惯是有连续性的，因此可以基于前一段时间（即前一天）的用电量来预测后一段时间（即今天）的用电量。结合时段标签和耗电历史数据则可以更准确地进行耗电量的预测。
需要说明的是，本实施例中，对于一个当前目标时间段（例如24小时），并不是直接预测出电动汽车集群在整个当前目标时间段的总体耗电量，而是分各个小的时间段（即预测时间分段，每一个预测时间分段的时间长度为1小时）来分别预测每一个电动汽车的耗电量（即预测耗电量），然后再计算整个电动汽车集群在整个当前目标时间段内的预测总体耗电量。如此，针对单体并区分各个小的时间段进行预测，可以使得预测结果更加精准。
对应的，输入Transformer模型的耗电历史数据也是电动汽车在前一目标时间段内每一个小的历史时间分段的实际耗电量。其中，上述前一目标时间段是在当前目标时间段之前的一个时间段，两者的时间长度相同，且前一目标时间段的结束时刻不早于当前目标时间段的起始时刻。同时，上述历史时间分段的时间长度也与预测时间分段的时间长度相同（即1小时）。
本实施例中，上述时段标签是预先设置的标签数据中的任意一种，上述标签数据包括工作日、周末和特殊节假日。进一步的，上述特殊节假日还可以包括具体的节假日名称，例如中秋节、端午节、国庆节等。因为不同的节假日用户的出行习惯和用电习惯是不同的，例如中秋节出行较少，电动汽车耗电较少，而国庆节出行的可能性较大，电动汽车耗电较多。结合时段标签可以考虑用户在不同时间的习惯，提高耗电量预测的准确性。
本实施例中，上述Transformer模型根据如下步骤进行预先训练：
获取预先采集的多个真实耗电数据，其中，一个上述真实耗电数据包括一个耗电时间段对应的时段标签、真实耗电量标签和耗电历史采集数据，上述真实耗电量标签包括各上述电动汽车在上述耗电时间段的每一个耗电时间分段内的实际耗电量，上述耗电历史采集数据包括各上述电动汽车在耗电历史采集时间段的每一个耗电历史采集时间分段内的实际耗电量，上述耗电历史采集时间段是上述耗电时间段对应的前一个时间段；
对上述真实耗电数据划分获得训练耗电数据集和测试耗电数据集；
根据上述训练耗电数据集对上述Transformer模型进行训练，并通过上述测试耗电数据集对训练的Transformer模型进行测试，直到获得训练好的Transformer模型，其中，将上述测试耗电数据集中的任意一个真实耗电数据中的时段标签和耗电历史采集数据作为上述训练好的Transformer模型的输入数据时，上述训练好的Transformer模型输出的耗电预测数据与该真实耗电数据中的真实耗电量标签的之间的损失值不大于预设的耗电预测损失阈值。
其中，上述耗电时间段（或耗电历史采集时间段）的时间长度与当前目标时间段的时间长度相同，即也为24小时，且每一个耗电时间分段（或耗电历史采集时间分段）的时间长度也与一个预测时间分段的时间长度相同，即为1小时。具体的，本实施例中，各个时间段对应的时间分段的划分方式是相同的。且上述耗电历史采集时间段是耗电时间段之前的24小时。
需要说明的是，对于获得的所有真实耗电数据，按照预设比例划分获得训练耗电数据集和测试耗电数据集，例如，将其中百分之八十的数据作为训练耗电数据集，百分之二十的数据作为测试耗电数据集。
然后根据上述训练耗电数据集对预设的Transformer模型进行训练，训练过程中进行模型参数的调整。并通过测试耗电数据集对Transformer模型进行测试，测试时，损失值根据预先设置的耗电损失公式进行计算。在一种应用场景中，还可以设置耗电训练迭代阈值以判断训练是否完成。
步骤S300，构建综合控制目标和控制约束条件，根据上述综合控制目标和上述控制约束条件，通过预设的粒子群算法进行优化求解并获得各上述电动汽车的充放电控制策略。
其中，上述综合控制目标包括目标充电方差最小、目标放电方差最小和消纳电量最大，上述目标充电方差是上述电动汽车集群在上述当前目标时间段内的充电量的方差，上述目标放电方差是上述电动汽车集群在上述当前目标时间段内的放电量的方差，上述消纳电量是上述电动汽车集群在上述当前目标时间段内的规划总体充电量，上述控制约束条件包括集群充电量范围约束和集群放电量范围约束，上述集群充电量范围约束用于限制上述消纳电量不小于上述目标发电量，上述集群放电量范围约束用于限制上述电动汽车集群在上述当前目标时间段内的规划总体放电量不小于上述预测总体耗电量。
可选的，还可以考虑充电过程中电动汽车与规划的充电区域之间的距离。本实施例中，上述综合控制目标还包括上述电动汽车集群的充电总距离最小，上述方法还包括：获取充电区域位置以及上述电动汽车集群中各上述电动汽车在上述当前目标时间段内的目标行驶路线，根据上述充电区域位置与上述目标行驶路线计算各上述电动汽车在各上述目标时间分段的充电距离数据。
在一种应用场景中，上述电动汽车是电动公交车，电动汽车集群是对应的公交车集群。因此可以根据公交车的行驶路线确定各电动汽车对应的目标行驶路线。在另一种应用场景中，电动汽车的目标行驶路线可以由用户直接输入，或者根据各个电动汽车的历史行驶数据和行驶习惯预测获得目标行驶路线，在此不作具体限定。
一个电动汽车对应的充电距离数据包括该电动汽车在各个目标时间分段的起始时刻，距离各个充电区域的距离。
在一种应用场景中，上述控制约束条件还可以包括各个充电区域内的最大接纳车辆数量，在一个时间分段内，各个充电区域内规划的电动汽车的数量不大于该充电区域预设的最大接纳车辆数量，避免出现某充电区域出现拥塞。进一步的，上述综合控制目标还可以包括电动汽车集群在上述当前目标时间段内的放电功率的方差最小和充电功率的方差最小，在此不作具体限定。
本实施例中，上述根据上述综合控制目标和上述控制约束条件，通过预设的粒子群算法进行优化求解并获得各上述电动汽车的充放电控制策略，包括：
根据上述综合控制目标构建目标函数，其中，上述目标函数是上述消纳电量的相反数、上述目标充电方差、上述目标放电方差以及上述充电总距离之和；
以上述目标函数取得最小的函数值为目标，根据上述控制约束条件，通过上述粒子群算法进行优化求解并获得各上述电动汽车的充放电控制策略。
其中，上述预设的粒子群算法可以是预设的多目标粒子群优化算法，例如社会学习粒子群算法。上述充放电控制策略包括，各预测时间分段内各个电动汽车的充电功率（或充电量）和放电功率（和放电量），其中，仅充电时放电功率为0，仅放电时充电功率为0。
在一种应用场景中，还可以设置各个电动汽车只选择一个预测时间分段进行充电，作为约束条件，以减少用户的充电次数，避免频繁充电。此时，对应的充放电策略包括电动汽车在一个目标充电时间分段，该目标充电时间分段内的充电功率，以及其它时间分段的放电功率。
需要说明的是，本实施例中的目标之一是对新能源电进行更多的消纳，本实施例中设置划总体放电量不小于上述预测总体耗电量，可以更多的消纳新能源电，且可以尽可能保证电动汽车不会缺电，同时，用户也可以根据实际需求对放电量进行调整，对于未使用完的电量，可以存储在电动汽车上。
步骤S400，根据上述充放电控制策略对上述电动汽车集群中的各上述电动汽车进行控制。
具体的，根据上述充放电控制策略对上述电动汽车集群中的各个电动汽车进行控制，对于一个电动汽车，根据充放电控制策略控制其在每一个预测时间分段内的充电功率和放电功率。上述充放电控制策略还可以包括具体的充电区域，从而控制电动汽车在对应的预测时间分段到达指定的充电区域进行充电。
需要说明的是，本实施例中求解的充放电控制策略是局部最优解，因此最终获得的控制策略并不一定能满足所有目标，而只需要满足目标函数的函数值最小即可。因此，上述可再生能源所产生的电能不一定能满足需求，或者不一定能完全被消纳。当可再生能源所产生的电能未被完全消纳时，可以使用超级电容进行储能，储存的能量可以在用电高峰期进行放电，以缓解电力系统的压力，降低发电成本。当可再生能源所产生的电能不能满足需求时，可以释放超级电容中的电能，或者从常规电网中获取电能来满足需求，而本实施例中限定了目标充电方差最小，所以从常规电网中获取电能的过程也较为稳定，不会带来剧烈的波动，有利于避免传统电网的波动。
由上可见，本实施例中综合考虑在当前目标时间段内可再生能源的目标发电量和电动汽车集群的预测总体耗电量，构建对于电动汽车集群的综合控制目标和控制约束条件，以目标充电方差最小、目标放电方差最小和消纳电量最大为目标进行优化求解并获得优化的充放电控制策略，有利于对电动汽车进行更好的充放电控制。其中，目标充电方差最小用于保证电动汽车集群带来的电网波动小，有利于保护电网，目标放电方差最小用于保证电动汽车集群中各个电动汽车的放电波动小，有利于保护电动汽车，消纳电量最大则有利于提高可再生能源的利用率。
示例性设备
如图3中所示，对应于上述电动汽车集群充放电控制方法，本发明实施例还提供一种电动汽车集群充放电控制系统，上述电动汽车集群充放电控制系统包括：
发电量获取模块510，用于获取当前目标时间段内可再生能源的目标发电量。
耗电量预测模块520，用于获取上述当前目标时间段对应的时段标签和电动汽车集群中各电动汽车的耗电历史数据，将上述时段标签和上述耗电历史数据输入预先训练好的Transformer模型，通过上述预先训练好的Transformer模型获取各上述电动汽车在上述当前目标时间段对应的耗电预测数据，并获得上述电动汽车集群在上述当前目标时间段内的预测总体耗电量，其中，上述电动汽车的耗电历史数据包括该电动汽车在前一目标时间段内每一个历史时间分段的实际耗电量，上述电动汽车的耗电预测数据包括该电动汽车在上述当前目标时间段内每一个预测时间分段的预测耗电量。
策略获取模块530，用于构建综合控制目标和控制约束条件，根据上述综合控制目标和上述控制约束条件，通过预设的粒子群算法进行优化求解并获得各上述电动汽车的充放电控制策略，其中，上述综合控制目标包括目标充电方差最小、目标放电方差最小和消纳电量最大，上述目标充电方差是上述电动汽车集群在上述当前目标时间段内的充电量的方差，上述目标放电方差是上述电动汽车集群在上述当前目标时间段内的放电量的方差，上述消纳电量是上述电动汽车集群在上述当前目标时间段内的规划总体充电量，上述控制约束条件包括集群充电量范围约束和集群放电量范围约束，上述集群充电量范围约束用于限制上述消纳电量不小于上述目标发电量，上述集群放电量范围约束用于限制上述电动汽车集群在上述当前目标时间段内的规划总体放电量不小于上述预测总体耗电量。
控制模块540，用于根据上述充放电控制策略对上述电动汽车集群中的各上述电动汽车进行控制。
具体的，本实施例中，上述电动汽车集群充放电控制系统及其各模块的具体功能可以参照上述电动汽车集群充放电控制方法中的对应描述，在此不再赘述。
需要说明的是，上述电动汽车集群充放电控制系统的各个模块的划分方式并不唯一，在此也不作为具体限定。
基于上述实施例，本发明还提供了一种智能终端，其原理框图可以如图4所示。上述智能终端包括处理器及存储器。该智能终端的存储器包括电动汽车集群充放电控制程序，存储器为电动汽车集群充放电控制程序的运行提供环境。该电动汽车集群充放电控制程序被处理器执行时实现上述任意一种电动汽车集群充放电控制方法的步骤。需要说明的是，上述智能终端还可以包括其它功能模块或单元，在此不作具体限定。
本领域技术人员可以理解，图4中示出的原理框图，仅仅是与本发明方案相关的部分结构的框图，并不构成对本发明方案所应用于其上的智能终端的限定，具体地智能终端可以包括比图中所示更多或更少的部件，或者组合某些部件，或者具有不同的部件布置。
本发明实施例还提供一种计算机可读存储介质，上述计算机可读存储介质上存储有数控机床主轴误差预测与补偿程序，上述数控机床主轴误差预测与补偿程序被处理器执行时实现本发明实施例提供的任意一种电动汽车集群充放电控制方法的步骤。
应理解，上述实施例中各步骤的序号大小并不意味着执行顺序的先后，各过程的执行顺序应以其功能和内在逻辑确定，而不应对本发明实施例的实施过程构成任何限定。
所属领域的技术人员可以清楚地了解到，为了描述的方便和简洁，仅以上述各功能单元、模块的划分进行举例说明，实际应用中，可以根据需要而将上述功能分配由不同的功能单元、模块完成，即将上述系统的内部结构划分成不同的功能单元或模块，以完成以上描述的全部或者部分功能。实施例中的各功能单元、模块可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以两个或两个以上单元集成在一个单元中，上述集成的单元既可以采用硬件的形式实现，也可以采用软件功能单元的形式实现。另外，各功能单元、模块的具体名称也只是为了便于相互区分，并不用于限制本发明的保护范围。上述系统中单元、模块的具体工作过程，可以参考前述方法实施例中的对应过程，在此不再赘述。
在上述实施例中，对各个实施例的描述都各有侧重，某个实施例中没有详述或记载的部分，可以参见其它实施例的相关描述。
本领域普通技术人员可以意识到，结合本文中所公开的实施例描述的各实例的单元及算法步骤，能够以电子硬件、或者计算机软件和电子硬件的结合来实现。这些功能究竟是以硬件还是软件方式来执行，取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同法来实现所描述的功能，但是这种实现不应认为超出本发明的范围。
在本发明所提供的实施例中，应该理解到，所揭露的系统/智能终端和方法，可以通过其它的方式实现。例如，以上所描述的系统/智能终端实施例仅仅是示意性的，例如，上述模块或单元的划分，仅仅为一种逻辑功能划分，实际实现时可以由另外的划分方式，例如多个单元或组件可以结合或者可以集成到另一个系统，或一些特征可以忽略，或不执行。
上述集成的模块/单元如果以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个计算机可读存储介质中。基于这样的理解，本发明实现上述实施例方法中的全部或部分流程，也可以通过计算机程序来指令相关的硬件来完成，上述计算机程序可存储于一计算机可读存储介质中，该计算机程序在被处理器执行时，可实现上述各个方法实施例的步骤。其中，上述计算机程序包括计算机程序代码，上述计算机程序代码可以为源代码形式、对象代码形式、可执行文件或某些中间形式等。上述计算机可读介质可以包括：能够携带上述计算机程序代码的任何实体或装置、记录介质、U盘、移动硬盘、磁碟、光盘、计算机存储器、只读存储器（ROM，Read-Only Memory）、随机存取存储器（RAM，Random Access Memory）、电载波信号、电信信号以及软件分发介质等。需要说明的是，上述计算机可读存储介质包含的内容可以根据司法管辖区内立法和专利实践的要求进行适当的增减。
以上所述实施例仅用以说明本发明的技术方案，而非对其限制；尽管参照前述实施例对本发明进行了详细的说明，本领域的普通技术人员应当理解；其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分技术特征进行等同替换；而这些修改或者替换，并不是相应技术方案的本质脱离本发明各实施例技术方案的精神和范围，均应包含在本发明的保护范围之内。

Claims (10)

  1. 一种电动汽车集群充放电控制方法，其特征在于，所述方法包括：
    获取当前目标时间段内可再生能源的目标发电量；
    获取所述当前目标时间段对应的时段标签和电动汽车集群中各电动汽车的耗电历史数据，将所述时段标签和所述耗电历史数据输入预先训练好的Transformer模型，通过所述预先训练好的Transformer模型获取各所述电动汽车在所述当前目标时间段对应的耗电预测数据，并获得所述电动汽车集群在所述当前目标时间段内的预测总体耗电量，其中，所述电动汽车的耗电历史数据包括该电动汽车在前一目标时间段内每一个历史时间分段的实际耗电量，所述电动汽车的耗电预测数据包括该电动汽车在所述当前目标时间段内每一个预测时间分段的预测耗电量；
    构建综合控制目标和控制约束条件，根据所述综合控制目标和所述控制约束条件，通过预设的粒子群算法进行优化求解并获得各所述电动汽车的充放电控制策略，其中，所述综合控制目标包括目标充电方差最小、目标放电方差最小和消纳电量最大，所述目标充电方差是所述电动汽车集群在所述当前目标时间段内的充电量的方差，所述目标放电方差是所述电动汽车集群在所述当前目标时间段内的放电量的方差，所述消纳电量是所述电动汽车集群在所述当前目标时间段内的规划总体充电量，所述控制约束条件包括集群充电量范围约束和集群放电量范围约束，所述集群充电量范围约束用于限制所述消纳电量不小于所述目标发电量，所述集群放电量范围约束用于限制所述电动汽车集群在所述当前目标时间段内的规划总体放电量不小于所述预测总体耗电量；
    根据所述充放电控制策略对所述电动汽车集群中的各所述电动汽车进行控制。
  2. 根据权利要求1所述的电动汽车集群充放电控制方法，其特征在于，所述可再生能源包括预设的目标区域内的风能和太阳能，所述获取当前目标时间段内可再生能源的目标发电量，包括：
    获取所述当前目标时间段内的环境预测信息，其中，所述环境预测信息包括风力强度、风向、光照强度和光照时间；
    根据所述环境预测信息获取所述目标发电量。
  3. 根据权利要求2所述的电动汽车集群充放电控制方法，其特征在于，所述根据所述环境预测信息获取所述目标发电量，包括：
    将所述环境预测信息输入预先训练好的发电量预测模型，将所述训练好的发电量预测模型输出的预测发电量作为所述当前目标时间段内的目标发电量。
  4. 根据权利要求3所述的电动汽车集群充放电控制方法，其特征在于，所述发电量预测模型根据如下步骤进行预先训练：
    获取预先采集的多个真实发电数据，其中，一个所述真实发电数据包括一个发电时间段内的实际环境信息和该发电时间段内的实际发电量，所述实际环境信息包括风力强度、风向、光照强度和光照时间；
    对所述真实发电数据划分获得训练发电数据集和测试发电数据集；
    根据所述训练发电数据集对所述发电量预测模型进行训练，并通过所述测试发电数据集对训练的发电量预测模型进行测试，直到获得训练好的发电量预测模型，其中，将所述测试发电数据集中的任意一个实际环境信息作为所述训练好的发电量预测模型的输入数据时，所述训练好的发电量预测模型输出的预测发电量与该实际环境信息对应的实际发电量之间的损失值不大于预设的发电预测损失阈值。
  5. 根据权利要求1所述的电动汽车集群充放电控制方法，其特征在于，所述时段标签是预先设置的标签数据中的任意一种，所述标签数据包括工作日、周末和特殊节假日，所述Transformer模型根据如下步骤进行预先训练：
    获取预先采集的多个真实耗电数据，其中，一个所述真实耗电数据包括一个耗电时间段对应的时段标签、真实耗电量标签和耗电历史采集数据，所述真实耗电量标签包括各所述电动汽车在所述耗电时间段的每一个耗电时间分段内的实际耗电量，所述耗电历史采集数据包括各所述电动汽车在耗电历史采集时间段的每一个耗电历史采集时间分段内的实际耗电量，所述耗电历史采集时间段是所述耗电时间段对应的前一个时间段；
    对所述真实耗电数据划分获得训练耗电数据集和测试耗电数据集；
    根据所述训练耗电数据集对所述Transformer模型进行训练，并通过所述测试耗电数据集对训练的Transformer模型进行测试，直到获得训练好的Transformer模型，其中，将所述测试耗电数据集中的任意一个真实耗电数据中的时段标签和耗电历史采集数据作为所述训练好的Transformer模型的输入数据时，所述训练好的Transformer模型输出的耗电预测数据与该真实耗电数据中的真实耗电量标签的之间的损失值不大于预设的耗电预测损失阈值。
  6. 根据权利要求1-5任意一项所述的电动汽车集群充放电控制方法，其特征在于，所述综合控制目标还包括所述电动汽车集群的充电总距离最小，所述方法还包括：
    获取充电区域位置以及所述电动汽车集群中各所述电动汽车在所述当前目标时间段内的目标行驶路线，根据所述充电区域位置与所述目标行驶路线计算各所述电动汽车在各所述目标时间分段的充电距离数据。
  7. 根据权利要求6所述的电动汽车集群充放电控制方法，其特征在于，所述根据所述综合控制目标和所述控制约束条件，通过预设的粒子群算法进行优化求解并获得各所述电动汽车的充放电控制策略，包括：
    根据所述综合控制目标构建目标函数，其中，所述目标函数是所述消纳电量的相反数、所述目标充电方差、所述目标放电方差以及所述充电总距离之和；
    以所述目标函数取得最小的函数值为目标，根据所述控制约束条件，通过所述粒子群算法进行优化求解并获得各所述电动汽车的充放电控制策略。
  8. 一种电动汽车集群充放电控制系统，其特征在于，所述系统包括：
    发电量获取模块，用于获取当前目标时间段内可再生能源的目标发电量；
    耗电量预测模块，用于获取所述当前目标时间段对应的时段标签和电动汽车集群中各电动汽车的耗电历史数据，将所述时段标签和所述耗电历史数据输入预先训练好的Transformer模型，通过所述预先训练好的Transformer模型获取各所述电动汽车在所述当前目标时间段对应的耗电预测数据，并获得所述电动汽车集群在所述当前目标时间段内的预测总体耗电量，其中，所述电动汽车的耗电历史数据包括该电动汽车在前一目标时间段内每一个历史时间分段的实际耗电量，所述电动汽车的耗电预测数据包括该电动汽车在所述当前目标时间段内每一个预测时间分段的预测耗电量；
    策略获取模块，用于构建综合控制目标和控制约束条件，根据所述综合控制目标和所述控制约束条件，通过预设的粒子群算法进行优化求解并获得各所述电动汽车的充放电控制策略，其中，所述综合控制目标包括目标充电方差最小、目标放电方差最小和消纳电量最大，所述目标充电方差是所述电动汽车集群在所述当前目标时间段内的充电量的方差，所述目标放电方差是所述电动汽车集群在所述当前目标时间段内的放电量的方差，所述消纳电量是所述电动汽车集群在所述当前目标时间段内的规划总体充电量，所述控制约束条件包括集群充电量范围约束和集群放电量范围约束，所述集群充电量范围约束用于限制所述消纳电量不小于所述目标发电量，所述集群放电量范围约束用于限制所述电动汽车集群在所述当前目标时间段内的规划总体放电量不小于所述预测总体耗电量；
    控制模块，用于根据所述充放电控制策略对所述电动汽车集群中的各所述电动汽车进行控制。
  9. 一种智能终端，其特征在于，所述智能终端包括存储器、处理器以及存储在所述存储器上并可在所述处理器上运行的电动汽车集群充放电控制程序，所述电动汽车集群充放电控制程序被所述处理器执行时实现如权利要求1-7任意一项所述电动汽车集群充放电控制方法的步骤。
  10. 一种计算机可读存储介质，其特征在于，所述计算机可读存储介质上存储有电动汽车集群充放电控制程序，所述电动汽车集群充放电控制程序被处理器执行时实现如权利要求1-7任意一项所述电动汽车集群充放电控制方法的步骤。
PCT/CN2022/137650 2022-05-27 2022-12-08 电动汽车集群充放电控制方法、系统及相关设备 WO2023226368A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202210587210.X 2022-05-27
CN202210587210.XA CN114919433B (zh) 2022-05-27 2022-05-27 电动汽车集群充放电控制方法、系统及相关设备

Publications (1)

Publication Number Publication Date
WO2023226368A1 true WO2023226368A1 (zh) 2023-11-30

Family

ID=82811218

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/137650 WO2023226368A1 (zh) 2022-05-27 2022-12-08 电动汽车集群充放电控制方法、系统及相关设备

Country Status (2)

Country Link
CN (1) CN114919433B (zh)
WO (1) WO2023226368A1 (zh)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114919433B (zh) * 2022-05-27 2022-12-09 深圳先进技术研究院 电动汽车集群充放电控制方法、系统及相关设备
CN115465141A (zh) * 2022-09-16 2022-12-13 上海挚达科技发展有限公司 电动汽车充放电控制方法、装置、电子设备及存储介质

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2746093A1 (en) * 2012-12-21 2014-06-25 Fundació Privada Barcelona Digital Centre Tecnologic Method and apparatus for optimized management of an electric vehicle charging infrastructure
CN106532764A (zh) * 2016-10-18 2017-03-22 国网山东省电力公司电力科学研究院 一种就地消纳光伏发电的电动汽车充电负荷调控方法
JP2020061800A (ja) * 2018-10-04 2020-04-16 パナソニックＩｐマネジメント株式会社 充電方法、及び、制御装置
CN111027842A (zh) * 2019-12-04 2020-04-17 清华大学 电动汽车充电与新能源发电的协同调度方法及装置
CN112257897A (zh) * 2020-09-17 2021-01-22 华北电力大学 基于改进多目标粒子群的电动汽车充电优化方法和系统
CN114330923A (zh) * 2022-01-06 2022-04-12 国电南瑞南京控制系统有限公司 一种基于公共气象数据的光伏发电功率预测方法
CN114407661A (zh) * 2022-02-10 2022-04-29 北京理工大学重庆创新中心 基于数据驱动的电动车能耗预测方法、系统、设备及介质
CN114919433A (zh) * 2022-05-27 2022-08-19 深圳先进技术研究院 电动汽车集群充放电控制方法、系统及相关设备

Family Cites Families (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104680258A (zh) * 2015-03-12 2015-06-03 北京交通大学 电动出租车调度方法及装置
CN105139096A (zh) * 2015-09-28 2015-12-09 东南大学 一种基于两阶段优化的电动汽车充电站选址定容方法
CN107640049B (zh) * 2017-09-20 2020-04-21 东北大学 一种针对充电站的电动汽车有序充电的移动终端系统
CN110015138B (zh) * 2017-11-20 2020-10-02 深圳先进技术研究院 电动汽车电池监测方法、装置、系统及电动汽车
JP2019158646A (ja) * 2018-03-14 2019-09-19 本田技研工業株式会社 車両制御装置、車両制御方法、及びプログラム
CN108510006B (zh) * 2018-04-08 2020-06-09 重庆邮电大学 一种基于数据挖掘的企业用电量分析与预测方法
CN109462231B (zh) * 2018-11-15 2020-09-01 合肥工业大学 居民微电网的负荷优化调度方法、系统和存储介质
EP3702202B1 (en) * 2019-02-28 2023-01-18 Bayerische Motoren Werke Aktiengesellschaft Method and system to optimize distributed charging station efficiency and user experience
CN110322120B (zh) * 2019-06-12 2021-08-06 深圳大学 一种基于效益最大化的电动车充电调度方法和系统
CN110599042A (zh) * 2019-09-12 2019-12-20 国网电子商务有限公司 一种基于云计算的智能用电分析方法及系统
US11865939B2 (en) * 2020-08-28 2024-01-09 Toyota Motor North America, Inc. Power allocation to transports
CN112200373A (zh) * 2020-10-15 2021-01-08 中国科学院深圳先进技术研究院 负荷预测模型的训练方法及训练装置、存储介质、设备
CN113043863A (zh) * 2021-03-12 2021-06-29 联合汽车电子有限公司 车辆充电方法及系统、可读存储介质及电子设备
CN113191682B (zh) * 2021-05-24 2022-07-29 温州大学 一种适用于电动汽车移动充电设备充电的多能源分配管理方法
CN113675839B (zh) * 2021-07-07 2022-12-20 深圳奥特迅电力设备股份有限公司 一种电动汽车充电设备与电网协调互动控制系统及方法
CN113962446B (zh) * 2021-10-08 2024-06-07 国网安徽省电力有限公司电力科学研究院 一种微电网群协同调度方法、装置、电子设备和存储介质

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2746093A1 (en) * 2012-12-21 2014-06-25 Fundació Privada Barcelona Digital Centre Tecnologic Method and apparatus for optimized management of an electric vehicle charging infrastructure
CN106532764A (zh) * 2016-10-18 2017-03-22 国网山东省电力公司电力科学研究院 一种就地消纳光伏发电的电动汽车充电负荷调控方法
JP2020061800A (ja) * 2018-10-04 2020-04-16 パナソニックＩｐマネジメント株式会社 充電方法、及び、制御装置
CN111027842A (zh) * 2019-12-04 2020-04-17 清华大学 电动汽车充电与新能源发电的协同调度方法及装置
CN112257897A (zh) * 2020-09-17 2021-01-22 华北电力大学 基于改进多目标粒子群的电动汽车充电优化方法和系统
CN114330923A (zh) * 2022-01-06 2022-04-12 国电南瑞南京控制系统有限公司 一种基于公共气象数据的光伏发电功率预测方法
CN114407661A (zh) * 2022-02-10 2022-04-29 北京理工大学重庆创新中心 基于数据驱动的电动车能耗预测方法、系统、设备及介质
CN114919433A (zh) * 2022-05-27 2022-08-19 深圳先进技术研究院 电动汽车集群充放电控制方法、系统及相关设备

Also Published As

Publication number Publication date
CN114919433A (zh) 2022-08-19
CN114919433B (zh) 2022-12-09

Similar Documents

Publication Publication Date Title
Sun A multi-objective optimization model for fast electric vehicle charging stations with wind, PV power and energy storage
Liu et al. Optimal sizing of a wind-energy storage system considering battery life
CN112103946B (zh) 一种基于粒子群算法的微电网储能优化配置方法
Zhang et al. An improved charging/discharging strategy of lithium batteries considering depreciation cost in day-ahead microgrid scheduling
CN109217290B (zh) 计及电动汽车充放电的微网能量优化管理方法
WO2023226368A1 (zh) 电动汽车集群充放电控制方法、系统及相关设备
CN110661246B (zh) 一种城市轨道交通光伏储能系统的容量优化配置方法
CN110378548B (zh) 一种电动汽车虚拟电厂多时间尺度响应能力评估模型构建方法
CN110084443B (zh) 一种基于qpso优化算法的换电站运行优化模型分析方法
CN104966127A (zh) 一种基于需求响应的电动汽车经济调度方法
CN106532764A (zh) 一种就地消纳光伏发电的电动汽车充电负荷调控方法
CN111064214A (zh) 基于电动汽车两阶段滚动式策略的配电网优化调度方法
CN109742755B (zh) 基于节点分区的含电动汽车的主动配电网运行调度方法
CN115147245B (zh) 一种工业负荷参与调峰辅助服务的虚拟电厂优化调度方法
CN112101632B (zh) 基于分布式估计算法楼宇负荷优化调度方法
CN111525624A (zh) 基于蓄电池储能系统的家庭分布式能源调度方法
CN117543581A (zh) 考虑电动汽车需求响应的虚拟电厂优化调度方法及其应用
CN115940284B (zh) 一种考虑分时电价的新能源制氢系统的运行控制策略
CN114069669B (zh) 一种共享储能运行模式控制方法
CN113705914B (zh) 利用区块链的电动车充电站管理方法
CN115759361A (zh) 基于双层规划的交通能源调度方法、系统、设备和介质
CN111216586B (zh) 考虑风电消纳的居民小区电动汽车有序充电控制方法
CN112803493B (zh) 基于边际成本的分散式储能与分布式光伏协同方法及系统
CN114723278A (zh) 一种考虑光伏储能的社区微电网调度方法及系统
CN115882460A (zh) 一种考虑需求侧管理的两阶段新能源微电网优化调度方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22943555

Country of ref document: EP

Kind code of ref document: A1