A method for promoting electricity consumption by new energy vehicles

Abstract

The present invention relates to the technical field of power system optimization and dispatching, and discloses a method for promoting electricity consumption by new energy vehicles. The method first obtains real-time load data of the power grid to establish a prediction model, combines the user behavior data of new energy vehicles to generate a spatiotemporal matrix of charging demand, and simultaneously calculates the renewable energy consumption threshold to construct a multi-objective optimization function. On this basis, the electricity price incentive levels are divided, the dynamic electricity price adjustment coefficient is calculated, a user response probability model is established, and differentiated charging guidance instructions are generated. The method has significant beneficial effects, can improve the utilization rate of renewable energy, reduce the power abandonment rate; balance the power grid load, reduce the peak-valley difference; reduce the user's charging cost; improve the level of intelligent energy management, promote the coordinated development of new energy vehicles and power grids, and also has good economic and environmental benefits, effectively solving the difficult problems of new energy vehicle charging and renewable energy consumption.

Description

Technical Field

[0001] The present invention relates to the technical field of power system optimization and dispatching, and specifically to a method for promoting electricity consumption by new energy vehicles. Background Art

[0002] Against the backdrop of global efforts to address climate change and promote energy transition, the new energy vehicle industry is booming, becoming a key force in achieving energy conservation and emissions reduction in the transportation sector. At the same time, as the proportion of distributed renewable energy (such as solar and wind power) in the power supply system continues to rise, the problem of electricity consumption is becoming increasingly prominent, becoming a major constraint on sustainable energy development. In this context, exploring collaborative development models between new energy vehicles, the power grid, and renewable energy sources is of great significance for achieving efficient energy utilization and stable power system operation.

[0003] In recent years, the number of new energy vehicles has exploded. While this growth trend helps reduce carbon emissions from traditional fuel vehicles, it also presents new challenges for the power system. New energy vehicle charging is characterized by randomness and concentration. If a large number of vehicles charge at the same time, it can easily cause a sudden increase in localized grid load, increase peak-to-valley variations, and impact the stability and reliability of power supply.

[0004] Renewable energy, as a major clean energy source, is experiencing rapid growth. However, its intermittent and volatile generation is a prominent problem. For example, solar photovoltaic power generation is subject to fluctuations in weather and daytime, making its output unstable. During cloudy days or at night, photovoltaic power generation can drop significantly or even cease. Wind power generation is similarly constrained by volatile wind speeds and directions, making it difficult to maintain a stable supply. This makes integrating renewable energy into the grid extremely challenging for power balance and power quality control, resulting in a significant inability to effectively absorb large amounts of renewable energy and frequent wind and solar power curtailment.

[0005] To address the above-mentioned issues, the industry has currently adopted a variety of measures. In terms of new energy vehicle charging management, some charging station operators adopt a timed charging strategy to encourage users to charge at night when electricity prices are low. However, this approach does not fully consider the actual needs of users and the real-time load changes of the power grid. Users' flexibility is limited, and charging plans cannot be dynamically adjusted according to the power generation of renewable energy. In terms of renewable energy consumption, some regions have built large-scale energy storage facilities, such as battery energy storage power stations, to store electricity when renewable energy generation is in excess and release electricity when power generation is insufficient. However, the construction cost of energy storage facilities is high, the investment payback period is long, and large-scale promotion faces economic pressure. At the same time, the existing energy storage capacity is limited, and it is difficult to fully balance the large power fluctuations of renewable energy.

[0006] The development of smart grid and big data analytics technologies has provided new opportunities for addressing the challenges of new energy vehicle electricity consumption and renewable energy consumption. Smart grids enable intelligent monitoring, control, and management of power systems, providing real-time access to grid load data, power generation equipment output data, and other information. Big data analytics can mine and analyze massive amounts of user behavior and energy data, providing a basis for precise energy management strategies. However, a mature and effective electricity consumption solution that leverages these technologies and comprehensively considers grid load, new energy vehicle charging needs, and the characteristics of renewable energy generation has yet to be established. Summary of the Invention

[0007] The purpose of the present invention is to provide a method for promoting electricity consumption of new energy vehicles to solve the problems raised in the above background technology.

[0008] To achieve the above objectives, the present invention provides the following technical solution: a method for promoting electricity consumption by new energy vehicles, the method comprising:

[0009] Step 1: Obtain real-time load data of the target area power grid, extract historical load fluctuation characteristics, and establish a dynamic load forecasting model for the power grid;

[0010] Step 2: Based on the new energy vehicle user behavior data, the charging demand distribution in different time periods and geographical locations is counted to generate a spatiotemporal matrix of charging demand;

[0011] Step 3: Identify the real-time output data of distributed renewable energy generation equipment in the target area and calculate the renewable energy absorption threshold based on the remaining capacity of the energy storage equipment;

[0012] Step 4: Construct a multi-objective optimization function based on the grid dynamic load forecasting model and the charging demand spatiotemporal matrix. The objectives of the multi-objective optimization function include grid load balance, renewable energy utilization rate, and user charging cost.

[0013] Step 5: Based on the absorbable threshold and the charging period optimization allocation result, the electricity price incentive levels for different periods are divided, and the dynamic electricity price adjustment coefficient is calculated;

[0014] Step 6: Extract historical charging behavior data of users, analyze their sensitivity to dynamic electricity prices, and establish a user response probability model;

[0015] Step 7: Generate differentiated charging guidance instructions based on the user response probability model and the dynamic electricity price adjustment coefficient, and send them to the user terminal through the communication network.

[0016] Preferably, the method for establishing the power grid dynamic load prediction model in step 1 is:

[0017] The historical load data is divided into multiple periods according to the time series, and the load characteristic parameters of each period are extracted, including peak load, valley load and load change gradient; the formula for constructing the load forecasting model is:

[0018]

[0019] in, For time The predicted load value; For time The benchmark load is calculated by averaging historical data; For the Abnormal load fluctuations are extracted through outlier detection algorithm; 、 、 are the model parameters fitted by machine learning; The historical load data The moment corresponding to the abnormal fluctuation.

[0020] Preferably, the process of generating the charging demand spatiotemporal matrix in step 2 includes:

[0021] Divide the target area into multiple grid units and count the charging start time, charging duration, and charging power of new energy vehicles in each grid unit;

[0022] Define the charging demand space-time matrix as a three-dimensional tensor , where i represents the longitude partition, j represents the latitude partition, and t represents the time slice. The value of each element is the total charging demand of the corresponding grid cell in time slice t.

[0023] Preferably, the method for calculating the renewable energy absorbability threshold in step 3 is:

[0024] Obtain real-time output of distributed renewable energy and the remaining capacity of the energy storage device ;

[0025] Set the absorption threshold to:

[0026]

[0027] in, is the scheduling time interval, The maximum reverse feed power allowed by the grid.

[0028] Preferably, the specific formula for constructing the multi-objective optimization function in step 4 is:

[0029]

[0030] in, is the net load of the grid, is the average load, Assign a value to the charging power, Charging costs for users, 、 、 is the weight coefficient and satisfies .

[0031] Preferably, the calculation process of the dynamic electricity price adjustment coefficient in step 5 includes:

[0032] According to the absorbable threshold and charging power distribution value The difference between the two is used to divide the electricity price incentive level;

[0033] Set the dynamic electricity price adjustment coefficient to:

[0034]

[0035] in, is the base electricity price, is the preset elastic coefficient.

[0036] Preferably, the method for establishing the user response probability model in step 6 is:

[0037] Extracting electricity price sensitivity indicators from users' historical charging behavior and charging urgency indicator ;

[0038] Define the user response probability as:

[0039] Where a, b, and c are the coefficients fitted by logistic regression.

[0040] Preferably, the process of generating differentiated charging guidance instructions in step 7 includes:

[0041] Based on user response probability Divide user groups, send priority charging instructions to users with high response probability, and send delayed charging suggestions to users with low response probability;

[0042] The charging guidance instruction includes recommended charging time period, estimated cost and renewable energy ratio information.

[0043] Preferably, the present invention further includes an electronic device, comprising:

[0044] a processor; a memory for storing instructions executable by the processor;

[0045] The processor is configured to call the instructions stored in the memory to execute a method for promoting electricity consumption by new energy vehicles.

[0046] Preferably, the present invention also includes a computer-readable storage medium having computer program instructions stored thereon, and when the computer program instructions are executed by a processor, the above-mentioned method of promoting electricity consumption by new energy vehicles is implemented.

[0047] Compared with the prior art, the present invention has the following beneficial effects:

[0048] Accurately calculate the renewable energy absorption threshold and integrate it into the multi-objective optimization function and electricity price incentive mechanism. By identifying the real-time output of distributed renewable energy generation equipment and the remaining capacity of energy storage equipment, the potential for renewable energy absorption can be fully tapped.

[0049] A multi-objective optimization function is constructed using the grid's dynamic load forecasting model and the spatiotemporal matrix of charging demand to effectively regulate new energy vehicle charging behavior. The model takes into account historical load fluctuations, accurately predicts load changes, and optimizes charging times based on the distribution of charging demand. During peak load periods, the charging load for new energy vehicles is reduced; during off-peak periods, the charging load is increased. Based on a dynamic electricity price adjustment coefficient and a user response probability model, differentiated charging guidance is provided to users. Users with high price sensitivity and low charging urgency receive priority charging recommendations during periods with lower electricity prices, saving on charging costs. The integration of smart grid technologies and big data analytics enables comprehensive monitoring and precise regulation of the power grid, new energy vehicles, and renewable energy. Real-time data acquisition enables in-depth analysis of user behavior and energy supply and demand patterns, providing a scientific basis for energy management decisions.

[0050] By rationally guiding the charging of new energy vehicles, we can tap into the potential of two-way interaction between vehicles and the power grid (V2G). When the grid is under load, new energy vehicles can feed power back to the grid, alleviating power shortages. When renewable energy generation is in excess, excess energy can be stored. This enhances the flexibility and resilience of the power grid, expands the application value of new energy vehicles, and promotes collaborative innovation and development between the new energy vehicle industry and the power industry.

[0051] Reducing renewable energy curtailment, improving energy efficiency, and lowering power generation costs; reducing grid upgrade investment, alleviating the imbalance between power supply and demand, and generating significant economic benefits; and using renewable energy on a large scale to charge new energy vehicles, reducing carbon emissions. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 This is a working principle diagram of the method for promoting electricity consumption by new energy vehicles according to the present invention;

[0053] Figure 2 Flowchart for building a dynamic load forecasting model for the power grid;

[0054] Figure 3 Flowchart for generating the spatiotemporal matrix of charging demand. DETAILED DESCRIPTION

[0055] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0056] See also Figure 1-3 The present invention provides a method for promoting electricity consumption by new energy vehicles, and its overall implementation scheme is as follows:

[0057] Step 1: Obtain real-time grid load data and establish a forecasting model: This process involves exchanging data with the target area's grid monitoring system to obtain real-time grid load data. Load data for the past period (e.g., one year) is extracted from the historical data repository and divided into multiple time periods based on the time series. Each time period can be customized, for example, in hourly units. For each time period, load characteristic parameters such as peak load, valley load, and load gradient are extracted.

[0058] Step 2: Calculate the charging demand distribution to generate a spatiotemporal matrix: Collect user behavior data for new energy vehicles (NEVs). This data can come from sources such as the vehicle's onboard system, charging station operational data, and user registration information. Divide the target area geographically into multiple grid cells, with the size of each cell determined based on the region's actual conditions and data accuracy requirements. For each grid cell, calculate the charging start time, charging duration, and charging power of the NEVs within it. Construct a three-dimensional spatiotemporal matrix of charging demand, representing the total charging demand at different locations and times.

[0059] Step 3: Calculate the renewable energy absorption threshold: Real-time monitoring is performed on the output of distributed renewable energy generation equipment (such as solar photovoltaic panels and wind turbines) within the target area, as well as the remaining capacity of energy storage equipment (such as battery energy storage systems). The absorption threshold is calculated based on the dispatch interval and the maximum backfeed power allowed by the grid. This threshold measures the upper limit of renewable energy capacity that can be absorbed at a specific moment.

[0060] Step 4: Construct a multi-objective optimization function: Based on the previously generated grid dynamic load forecast model and the spatiotemporal matrix of charging demand, a multi-objective optimization function is constructed. This function balances the three objectives of grid load balance, renewable energy utilization, and user charging costs by setting different weight coefficients to find the optimal charging power allocation solution.

[0061] Step 5: Classify electricity price incentive levels and calculate adjustment coefficients: Based on the difference between the renewable energy absorption threshold and the charging power allocation value, electricity price incentive levels are divided according to specific rules for different time periods. This coefficient will be used to adjust electricity prices for different time periods to encourage users to charge during periods with abundant renewable energy and low grid load.

[0062] Step 6: Build a user response probability model: Extract price sensitivity and charging urgency metrics from historical user charging behavior data. Using these metrics, we use machine learning methods like logistic regression to derive correlation coefficients to predict the probability of a user responding to different electricity prices.

[0063] Step 7: Generate and send differentiated charging guidance instructions: Based on the results of the user response probability model, user groups are divided. Prioritized charging instructions are sent to users with a high response probability, while delayed charging recommendations are sent to users with a low response probability. These charging guidance instructions are sent to user terminals via communication networks (such as 4G / 5G networks and Wi-Fi). The instructions include recommended charging times, estimated costs, and the proportion of renewable energy, helping users make informed charging decisions.

[0064] The implementation of the present invention will be further described below with reference to Examples 1 to 5.

[0065] Example 1:

[0066] The purpose of this embodiment is to explain in detail how to accurately establish a dynamic load forecasting model for a power grid, so as to improve the accuracy of load forecasting and provide a reliable basis for subsequent charging strategy formulation.

[0067] After obtaining the real-time load data and historical load data of the target area power grid, the historical load data is divided into multiple time periods per hour. For each time period, the peak load, valley load and load change gradient are extracted through data analysis tools. Assume that the start time of a time period is , and the end time is Hours, during this period, by sampling and processing the power grid monitoring data, the maximum load value in this period is obtained as the peak load , the minimum load value is taken as the valley load The load change gradient is obtained by calculating the ratio of the load change to the time during the period, that is, ,in Indicates time The grid load value.

[0068] When building a load forecasting model, the formula is:

[0069] in, For time The predicted load value; For time The base load is calculated by calculating the time and The historical load data of the same time period (such as the same hour of each day) is obtained. For example, the average load value from 10:00 to 11:00 every day in the past year is calculated as the benchmark load at 10:00. .

[0070] For the Abnormal load fluctuations are extracted using outlier detection algorithms (such as density-based spatial clustering algorithm DBSCAN). Assume that in a certain period of historical load data, three abnormal load fluctuation points are found after outlier detection, and the corresponding abnormal load fluctuations are 、 、 The corresponding moments are 、 、 .

[0071] 、 、 are the model parameters fitted by machine learning. These parameters can be optimized using the gradient descent algorithm. First, the historical load data is divided into a training set and a test set. The training set is used to train the model, and the test set is used to evaluate the accuracy of the model. During the training process, the 、 、 The value of The error (such as mean square error) between the model and the actual load value is minimized. After multiple iterative calculations, the optimal model parameter values ​​are obtained.

[0072] The dynamic load forecasting model for the power grid established in this way can fully consider the normal trends and abnormal fluctuations in historical load data, improve the accuracy of load forecasting, and provide accurate data support for subsequent optimal allocation of power resources.

[0073] Example 2:

[0074] The purpose of this embodiment is to elaborate on how to generate a spatiotemporal matrix of charging demand, so as to more accurately grasp the charging demand of new energy vehicles at different times and spaces, and provide a basis for optimizing charging strategies.

[0075] Taking a certain urban area as the target area, the area is divided into The grid unit can be set to 1 km on each side according to actual conditions. By connecting with the new energy vehicle charging pile operation platform and vehicle management system, data on the charging start time, charging duration, and charging power of new energy vehicles in each grid unit can be obtained.

[0076] Assume that in a grid cell In the time slice (Time slices can be set to one every 30 minutes), multiple new energy vehicles are charging. For one of the vehicles, the charging start time is , charging time is , the charging power is .if In time slice Inside, and Also in time slice Then the total charging demand of this car in this time slice is .if In time slice Before, but Part in time slice In the time slice The charging time is , then the total charging demand is .

[0077] Define the charging demand space-time matrix as a three-dimensional tensor ,in Represents the longitude partition, Show latitude partition, Indicates a time slice. For example, at a certain moment, the grid cell with longitude partition 3 and latitude partition 5 is in the time slice There are multiple vehicles charging within this time slice. After statistical calculation, the total charging demand of this grid unit in this time slice is kilowatt-hours, then .

[0078] The charging demand spatiotemporal matrix generated in this way can intuitively reflect the distribution of charging demand for new energy vehicles in different geographical locations and times, providing a detailed data basis for subsequent multi-objective optimization and charging strategy formulation.

[0079] Example 3:

[0080] The purpose of this embodiment is to explain in detail the calculation process of the renewable energy absorbable threshold value, so as to ensure that the absorbable renewable energy power is reasonably determined under the premise of considering the safety of the power grid and the energy storage capacity.

[0081] In the target area, distributed renewable energy power generation equipment (such as solar photovoltaic power stations, wind farms, etc.) monitors its output data in real time through sensors. , energy storage equipment (such as large battery energy storage systems) also provides real-time feedback on their remaining capacity Assume that the scheduling interval Set to 1 hour, the maximum reverse feed power allowed by the grid is .

[0082] The formula for calculating the renewable energy absorption threshold is:

[0083] For example, at a certain moment , the real-time output of a distributed solar photovoltaic power station in a certain area is , the remaining capacity of the energy storage equipment in this area is , scheduling time interval hours, the maximum reverse feed power allowed by the grid .

[0084] First calculate , then with Compare and take the smaller value. At this moment, the renewable energy consumption threshold of the region is .

[0085] Through this calculation method, the real-time output of renewable energy power generation equipment, the remaining capacity of energy storage equipment and the safety limitations of the power grid can be comprehensively considered to accurately calculate the upper limit of renewable energy electricity that can be absorbed at each moment, providing an important reference basis for the subsequent optimal allocation of power resources.

[0086] Example 4:

[0087] The purpose of this embodiment is to explain in detail the construction process of the multi-objective optimization function and the significance of each parameter, so as to achieve comprehensive optimization of grid load balance, renewable energy utilization and user charging costs.

[0088] The net load of the power grid is obtained according to the dynamic load prediction model of the power grid , the average load is obtained by statistically calculating the grid load data within a period of time (such as a day) In terms of charging power allocation, the charging power allocation value is determined based on the charging demand space-time matrix and optimization algorithm. , while considering the cost of charging for users at different times .

[0089] The formula for constructing the multi-objective optimization function is:

[0090]

[0091] in, 、 、 is the weight coefficient and satisfies .

[0092] Assume that in a certain optimization scenario, the weight coefficient is set to , , In 24 hours a day, the net load of the power grid for each hour is obtained through the power grid dynamic load forecasting model. , calculate the average load for the day Determine the charging power allocation value for each hour based on the charging demand space-time matrix and optimization algorithm At the same time, the user's charging cost is calculated based on the electricity price at different times .

[0093] For the load balancing part of the power grid, calculate , for example, in hour, , , then the value of this item is ;exist hour, , then the value of this item is , and so on, calculate the total of this item for 24 hours.

[0094] For the renewable energy utilization rate, first calculate and , assuming , , then the value of this item is .

[0095] For the user charging cost part, calculate , assuming that the total charging cost of the user on this day is calculated to be Yuan, the value of this item is .

[0096] Add the values ​​of these three parts to get the value of the multi-objective optimization function By continuously adjusting the charging power distribution value , making The value of is minimized, thereby achieving comprehensive optimization of grid load balance, renewable energy utilization rate and user charging cost.

[0097] Example 5:

[0098] This embodiment aims to explain in detail how to adjust the electricity price based on the relationship between the absorptive threshold and the charging power allocation value, build a user response probability model, and generate targeted charging guidance instructions based on this model to guide users to reasonably arrange charging behavior and improve electricity efficiency.

[0099] In terms of dynamic electricity price adjustment coefficient calculation, based on the absorbable threshold and charging power distribution value The difference in power consumption is divided into three incentive levels. A difference greater than 50kW is considered a high incentive level; a difference between 20kW (inclusive) and 50kW is considered a medium incentive level; and a difference of 20kW or less is considered a low incentive level.

[0100] Taking a certain region as an example, set the benchmark electricity price 0.6 yuan / kWh, with a preset elasticity coefficient At 16:00, the area can absorb the threshold The charging power distribution value is 300kW The difference between the two is 70kW, which is a high incentive level. Calculation, dynamic electricity price adjustment coefficient Compared with the benchmark electricity price, the electricity price during this period has increased, aiming to encourage more users to charge during periods with greater room for renewable energy consumption.

[0101] When building a user response probability model, key indicators are extracted from the user's historical charging data. For example, when the electricity price dropped by 15% in the past, user A increased his charging power by 25%, which indicates that he is highly sensitive to electricity prices. If the user often charges when the power level is less than 30%, it indicates that he has a high sense of urgency in charging. By analyzing a large amount of user data through logistic regression, the coefficient 、 、 .

[0102] For user A, assuming that its electricity price sensitivity index is 0.7, charging urgency index At 16:00, the dynamic electricity price adjustment coefficient is 0.72, according to the formula , calculate its response probability .

[0103] Users are divided into groups based on their response probability, with a response probability greater than 0.7 defined as a high-response probability user and a response probability less than 0.5 as a low-response probability user. For high-response probability users like User A, the system sends a priority charging instruction, recommending a charging period of 4:00 PM to 6:00 PM, informing them of the estimated cost of 0.72 RMB / kWh multiplied by the estimated charge volume (assuming an estimated charge volume of 40 kWh, the estimated cost is 28.8 RMB), and indicating that renewable energy accounts for 70% of the electricity consumption during this period. For users with a low response probability, the system sends a delayed charging recommendation, recommending charging between 10:00 PM and midnight. During this time, electricity prices may be lower due to different incentive levels, and the proportion of renewable energy is also relatively significant, guiding users to charge during off-peak hours and balancing the grid load.

[0104] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "includes," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0105] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A method for promoting electricity consumption by new energy vehicles, characterized in that: The method comprises: Step 1: Obtain real-time load data of the target area power grid, extract historical load fluctuation characteristics, and establish a dynamic load forecasting model for the power grid; Step 2: Based on the new energy vehicle user behavior data, the charging demand distribution in different time periods and geographical locations is counted to generate a spatiotemporal matrix of charging demand; Step 3: Identify the real-time output data of distributed renewable energy generation equipment in the target area and calculate the renewable energy absorption threshold based on the remaining capacity of the energy storage equipment; Step 4: Construct a multi-objective optimization function based on the grid dynamic load forecasting model and the charging demand spatiotemporal matrix. The objectives of the multi-objective optimization function include grid load balance, renewable energy utilization rate, and user charging cost. The specific formula for constructing the multi-objective optimization function in step 4 is: Among them, L net (t) is the net load of the power grid, L avg is the average load, P charge (t) is the charging power distribution value, C user is the user charging cost, w1, w2, w3 are weight coefficients, and satisfy w1+w2+w3=1; the net load L of the grid is obtained according to the dynamic load prediction model of the grid net (t); Determine the charging power allocation value P based on the charging demand spatiotemporal matrix and optimization algorithm charge (t); By continuously adjusting the charging power distribution value P charge (t), so that the value of F is minimized; Step 5: Based on the absorbable threshold and the charging period optimization allocation result, the electricity price incentive levels for different periods are divided, and the dynamic electricity price adjustment coefficient is calculated; The calculation process of the dynamic electricity price adjustment coefficient in step 5 includes: According to the absorbable threshold θ(t) and the charging power allocation value P charge The difference of (t) is used to divide the electricity price incentive level; Set the dynamic electricity price adjustment coefficient to: Among them, λ base is the benchmark electricity price, ε is the preset elasticity coefficient; Step 6: Extract historical charging behavior data of users, analyze their sensitivity to dynamic electricity prices, and establish a user response probability model; Step 7: Generate differentiated charging guidance instructions based on the user response probability model and the dynamic electricity price adjustment coefficient, and send them to the user terminal through the communication network.

2. The method for promoting electricity consumption by new energy vehicles according to claim 1, characterized in that: The method for establishing the power grid dynamic load prediction model in step 1 is: The historical load data is divided into multiple periods according to the time series, and the load characteristic parameters of each period are extracted, including peak load, valley load and load change gradient; the formula for constructing the load forecasting model is: Where L(t) is the predicted load value at time t; L base (t) is the baseline load at time t, calculated by the mean of historical data; ΔL k (t) is the kth abnormal load fluctuation, extracted by the outlier detection algorithm; α, β, and γ are the model parameters fitted by machine learning; t k is the time corresponding to the kth abnormal fluctuation in the historical load data.

3. The method for promoting electricity consumption by new energy vehicles according to claim 2, characterized in that: The process of generating the charging demand spatiotemporal matrix in step 2 includes: Divide the target area into multiple grid units and count the charging start time, charging duration, and charging power of new energy vehicles in each grid unit; Define the charging demand space-time matrix as a three-dimensional tensor D i,j,t , where i represents the longitude partition, j represents the latitude partition, and t represents the time slice. The value of each element is the total charging demand of the corresponding grid cell in time slice t.

4. The method for promoting electricity consumption by new energy vehicles according to claim 3, characterized in that: The method for calculating the renewable energy absorbability threshold in step 3 is: Obtain the real-time output P of distributed renewable energy gen (t) and the remaining capacity C of the energy storage device bat (t); Set the absorption threshold to: Among them, Δt is the scheduling time interval, P grid_max The maximum reverse feed power allowed by the grid.

5. The method for promoting electricity consumption by new energy vehicles according to claim 1, characterized in that: The method for establishing the user response probability model in step 6 is: Extract the electricity price sensitivity index s from the user's historical charging behavior U and charging urgency index d U ; Define the user response probability as: Where a, b, and c are the coefficients fitted by logistic regression.

6. The method for promoting electricity consumption by new energy vehicles according to claim 5, characterized in that: The process of generating differentiated charging guidance instructions in step 5 includes: According to the user response probability P U (t) Divide user groups, send priority charging instructions to users with high response probability, and send delayed charging suggestions to users with low response probability; The charging guidance instruction includes recommended charging time period, estimated cost and renewable energy ratio information.

7. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the method according to any one of claims 1 to 6.

8. A computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Images