A dynamic sharing-charging coupling scheduling optimization method and system

Abstract

This invention relates to a dynamic shared-charging coupled scheduling optimization method and system. The method includes: dividing a target area into multiple sub-regions; acquiring grid data, order data, charging pile data, and vehicle data for each region; and obtaining an objective function; constructing a loss function for dynamic shared-charging coupled scheduling based on the grid data, order data, charging pile data, and vehicle data; and using a federated learning algorithm for scheduling based on the objective function and the loss function to obtain vehicle behavior when the objective function is optimal. The system is used to implement the above method. Compared with existing technologies, this invention incorporates photovoltaic power generation into the vehicle charging scheduling considerations, constructs an objective function based on the benefits of multiple stakeholders in the charging scheduling process, ensures maximum benefit, and utilizes federated learning to protect the privacy and security of each charging aggregator during the scheduling process.

Description

Technical Field

[0001] This invention relates to the fields of vehicle scheduling and energy, specifically to a dynamic shared-charging coupled scheduling optimization method and system. Background Technology

[0002] New energy vehicles, as an emerging representative of flexible user-side resources, have achieved rapid development globally in recent years. With the dramatic increase in the number of electric vehicles, their charging load has become a significant factor that cannot be ignored in power grid operation. Especially against the backdrop of urban power grids already facing peak evening electricity demand, the charging behavior of electric vehicles during these peak hours further exacerbates the grid burden, making the "duck curve" phenomenon (i.e., lower electricity demand during the day and higher demand at night) more pronounced. Given the randomness and flexibility of electric vehicle charging time and location, they actually possess enormous scheduling potential. However, current research on electric vehicle charging scheduling mainly focuses on the perspectives of shared mobility platforms and urban power grids, neglecting the impact of photovoltaic power generation on the charging scheduling of new energy vehicles. This limitation leads to scheduling strategies failing to fully utilize the potential of photovoltaic energy and effectively addressing the grid load pressure brought by electric vehicle charging. Furthermore, when formulating collective scheduling strategies, it is impossible to avoid the exposure of charging-related data from various electric vehicle aggregators, failing to achieve privacy protection.

[0003] Therefore, the problem that needs to be solved is to provide a method that incorporates photovoltaic power generation into the charging scheduling considerations while paying attention to the data privacy of each electric vehicle aggregator. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the existing technology by providing a dynamic shared-charging coupled scheduling optimization method and system. This method constructs a benefit objective function for the main entities participating in the charging scheduling process, such as photovoltaic power generation companies, traditional power grids and shared mobility platforms, incorporates photovoltaic power generation into the charging scheduling considerations, makes full use of the potential of photovoltaic energy, and can generate more accurate scheduling schemes.

[0005] The objective of this invention can be achieved through the following technical solutions:

[0006] According to a first aspect of the present invention, a dynamic shared-charging coupled scheduling optimization method is provided, the method comprising:

[0007] The target area is divided into multiple sub-regions and numbered. Power grid data, order data, charging pile data and vehicle data of each region are obtained, and the objective function is obtained. The objective function includes the shared platform benefit function, the regional fairness benefit function and the photovoltaic power generation utilization function.

[0008] Based on the aforementioned power grid data, order data, charging pile data, and vehicle data, a loss function for dynamic sharing-charging coupled scheduling is constructed. The loss function includes the photovoltaic power generation company benefit loss function, the traditional power grid loss benefit function, and the shared mobility platform benefit loss function.

[0009] Based on the aforementioned objective function and loss function, a federated learning algorithm is used for scheduling to obtain the vehicle behavior when the objective function is optimal. The vehicle behavior includes: charging and charging sub-area number or order acceptance and heading to order acceptance sub-area number.

[0010] As a preferred technical solution, the objective function is obtained by weighted summation of the shared platform benefit function, the regional fairness benefit function, and the photovoltaic power generation utilization rate function.

[0011] As a preferred technical solution, the expression for the photovoltaic power generation company's benefit loss function is as follows:

[0012]

[0013] Where T represents the total time, P t G represents the electricity price during time period t. t sp represents the amount of electricity generated by photovoltaic power during time period t. t d represents the penalty cost for overflowing electricity. t Let t be the power grid demand during time period t.

[0014] As a preferred technical solution, the expression for the traditional power grid loss benefit function is as follows:

[0015]

[0016] L grid,peak =λ∑ t∈PeakHours s t ,

[0017] Among them, s t The time period t represents the total amount of electricity supplied, including electricity generated by photovoltaic power generation and other methods. t Let λ represent the grid demand during time period t, and λ represent the additional cost factor during peak hours.

[0018] As a preferred technical solution, the expression for the benefit loss function of the shared mobility platform is:

[0019]

[0020] Where, d j c represents the mileage traveled by vehicle j. j This represents the cost per unit mileage, and M represents the total number of vehicles.

[0021] As a preferred technical solution, the loss function is obtained by weighted summation of the photovoltaic power generation company's benefit loss function, the traditional power grid's benefit loss function, and the shared mobility platform's benefit loss function, and its expression is:

[0022] L=αL PV +βL grid +γL platform ,

[0023] Where α, β, and γ represent the weight coefficients of each loss function; L PV Represents the profit loss function of photovoltaic power generation companies, based on L PV,usage and L PV,spill Obtain; L grid Represents the traditional power grid loss benefit function, based on L grid,demand and L grid,peak Obtain; L platform This represents the loss function of revenue for ride-sharing platforms.

[0024] As a preferred technical solution, the method for scheduling using a federated learning algorithm includes:

[0025] Local update: Obtain all electric aggregators in each sub-region, calculate the loss function and gradient value for each electric aggregator based on its objective function, and perform local optimization and update on the loss function;

[0026] Aggregate Update: Aggregate the loss functions of all electric vehicle aggregators after local optimization updates to obtain an aggregated loss function. Perform global optimization updates based on the aggregated loss function and distribute the results of the global optimization updates to each electric vehicle aggregator for the next local optimization update.

[0027] Iteration: Repeat local updates and aggregate updates until the algorithm converges or the training objective is achieved;

[0028] Scheduling: Using a pre-trained federated learning algorithm, the vehicle behavior is obtained when the objective function is optimal.

[0029] As a preferred technical solution, the aggregation method is as follows:

[0030]

[0031] in, |D represents the local optimization update of the i-th sub-region at time t. i | represents the local data of the i-th sub-region, N represents the total number of sub-regions, and j represents the j-th sub-region.

[0032] As a preferred technical solution, the training objective is: the rate of change between the loss function after the previous optimization and the loss function after the next optimization and update is less than a preset value.

[0033] According to a second aspect of the present invention, a dynamic shared-charging coupled scheduling optimization system is provided, the system being used to implement the above-described method.

[0034] Compared with the prior art, the present invention has the following advantages:

[0035] 1) This invention provides a charging scheduling method for electric ride-hailing vehicles under the background of the new stage of photovoltaic development. By constructing an objective function and a loss function that include the benefits of photovoltaic power generation, the method aims to improve the utilization rate of photovoltaic power and schedule the charging of ride-hailing vehicles more rationally. This can effectively reduce the adverse effects of the "duck curve" and achieve peak shaving and valley filling, thus promoting the efficient use of green energy. It also ensures that the comprehensive benefits of photovoltaic power generation companies, ride-hailing platforms and power grid companies are maximized in the charging and scheduling process.

[0036] 2) This invention also employs federated learning technology, enabling each charging aggregator to train independently based on local data, thereby improving data privacy protection and model training efficiency, and avoiding privacy leakage issues in centralized data processing. Attached Figure Description

[0037] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0038] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0039] Obviously, the accompanying drawings described below are merely some examples or embodiments of this application. Those skilled in the art can apply this application to other similar scenarios based on these drawings without any inventive effort. Furthermore, it is understood that although the efforts made in this development process may be complex and lengthy, for those skilled in the art related to the content disclosed in this application, any changes to design, manufacturing, or production based on the technical content disclosed in this application are merely conventional technical means and should not be construed as insufficient disclosure of the content of this application.

[0040] This embodiment uses a city as an example to optimize the dynamic vehicle sharing-charging coupled scheduling. The process is as follows: Figure 1 As shown, the following steps are included:

[0041] S1. Data and Objective Function Acquisition:

[0042] S11. Divide the sample city into multiple sub-regions and number them, and determine the main body in each sub-region and the purpose that each main body hopes to achieve.

[0043] The main entities include: traditional power companies, ride-sharing platforms, photovoltaic power generation companies, vehicle fleets, passengers, and charging stations. Specifically, 1) traditional power grid companies hope to rationally schedule vehicle charging to achieve peak shaving and valley filling, and alleviate grid pressure; 2) ride-sharing platforms hope to maximize profits while ensuring passenger service quality; 3) photovoltaic power generation companies hope to improve photovoltaic utilization and reduce photovoltaic waste; 4) vehicle fleets are completely controlled by ride-sharing platforms in the model; 5) passengers hope to have their service quality guaranteed; and 6) charging stations hope to serve as many vehicles as possible.

[0044] S12. Obtain the data generated by each entity in each sub-region, including power grid data, order data, charging pile data, and vehicle data.

[0045] The sample data includes partial data requested throughout the day from the city. Order data includes: number of orders (30,309), order ID, order start time, order start point, and latitude and longitude of order destination; vehicle data includes: maximum vehicle range, full charge time, and mileage per kilowatt-hour; power grid data includes: photovoltaic power generation, photovoltaic power generation percentage, photovoltaic subsidies, power grid demand, and total power grid generation; and charging pile data includes: number of charging piles, charging pile number, and charging voltage.

[0046] S13. Combining the objective the subject hopes to achieve and the data obtained in step S12, obtain the objective function. The expression of the objective function is: F = max{F1, F2, F3}. More specifically, the objective function can be represented by constructing a Pareto formula, whose expression is:

[0047]

[0048] in, Let represent the normalization function of the ζ-th objective function in the k-th optimal solution; This represents the ζ-th objective function in the k-th optimal solution; Let represent the minimum value of the ζ-th objective function; Let ζ represent the maximum value of the ζ-th objective function; ζ represents the type of objective function, and when ζ = 1, it represents the travel benefits of the shared platform; when ζ = 2, it represents the regional equity benefits; and when ζ = 3, it represents the photovoltaic power generation utilization rate.

[0049] The objective function includes:

[0050] A. The benefit function of the shared platform, obtained from charging pile data and vehicle data, is expressed as follows:

[0051]

[0052] F1 represents the efficiency of ride-sharing platforms. This refers to the collection of ride-hailing vehicles participating in the dispatch. This represents the set of travel requests participating in the scheduling. This indicates the revenue generated from vehicle pick-up and drop-off orders. This represents the cost incurred from vehicle pick-up and drop-off orders. Indicates the regional photovoltaic power generation subsidy coefficient. This indicates the cost incurred by the vehicle traveling to a charging station. Let x represent the set of charging stations participating in the scheduling, k represent the charging station number, and x represent the charging station number. ij Let y represent the decision variable for vehicle i picking up and dropping off order j. ik Let P represent the decision variable for vehicle i to travel to charging station k. pv (t) represents the photovoltaic power generation, P max (t) represents the vehicle's charging power.

[0053] The revenue generated from vehicle pick-up and drop-off orders consists of two parts: a commission based on the sum of the starting price, mileage, and travel time, and a commission on the remaining amount as a service fee. The calculation formula is as follows:

[0054]

[0055] Among them, p sta p represents the starting price for the order. km This indicates the cost per kilometer of the order, s j p represents the order distance. min This represents the service fee per minute for the order, t j Indicates the order time, c pla θ represents the percentage of the total order amount that is taken as commission. pla This represents the commission paid out on the remaining amount of the order.

[0056] The cost of vehicle pick-up and drop-off orders is primarily determined by the distance between the vehicle's current location and the order's pick-up point. It's necessary to explain the platform's loss for every 1km the vehicle travels empty. km "This is calculated based on information such as the local drivers' hourly wage and the platform's order revenue, and the formula is as follows:"

[0057]

[0058] And d ij This indicates the distance between the vehicle and the order.

[0059] The regional photovoltaic power generation subsidy coefficient is determined by the local photovoltaic power generation company, as the company hopes that vehicles will charge during peak photovoltaic power generation periods to improve the utilization rate of photovoltaic power.

[0060] The cost of traveling to a charging station is primarily determined by the current electricity price, the distance between the vehicle and the charging station, and the estimated waiting time at the charging station. It can be represented by three parts: the cost of traveling the distance from the vehicle's current location to the charging station, the estimated waiting cost at the charging station, and the charging cost, as detailed below:

[0061]

[0062] in, Indicates the estimated waiting time for the vehicle to charge at the charging station, L min This indicates the loss incurred by the platform for every minute of idle time. Indicates the estimated charging time for the vehicle, e price This indicates the electricity price.

[0063] B. The regional equity benefit function, obtained from vehicle data, is expressed as follows:

[0064]

[0065] Where p represents a subregion, P represents the set of subregions, and SAV p TOTAL represents the number of vehicles in region p after scheduling. SAV RE represents the total number of vehicles participating in the dispatch. p TOTAL represents the predicted number of requests from region p to this subregion. RE This indicates the total number of requests; the number of vehicles in region p after scheduling includes the original number of vehicles in region p, the number of vehicles going to region p due to orders, and the number of vehicles going to region p due to charging.

[0066] The number of vehicles in region p after dispatch consists of the original number of vehicles in the region, the number of vehicles that went to the region due to orders, and the number of vehicles that went to the region for charging. The formula for calculating the number of vehicles that went to the region due to orders is as follows:

[0067]

[0068] The formula for calculating the number of vehicles heading to the charging area is:

[0069]

[0070] C. The photovoltaic power generation utilization rate function, obtained from grid data, is expressed as follows:

[0071]

[0072] Where F3 represents the photovoltaic power generation utilization rate, T represents the target time period, and P pv (t) represents the photovoltaic power generation, P sav (t) represents the vehicle's charging power.

[0073] S2. Loss Function Construction:

[0074] A loss function for dynamic shared-charging coupled scheduling is constructed based on grid data, order data, charging pile data, and vehicle data. This loss function includes the photovoltaic power generation company's benefit loss function, the traditional grid's benefit loss function, and the shared mobility platform's benefit loss function. Its expression is as follows:

[0075] L=αL PV +βL grid +γL platform ,

[0076] Where α, β, and γ represent the weight coefficients of each loss function; L PV Represents the profit loss function of photovoltaic power generation companies, based on L PV,usage and L PV,spill Obtain; L grid Represents the traditional power grid loss benefit function, based on L grid,demand and L grid,peak Obtain; L platform This represents the loss function of revenue for ride-sharing platforms.

[0077] Specifically, the expression for the profit loss function of a photovoltaic power generation company is as follows:

[0078]

[0079] Where T represents the total time, P t G represents the electricity price during time period t. t sp represents the amount of electricity generated by photovoltaic power during time period t. t d represents the penalty cost for overflowing electricity. t Let t be the power grid demand during time period t.

[0080] The traditional power grid loss-benefit function is expressed as follows:

[0081]

[0082] L grid,peak =λ∑ t∈PeakHours s t ,

[0083] Among them, s t The time period t represents the total amount of electricity supplied, including electricity generated by photovoltaic power generation and other methods. t Let λ represent the grid demand during time period t, and λ represent the additional cost factor during peak hours.

[0084] The expression for the benefit loss function of the ride-sharing platform is:

[0085]

[0086] Where, d j c represents the mileage traveled by vehicle j. j This represents the cost per unit mileage, and M represents the total number of vehicles.

[0087] S3. Scheduling using federated learning algorithms:

[0088] Based on the constructed objective function and loss function, a federated learning algorithm is used to obtain the vehicle behavior when the objective function has an optimal solution. The federated learning algorithm consists of multiple local blocks: electric vehicle aggregators and a central server.

[0089] Detailed steps include:

[0090] S31. Local Update: Obtain all electric vehicle aggregators in each sub-region, calculate the loss function and gradient value for each aggregator based on its local data, and perform local optimization and update of the loss function. After the local optimization and update is completed, each electric vehicle aggregator uploads the results of the local optimization and update to the central server.

[0091] S32, Aggregated Update: The central server aggregates the locally optimized and updated loss functions of all electric vehicle aggregators to obtain the aggregated loss function. This aggregation process is usually a weighted average.

[0092]

[0093] This represents the local optimization update result of the i-th charging aggregator, where N represents the total number of charging aggregators, and D represents the total number of charging aggregators. i This represents the result of the local optimization and update by the charging aggregator, where j represents the j-th charging aggregator.

[0094] Global optimization and updates are performed based on the aggregation loss function, and the results of the global optimization and updates are distributed to each electric vehicle aggregator for use in the next local optimization and updates of the charging aggregator.

[0095] S33. Iteration: Repeat local updates and aggregate updates until the algorithm converges or the rate of change between the loss function after the previous optimization update and the loss function after the next optimization update is less than 0.01%.

[0096] S34. Scheduling: Using a trained federated learning algorithm, obtain the vehicle behavior when the objective function is optimal, i.e., the solution to the Pareto equation in step S13 is at its maximum value. The specific vehicle behaviors include: charging and charging sub-area number or order acceptance and heading to the order acceptance sub-area number.

[0097] This embodiment also provides a dynamic shared-charging coupling scheduling optimization system, which is used to implement the above method. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process described can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0098] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A dynamic shared-charging coupled scheduling optimization method, characterized in that, The method includes: The target area is divided into multiple sub-regions and numbered. Power grid data, order data, charging pile data and vehicle data of each region are obtained, and the objective function is obtained. The objective function includes the shared platform benefit function, the regional fairness benefit function and the photovoltaic power generation utilization function. The benefit function of the shared platform is: in, This indicates the benefits of ride-sharing platforms. This refers to the collection of ride-hailing vehicles participating in the dispatch. This represents the set of travel requests participating in the scheduling. This indicates the revenue generated from vehicle pick-up and drop-off orders. This represents the cost incurred from vehicle pick-up and drop-off orders. Indicates the regional photovoltaic power generation subsidy coefficient. This indicates the cost incurred by the vehicle traveling to a charging station. This represents the set of charging stations participating in the scheduling, where k represents the charging station number. Let i represent the decision variable for vehicle i picking up and dropping off order j. Let $\mathbf$ represent the decision variable for vehicle $i$ to go to charging station $k$. Indicates photovoltaic power generation capacity. Indicates the vehicle's charging power; The aforementioned regional equity benefit function is: ， in, Indicates a sub-region. Represents a set of subregions. Indicates after scheduling Number of vehicles in the area This indicates the total number of vehicles participating in the dispatch. express The predicted number of requests to that subregion. Indicates the total number of requests; after scheduling The number of vehicles in the area includes The original number of vehicles in the area, and the number of vehicles traveling due to orders. The number of vehicles in the area and those heading to charging stations Number of vehicles in the area; Based on the aforementioned power grid data, order data, charging pile data, and vehicle data, a loss function for dynamic sharing-charging coupled scheduling is constructed. The loss function includes the photovoltaic power generation company benefit loss function, the traditional power grid loss benefit function, and the shared mobility platform benefit loss function. The expression for the profit loss function of the photovoltaic power generation company is as follows: ， ， Where T represents the total time. This represents the electricity price during time period t. This represents the amount of electricity generated by photovoltaic power during time period t. This represents the penalty cost for excess electricity. Let t be the grid demand during time period t; The expression for the traditional power grid loss benefit function is as follows: ， ， in, This indicates that time period t includes the total amount of electricity supplied by photovoltaic power generation and other methods. Let λ represent the grid demand during time period t, and let λ represent the additional cost coefficient during peak hours. The expression for the benefit loss function of the ride-sharing platform is as follows: ， in, This represents the mileage traveled by vehicle j. This represents the cost per unit mileage, where M represents the total number of vehicles. The aforementioned loss function is obtained by weighted summation of the photovoltaic power generation company's benefit loss function, the traditional power grid's benefit loss function, and the shared mobility platform's benefit loss function, and its expression is: ， in, , and These represent the weighting coefficients of each loss function; The function representing the profit loss of a photovoltaic power generation company is based on... and get; Representing the traditional power grid loss benefit function, based on and get; This represents the loss function of revenue for ride-sharing platforms. Based on the aforementioned objective function and loss function, a federated learning algorithm is used for scheduling to obtain the vehicle behavior when the objective function is optimal. The vehicle behavior includes: charging and charging sub-area number or order acceptance and heading to order acceptance sub-area number.

2. The dynamic shared-charging coupled scheduling optimization method according to claim 1, characterized in that, The objective function is obtained by weighted summation of the shared platform benefit function, the regional equity benefit function, and the photovoltaic power generation utilization rate function.

3. The dynamic sharing-charging coupled scheduling optimization method according to claim 1, characterized in that, The method for scheduling using federated learning algorithms includes: Local update: Obtain all electric aggregators in each sub-region, calculate the loss function and gradient value for each electric aggregator based on its objective function, and perform local optimization and update on the loss function; Aggregate Update: Aggregate the loss functions of all electric vehicle aggregators after local optimization updates to obtain an aggregated loss function. Perform global optimization updates based on the aggregated loss function and distribute the results of the global optimization updates to each electric vehicle aggregator for the next local optimization update. Iteration: Repeat local updates and aggregate updates until the algorithm converges or the training objective is achieved; Scheduling: Using a pre-trained federated learning algorithm, the vehicle behavior when the objective function is optimal is obtained.

4. The dynamic sharing-charging coupled scheduling optimization method according to claim 3, characterized in that, The aggregation method is as follows: ， in, This represents the local optimization update of the i-th sub-region at time t. Let N represent the local data of the i-th sub-region, N represent the total number of sub-regions, and j represent the j-th sub-region.

5. The dynamic shared-charging coupled scheduling optimization method according to claim 3, characterized in that, The training objective is that the rate of change between the loss function after the previous optimization and the loss function after the next optimization is less than a preset value.

6. A dynamic shared-charging coupled scheduling optimization system, characterized in that, The system is used to implement the method as described in any one of claims 1 to 5.

Images