Shared electric vehicle multi-mode charging network optimization method based on data driving

Abstract

The invention discloses a shared electric vehicle multi-mode charging network optimization method based on data driving. The method comprises the following steps that S1, track data and charging events of a shared electric vehicle are collected and extracted; s2, carrying out the clustering of the track data of the shared electric vehicle, and recognizing the positions of candidate fixed charging stations and mobile charging stations; s3, constructing a dual-objective optimization model, balancing the early-stage cost of an operator and the average travel delay of a user, and generating multi-mode charging network configuration through solving; and S4, performing cost and travel delay evaluation on the configuration, and feeding back to the dual-objective optimization model for iterative optimization to finally obtain a Pareto optimal solution set for configuring and optimizing the fixed charging station and the mobile battery swap station. According to the invention, a traditional fixed station and on-demand mobile battery changing vehicle service are combined, the input cost of an operator and the service quality of a user are balanced by constructing double targets, and compared with the prior art, the multi-mode charging network has the advantage that the travel delay can be obviously reduced under the similar cost input.

Description

Technical Field

[0001] This invention relates to the technical fields of traffic management and new energy, and in particular to a data-driven method and system for optimizing a multi-mode charging network for shared electric vehicles. Background Technology

[0002] Shared electric vehicles (SEVs), as an emerging mobility technology, have potential in reducing reliance on private cars and lowering carbon emissions from urban travel. Despite these advantages, limited battery capacity can cause range anxiety for users. Specifically, on the demand side, shared electric vehicles exhibit dynamic spatiotemporal usage patterns, significantly different from private cars and electric buses. Some shared electric vehicle users drive to remote areas for outings, resulting in uneven spatial and temporal distribution. On the supply side, deploying fixed charging infrastructure is costly, and covering all areas with fixed charging stations and battery swapping stations is economically infeasible. Furthermore, the uneven distribution of charging resources further restricts the popularization and development of shared electric vehicles. In summary, the mismatch between supply and demand severely hinders the efficient operation and market development of shared electric vehicles.

[0003] Currently, shared electric vehicles operating in cities and suburbs face challenges such as fluctuating demand and a lack of charging infrastructure. In response, mobile battery swapping stations have emerged. These stations are flexible, mobile facilities equipped with battery storage and swapping equipment, capable of quickly replacing fully charged batteries for electric vehicles at home or designated locations. The demand for shared electric vehicles exhibits dynamic spatiotemporal usage patterns, making it difficult for a single charging solution to serve diverse travel needs. Currently, there is limited work on designing multi-modal charging networks, with most research focusing on the characteristics of private car travel. Research on how to design multi-modal charging networks that integrate the spatiotemporal characteristics of shared electric vehicle and car travel remains lacking. Summary of the Invention

[0004] The purpose of this invention is to provide a data-driven method and system for optimizing a multi-mode charging network for shared electric vehicles. It combines the design of a charging network integrating fixed charging stations and mobile battery swapping stations, addressing technical issues such as insufficient charging infrastructure and range anxiety for shared battery vehicles. Furthermore, this invention integrates the characteristics of fixed charging stations and mobile battery swapping stations, providing charging services in areas with high demand and demand-responsive battery swapping services in areas with low demand.

[0005] To achieve the above objectives, the present invention provides the following technical solution: On one hand, this invention provides a data-driven method for optimizing a multi-mode charging network for shared electric vehicles, comprising the following steps: S1. Obtain relevant parameters of shared electric vehicles by collecting GPS data, including vehicle ID, longitude, latitude, driving distance, instantaneous speed and battery state of charge (SOC). The processed data is used to extract the trajectory data of shared electric vehicles (analyze the travel patterns of shared electric vehicles) and extract charging events. S2. Cluster the trajectory data of shared electric vehicles to identify the locations of candidate fixed charging stations and mobile battery swapping stations; S3. Construct a dual-objective optimization model to balance the operator's initial costs and the average user travel delay, and generate a multi-mode charging network configuration by solving the model. S4. The configuration is evaluated for cost and travel delay, and fed back to the bi-objective optimization model for iterative optimization. Finally, the Pareto optimal solution set is obtained, which is used to configure and optimize fixed charging stations and mobile battery swapping stations.

[0006] This invention combines traditional fixed charging stations with on-demand mobile battery swapping services, and balances the operator's investment costs with the user's service quality by constructing a dual-objective framework. Using real-world shared electric vehicle data as an example, this invention first mines the dynamic spatiotemporal characteristics of shared electric vehicles; secondly, comparing pure fixed charging station networks and fixed charging station-fixed battery swapping station networks, the multi-mode charging network proposed in this invention, with similar cost inputs, can reduce travel delays by 42.9% and 11.7% compared to the previous two models; finally, even when supply and demand conditions fluctuate, this model can still continuously provide more reliable service.

[0007] On the other hand, the present invention provides a data-driven multi-mode charging network optimization system for shared electric vehicles, which uses the above method and includes the following modules: Data acquisition and preprocessing module: used to collect and preprocess parameters of shared electric vehicles, and extract trajectory data and charging events of shared electric vehicles; Clustering and identification module: Clusters shared electric vehicle trajectory data to identify the locations of candidate fixed charging stations and mobile battery swapping stations; Dual-objective optimization module: used to build a dual-objective optimization model to balance the operator's upfront costs and the average user travel delay, and generate multi-mode charging network configurations by solving the problem; Evaluation and optimization module: Evaluates the configuration for cost and travel delay, and feeds the results back to the dual-objective optimization module for iterative optimization, ultimately obtaining the Pareto optimal solution set, which is used to configure and optimize fixed charging stations and mobile battery swapping stations.

[0008] Compared with the prior art, the present invention has the following beneficial effects: This invention targets shared electric vehicles, employing a data-driven, multi-mode charging network design. This approach reduces upfront investment costs for operators while improving user convenience, fully leveraging the complementary advantages of two models: fixed charging stations efficiently serve high-demand areas but lack fairness when covering the entire region; while mobile battery swapping stations provide a flexible and cost-effective charging solution for low-demand or underserved areas. Extensive experiments demonstrate that the proposed fixed charging station-mobile battery swapping station model effectively addresses the spatiotemporal dynamics of shared electric vehicle demand. Compared to a baseline model, this invention reduces average travel delay by 42.9% with a fixed charging station-only network and by 11.7% with a fixed charging station-battery swapping station network.

[0009] Further analysis demonstrates that the fixed charging station-mobile battery swapping station model can provide reliable service under dynamic demand and supply disruptions. It maintains consistent performance in both urban and suburban areas, with delays minimized even during peak demand periods such as holidays (e.g., an increase of no more than 0.1 minutes), and exhibits strong resilience in the face of infrastructure failures such as regional power outages or charging station disruptions. These advantages are crucial for enhancing user confidence and encouraging wider adoption of shared electric vehicles. Furthermore, multi-scenario experiments show that the model can adapt to evolving charging technologies and is suitable for cities with existing charging infrastructure, highlighting its versatility and long-term planning value. Attached Figure Description

[0010] Figure 1 This is a schematic diagram of a fixed charging station-mobile battery swapping station network in Embodiment 1 of the present invention; Figure 2 This is a schematic diagram of the overall process of Embodiment 1 of the present invention; Figure 3 This is a schematic diagram illustrating the selection between FCS and MCD considering completion time in Embodiment 1 of the present invention; Figure 4 This is a schematic diagram illustrating the trade-off between initial cost and travel delay in the fixed charging station-mobile battery swapping station network in Embodiment 1 of the present invention; Figure 5 This is a schematic diagram of the overall structure of Embodiment 2 of the present invention. Detailed Implementation

[0011] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0012] Example 1: Please see Figures 1-4 A data-driven optimization method for multi-mode charging networks of shared electric vehicles includes the following steps: Step 1: Collect and preprocess the parameters of the shared electric vehicles, and extract the trajectory data and charging events of the shared electric vehicles.

[0013] The system collects driving trajectory data for shared electric vehicles using GPS driving records and other methods. Each record includes the vehicle ID, longitude, latitude, driving distance, instantaneous speed, and battery state of charge (SOC, defined as the percentage of remaining charge to full charge). To ensure data quality, the data is cleaned and denoised to remove abnormal records such as incorrect timestamps, speeding (>150km / h), invalid SOC values ​​(>100% or sudden changes), and coordinates outside the study area. The processed data is used to analyze the travel patterns of shared electric vehicles and extract charging events.

[0014] No. The trajectory data of a vehicle is represented as the following sequence of tuples: ; In the formula, For vehicle location, The coordinates are latitude and longitude. For timestamps; Instantaneous velocity This indicates the battery's state of charge.

[0015] The journey begins when the vehicle's speed increases from 0; the journey ends when the speed remains at 0 for 5 consecutive times for at least 3 minutes. The s-th journey of the v-th vehicle is: ; In the formula, , These are the coordinates of the start and end points of the journey, respectively. , These are the start and end times, , These represent the battery charge states at the start and end points, respectively. This represents the amount of electricity consumed.

[0016] Charging events are identified based on trip data and battery state of charge (SOC) level. The conditions under which a vehicle needs charging are defined as follows: (1) When the SOC level is below 20%, it means that the battery life may be affected; (2) The remaining power is not enough to meet the energy consumption requirements of the next trip.

[0017] The charging event number is Each event is described by the following five attributes: ; In the formula, The location where the charging event occurred; , These are the time the charging event was generated (event identification time) and the estimated departure time of the next trip, respectively. for The state of charge of the battery at that time; For indicator variables, default value , indicating unallocated This indicates that the vehicle will be assigned to a fixed charging station. This indicates allocation to mobile battery swapping stations. This indicates that charging is complete.

[0018] S2. Cluster the trajectory data of shared electric vehicles to identify the locations of candidate fixed charging stations and mobile battery swapping stations.

[0019] Candidate locations for stationary charging stations (FCS) and mobile battery swapping stations (MCD) are key inputs for designing optimization models for multimodal charging networks. These locations are independently identified from shared electric vehicle datasets, taking into full account the different characteristics of the two types of facilities.

[0020] For candidate fixed charging stations If the State of Charge (SOC) of a shared electric vehicle increases during a parking period and includes at least six consecutive records or remains stable for 15 minutes, it is considered a charging activity. To eliminate bias caused by GPS data errors or large parking lots, agglomerative hierarchical clustering is used to merge similar records into a single fixed charging station location, resulting in a candidate set of fixed charging stations. ; in, ; Each candidate fixed charging station The state is defined by a six-tuple: For fixed charging station locations; , These are the number of charging piles at fixed charging stations and their utilization rate; Service radius of fixed charging stations Internal demand coverage; For fixed charging stations With charging events The distance between them; The set of charging completion times; This is a binary variable indicating whether or not the site should be built. The condition symbol for a set; It means "any".

[0021] For candidate mobile battery swapping stations The DBSCAN (Density-Based Spatial Clustering of Applications with Noise) spatial clustering algorithm was used to cluster charging events. Cluster centers represent hotspots of shared electric vehicle charging events and serve as candidate mobile battery swapping station locations. The set of candidate mobile battery swapping stations is denoted as: ; in, ; Each candidate mobile battery swapping station The state is described by a six-tuple: Location of the mobile battery swapping station; , These are the number of mobile charging vehicles and their utilization rate; Service radius of mobile battery swapping stations Internal mobile battery swapping station The set of charging events covered; The set of times for the mobile battery swapping vehicle to return to the warehouse; This is a binary variable indicating whether the mobile battery swapping station should be deployed. for With charging events The distance between them.

[0022] Therefore, the set of charging events is: ; in, ; ; In addition to the five attributes already described, the charging demand attributes include: In order to be in The service radius of the fixed charging station at the location Similarly, the existing collection of fixed charging stations within the area, In order to be in Service radius of mobile battery swapping stations The collection of mobile battery swapping stations that have already been deployed.

[0023] S3. Construct a dual-objective optimization model to balance the operator's initial costs and the average user travel delay, and generate a multi-mode charging network configuration by solving the model.

[0024] For operators, deploying fixed charging stations requires determining the number of charging piles at each site, while deploying mobile battery swapping stations requires deciding on the number of mobile battery swapping vehicles at each site, with a primary focus on upfront costs. Assuming a stable power supply for fixed charging stations and that mobile battery swapping stations carry sufficient batteries, empirical analysis suggests that one battery per mobile charging vehicle per mission is sufficient to meet demand. For users, the main concern is travel delays, and they expect the mobile battery swapping vehicle to be fully charged when they place an order.

[0025] Therefore, the constructed dual-objective optimization model needs to satisfy two objectives: Minimize the upfront cost of the charging network: (1a) The first four terms of the formula represent the annualized infrastructure construction cost, and the last term represents the total number of charging events. The charging and battery swapping operating costs, through Annualization is used to maintain consistency.

[0026] In the formula, Cost per fixed charging station; Cost per charging station; The construction cost of each mobile battery swapping station; Cost per mobile charging vehicle; This is the coefficient used to convert to years; The charging operation cost of fixed charging stations; The operating cost of battery swapping for mobile battery swapping stations; For all charging events The charging and battery swapping operating costs, through Annualization to maintain consistency; Minimize average trip delays for users by optimizing the layout of charging facilities to reduce user wait times for charging: (1b) In the formula, For charging events The delay time; For the first A collection of vehicle journeys.

[0027] To solve the above model, the following constraints need to be set: To ensure that each completed fixed charging station and mobile battery swapping station covers at least C charging events, ; (1c) (1d) In the formula, For charging events Whether or not Covered; For charging events Whether or not Covered.

[0028] The Big M method is used to mandate that the spacing between facilities is no less than their service radius. (1e) (1f) In the formula, For fixed charging stations , The distance between them; The service radius of a fixed charging station; This is the Big M method; For mobile battery swapping stations , The distance between them; The service radius of a mobile battery swapping station is limited by the number of charging piles or mobile vehicles per facility, reflecting the upper limit of space or resources. (1g) (1h) In the formula, The maximum number of mobile charging vehicles or charging stations; This is a dimensionless coefficient that converts the coverage requirement of fixed charging stations into the number of charging piles. The value is a dimensionless coefficient, which converts the coverage demand of mobile battery swapping stations into the number of mobile charging vehicles.

[0029] Define the domain of decision variables for the parameters. Obtain it using the NSGA-II algorithm to provide insights into the trade-offs between cost and delay in planning schemes.

[0030] (1i) (1j) It is the set of positive integers.

[0031] S4. Evaluate the configuration for cost and travel delay, and feed the results back to the bi-objective optimization model for iterative optimization. Finally, obtain the Pareto optimal solution set, which is used to optimize and configure fixed charging stations and mobile battery swapping stations.

[0032] In one specific embodiment, in the current iteration, after the planning scheme is given by the bi-objective optimization model, data-driven operational simulation is used to evaluate charging costs and travel delays, and the results are fed back to the bi-objective optimization model for the next iteration.

[0033] First, the upfront cost is calculated using expression (1a). Given the multi-mode charging network configuration output by the bi-objective optimization model, the facility cost is: The charging cost is The charging costs are calculated differently for fixed charging stations and mobile battery swapping stations.

[0034] For users of fixed charging stations, the charging cost is: ; in, ; In the formula, The cost per kWh of electricity; Battery capacity; For shared electric vehicles to reach fixed charging stations SOC value at that time; For charging events Arrive at a fixed charging station The distance; To extend the driving range of shared electric vehicles.

[0035] For mobile battery swapping station users, the battery swapping cost is: ; In the formula, The cost per kWh for battery swapping (including service fees).

[0036] Secondly, when assessing average trip delay, charging events should be considered. The choice between a fixed charging station and a mobile battery swapping station depends on the completion time, which mainly consists of the time to access a fixed charging station, the waiting time, and the charging time at the fixed charging station or the battery swapping time at the mobile battery swapping station.

[0037] like Figure 3 As shown, charging event Defined by two timestamps: Time of charging event generation ; Expected departure time for the next trip ; If the completion time exceeds If this happens, travel delays will occur. (If some charging events are outside the service area of ​​the planned facilities, the nearest facility (fixed charging station or mobile battery swapping station) will provide service).

[0038] With a minimum retained SOC of no less than 20%, for charging events Depending on the completion time, you need to choose between fixed charging stations and mobile battery swapping stations. Completion time mainly includes: Total charging time at fixed charging stations : (2) In the formula, For use in calculating at fixed charging stations This is a charging event. Total charging time; The time at which the charging event is generated; For charging events Drive from your current location to a designated charging station. The time spent together, The average speed of shared electric vehicles; For fixed charging stations Charging incident The charging time, of which For battery capacity, The SOC value of shared electric vehicles upon arrival at a fixed charging station. This refers to the charging power. For fixed charging stations The earliest completion time of the charging task indicates that the vehicle must wait in line.

[0039] Mobile battery swapping station For charging events Total battery replacement time provided for:

[0040] In the formula, For mobile battery swapping vehicles from battery swapping stations Departure, Arrival, Charging Event Time; Time taken to replace the battery; The earliest return time for the vehicle currently on the mission; The time it takes for the mobile charging vehicle to travel to vehicle k. The average speed of the mobile charging vehicle; Time taken to replace the battery.

[0041] Due to the charging incident It may have multiple fixed charging station options. and mobile battery swapping station options Therefore, a fixed charging station or a mobile battery swapping station needs to be selected for service. To calculate the minimum travel delay, it is assumed that the user prioritizes the facility that can complete charging or battery swapping the earliest, ideally without causing delays to the next trip. If at least one fixed charging station meets the conditions, the fixed charging station is preferred (because its business model is more mature and charging costs are lower). If the earliest completion time of all options exceeds [a certain timeframe], [further details are needed]. If this happens, it will cause a trip delay. Multiple charging events will be served on a "first-come, first-served" basis.

[0042] For each charging event The following three scenarios need to be addressed: 1. If there is a fixed charging station, you can charge at the expected departure time. Complete charging before ( Charging event (User) selects the earliest completed fixed charging station. : .

[0043] 2. If there is no fixed charging station, you can... Complete charging before ( However, at least one mobile battery swapping station can complete the battery swap on time. Charging event Select the earliest completed mobile battery swapping station : .

[0044] 3. If neither fixed charging stations nor mobile battery swapping stations can be used... Pre-completion service ( Users can select the earliest completed service option (fixed charging station or mobile battery swapping station): .

[0045] In the first two scenarios, fixed charging stations or mobile battery swapping station Available Service not completed in advance, resulting in delays. .

[0046] In the third scenario, all facilities are unable to... The previous trip was completed, but the next leg of the journey will be delayed by [time]. for: ; In the formula, For each candidate stationary charging station (FCS) The relevant set of charging times; For each candidate mobile battery swapping station (MCD) The relevant set of battery swapping times; To provide charging events within the service radius A collection of fixed charging stations (FCS) that provide services; To provide charging events within the service radius A collection of mobile battery swapping stations (MCDs) providing services; It is an empty set.

[0047] This refers to charging events not covered by any fixed charging station or mobile battery swapping station. Based on the nearest neighbor principle, the nearest fixed charging station or mobile battery swapping station is selected; similarly, if the charging / swapping completion time exceeds the departure time of the next leg of the journey, the delay needs to be calculated, as shown in Table 1.

[0048] Table 1. Pseudocode Table for Trip Delay Assessment

[0049] In one specific embodiment, the invention extracted 3,125 shared electric vehicle trips and 514 charging events within the target area during the third week of January 2019, as shown in Table 2, providing a GPS trajectory sample of shared electric vehicles. Subsequently, a clustering algorithm was applied to obtain 67 candidate fixed charging stations and 15 candidate mobile battery swapping stations.

[0050] Table 2 Sample GPS trajectory data of shared electric vehicles

[0051] The four types of upfront cost parameters in formula (1a) are set as follows: Fixed charging stations One million yuan per station per year for fixed charging piles Yuan / pile·year, mobile battery swapping station One million yuan / warehouse / year, mobile charging vehicle Yuan / vehicle / year. The benchmark value is based on the average charging and battery swapping prices in the target area. Yuan / kWh and Yuan / kWh. Maximum number of mobile charging vehicles and charging stations. All values ​​are limited to 10 to accelerate the algorithm's solution. The time step size is [not specified] within the one-week research period. Seconds; average speed of shared electric vehicles =30 km / h (based on shared electric vehicle data), mobile charging vehicle average speed 25 km / h. Charging pile power kW, battery swapping time of mobile charging vehicle minute.

[0052] like Figure 4As shown, this illustrates the trade-off between upfront costs and average travel delay under a baseline setting for a fixed charging station-mobile battery swapping station network. The curves exhibit an inverse relationship: as upfront costs increase, average travel delay decreases. Specifically, delays decrease rapidly within the 3–4 million RMB range, indicating high sensitivity; the rate of decrease slows in the 4–5 million RMB range, showing diminishing marginal returns; and beyond 5 million RMB, additional investment has limited impact on delay improvement. A similar trend is observed for fixed charging station networks alone, but the reduction in delay is significantly smaller.

[0053] Table 3 compares the average travel delays of each model under similar upfront costs. Due to different charging configurations, direct comparisons at the same cost point are not possible. It can be seen that the fixed charging station-mobile battery swapping station network has the lowest average travel delay and the best performance. Compared to a fixed charging station-only network, the fixed charging station-mobile battery swapping station network reduces the average travel delay by 42.9% while saving 1.2 times the upfront cost; compared to a fixed charging station-battery swapping station network, the delay is reduced by 11.7%, and the upfront cost is reduced by 5.0%. The results highlight the cost-effectiveness of the fixed charging station-mobile battery swapping station network: achieving a higher service level with lower economic investment.

[0054] Table 3 compares the fixed charging station-mobile battery swapping station, fixed charging station only, and fixed charging station-battery swapping station configurations under similar initial costs.

[0055] Charging events can also surge at specific times, especially during holidays. To simulate such a surge, this invention introduces 423 additional real-world shared electric vehicle charging events. The upfront costs and average trip delays under this scenario are summarized in Table 3. First, the average trip delay for the fixed charging station-mobile battery swapping station network increases by only 0.09 minutes, highlighting the robustness of the proposed multi-mode charging network. Second, the upfront costs for all three charging networks increase because operators must bear the additional costs of charging or battery swapping. Finally, the advantages of the fixed charging station-mobile battery swapping station network remain and become even more pronounced.

[0056] Under unstable charging supply conditions, firstly, in a 10% random failure scenario, the "fast charging station - mobile charging vehicle" (fixed charging station - mobile battery swapping station) and "fast charging station - battery swapping station" (fixed charging station - battery swapping station) networks are less affected. When a fast charging station (fixed charging station) fails, mobile charging vehicles (mobile battery swapping stations) and battery swapping stations (battery swapping stations) can provide supplementary services to meet charging demand. Secondly, this invention studied a power outage scenario in Tongzhou District. The results show that the "fast charging station - mobile charging vehicle" (fixed charging station - mobile battery swapping station) network is least affected, while networks relying solely on fast charging stations (fixed charging stations only) face greater pressure, leading to more travel delays. Finally, this invention also studied the impact of a power outage in a target area where charging events are relatively sparse. As a suburb with popular tourist attractions, all three models in this area are served by only one fast charging station. It can be seen that the advantages of the "fast charging station - mobile charging vehicle" (fixed charging station - mobile battery swapping station) network still exist. It is worth noting that the disadvantages of relying solely on fast charging stations (fixed charging stations only) have widened, specifically with an increase in average travel delays of 1.12 minutes.

[0057] As charging technology continues to advance, it is natural to verify whether the advantages of mobile battery swapping stations or multi-mode networks will still exist in future higher-power scenarios. Conversely, many cities have already built fixed charging stations, and planners hope to incorporate existing assets into multi-mode network designs. The results, as shown in Tables 4 and 5, indicate that the fixed charging station-mobile battery swapping station model, with similar initial costs, consistently outperforms the baseline model as charging power increases, providing better service. However, it is worth noting that the advantage in average travel delay compared to relying solely on fast charging stations (fixed charging stations only) gradually diminishes (from 29.8% to 27.3%), and it demonstrates significant advantages and robustness in scenarios with spatial heterogeneity, surging demand, and uncertainty.

[0058] Table 4. Impact of Dynamic Charging Supply on Average Travel Delay with Similar Initial Costs

[0059] Table 5 Comparison between the "Fast Charging Station - Mobile Charging Vehicle" (Fixed Charging Station - Mobile Battery Swapping Station) and the baseline model under different charging power.

[0060] This invention combines traditional fixed charging stations with on-demand mobile battery swapping services, and balances the operator's investment costs with the user's service quality by constructing a dual-objective framework. Using real-world shared electric vehicle data as an example, the invention first mines the dynamic spatiotemporal characteristics of shared electric vehicles; secondly, comparing pure fixed charging station networks and fixed charging station-fixed battery swapping station networks, the multi-mode charging network proposed in this invention, with similar cost inputs, can reduce travel delays by 42.9% and 11.7% respectively compared to the first two models; finally, even when supply and demand conditions fluctuate, this model can still continuously provide more reliable service.

[0061] Example 2 like Figure 5 As shown, a data-driven multi-mode charging network optimization system for shared electric vehicles, using the above method, includes the following modules: Data acquisition and preprocessing module: used to collect and preprocess parameters of shared electric vehicles, and extract trajectory data and charging events of shared electric vehicles; Clustering and identification module: Clusters shared electric vehicle trajectory data to identify the locations of candidate fixed charging stations and mobile battery swapping stations; Dual-objective optimization module: used to build a dual-objective optimization model to balance the operator's upfront costs and the average user travel delay, and generate multi-mode charging network configurations by solving the problem; Evaluation and optimization module: Evaluates the configuration for cost and travel delay, and feeds the results back to the dual-objective optimization module for iterative optimization, ultimately obtaining the Pareto optimal solution set, which is used to configure and optimize fixed charging stations and mobile battery swapping stations.

[0062] This invention discloses a data-driven multi-mode charging network optimization system for shared electric vehicles, which can be installed in a computer device. The computer device includes a processor, a memory, and a computer program stored in the memory and executable on the processor, such as a data-driven multi-mode charging network optimization program for shared electric vehicles. The memory includes at least one type of readable storage medium, such as flash memory, portable hard drive, multimedia card, card-type memory (e.g., SD or DX memory), magnetic storage, disk, optical disk, etc. The processor is the control core of the electronic device, connecting various components of the computer device via various interfaces and lines. It executes programs or modules stored in the memory and calls data stored in the memory to perform various functions of the computer device and process data.

[0063] The module described in this invention refers to a series of computer program segments that can be executed by the processor of a computer device and can perform a fixed function, and which are stored in the memory of the computer device.

[0064] The system provided in this embodiment of the invention has the same implementation principle and technical effects as the aforementioned method embodiment. For the sake of brevity, any parts not mentioned in the device embodiment can be referred to the corresponding content in the aforementioned method embodiment.

[0065] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those shown in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0066] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall fall within the scope of the present invention.

Claims

1. A data-driven optimization method for multi-mode charging networks of shared electric vehicles, characterized in that, Includes the following steps: S1. Collect and preprocess the parameters of shared electric vehicles, and extract the trajectory data and charging events of shared electric vehicles; S2. Cluster the trajectory data of shared electric vehicles to identify the locations of candidate fixed charging stations and mobile battery swapping stations; S3. Construct a dual-objective optimization model to balance the operator's initial costs and the average user travel delay, and generate a multi-mode charging network configuration by solving the model. S4. The configuration is evaluated for cost and travel delay, and fed back to the bi-objective optimization model for iterative optimization. Finally, the Pareto optimal solution set is obtained, which is used to configure and optimize fixed charging stations and mobile battery swapping stations.

2. The data-driven multi-mode charging network optimization method for shared electric vehicles according to claim 1, characterized in that, The parameters of the shared electric vehicles include vehicle ID, longitude, latitude, driving distance, instantaneous speed, and battery state of charge.

3. The data-driven multi-mode charging network optimization method for shared electric vehicles according to claim 1, characterized in that, The charging event is identified based on the shared electric vehicle's trip data and battery state of charge level. The conditions requiring charging are as follows: When the battery state of charge level is below 20%; The remaining battery power is insufficient to meet the energy requirements for the next trip.

4. The data-driven multi-mode charging network optimization method for shared electric vehicles according to claim 1, characterized in that, The location of the candidate fixed charging station is determined as follows: if the state of charge of the battery of a shared electric vehicle increases during a parking period and remains stable for at least 6 consecutive records or 15 minutes, it is determined as a charging behavior. Agglomerative hierarchical clustering is used to merge similar records into a single fixed charging station location. The location of the mobile battery swapping station is determined as follows: the charging events are clustered using the DBSCAN spatial clustering algorithm. The cluster centers represent the hotspots of shared electric vehicle charging events and serve as candidate mobile battery swapping station locations.

5. The data-driven multi-mode charging network optimization method for shared electric vehicles according to claim 1, characterized in that, The dual-objective optimization model needs to satisfy two objectives: Minimize the upfront cost of the charging network: ； Minimize the average travel delay for users: ； In the formula, Cost per fixed charging station; This is a binary variable indicating whether or not the fixed charging station should be built. Cost per charging station; This refers to the number of charging stations; The construction cost of each mobile battery swapping station; This is a binary variable indicating whether the mobile battery swapping station should be deployed. Cost per mobile charging vehicle; Number of mobile charging vehicles; This is the coefficient used to convert to years; The charging operation cost of fixed charging stations; The operating cost of battery swapping for mobile battery swapping stations; For indicator variables, default value , indicating unallocated This indicates that the vehicle will be assigned to a fixed charging station. This indicates allocation to mobile battery swapping stations. This indicates that charging is complete; For all charging events The charging and battery swapping operating costs, through Annualization to maintain consistency; For charging events The delay time; For the first A collection of vehicle journeys.

6. The data-driven multi-mode charging network optimization method for shared electric vehicles according to claim 5, characterized in that, Minimizing upfront costs and minimizing average travel delays requires satisfying the following constraints: Ensure that each completed fixed charging station and mobile battery swapping station covers at least C charging events. ; The distance between fixed charging stations and between mobile battery swapping stations shall not be less than their service radius. Limit the number of charging piles at fixed charging stations or the number of mobile charging vehicles at fixed charging stations; Define the feasible region of the decision variables.

7. The data-driven multi-mode charging network optimization method for shared electric vehicles according to claim 1, characterized in that, The trip delay assessment includes: Regarding charging events The choice between a fixed charging station and a mobile battery swapping station depends on the completion time, which includes the time to access the fixed charging station, the waiting time, and the charging time, and the battery swapping time at the mobile battery swapping station. Charging event Defined by two timestamps: Time of charging event generation ; Expected departure time for the next trip ; If the completion time exceeds If this happens, travel delays will occur.

8. The data-driven multi-mode charging network optimization method for shared electric vehicles according to claim 7, characterized in that, The charging events are handled on a first-come, first-served basis. For each charging event, the following three scenarios need to be processed: If there is a fixed charging station, it can be reached at the expected departure time. Before charging is completed, the user Select the earliest completed fixed charging station. ; If there is no fixed charging station, you can do so at your expected departure time. Charging must be completed beforehand, but at least one mobile battery swapping station can complete the swap on time for the user. Select the earliest completed mobile battery swapping station ; If neither the fixed charging station nor the mobile battery swapping station can be reached at the expected departure time For services to be completed earlier, users can select the earliest completed service option. ； The delay time is: ； In the formula, For fixed charging stations For charging events Total charging time for the services provided; For mobile battery swapping stations For charging events Total battery replacement time for services provided; For the first One fixed charging station; For the first One mobile battery swapping station; To provide charging events within the service radius A collection of fixed charging stations that provide services; To provide charging events within the service radius A collection of mobile battery swapping stations providing services; For each candidate fixed charging station The relevant set of charging times; To match each candidate mobile battery swapping station The relevant set of battery swapping times; The condition symbol for a set.

9. A data-driven multi-mode charging network optimization system for shared electric vehicles, using the method described in any one of claims 1-8, characterized in that, include: Data acquisition and preprocessing module: used to collect and preprocess parameters of shared electric vehicles, and extract trajectory data and charging events of shared electric vehicles; Clustering and identification module: Clusters shared electric vehicle trajectory data to identify the locations of candidate fixed charging stations and mobile battery swapping stations; Dual-objective optimization module: used to build a dual-objective optimization model to balance the operator's upfront costs and the average user travel delay, and generate multi-mode charging network configurations by solving the problem; Evaluation and optimization module: Evaluates the configuration for cost and travel delay, and feeds the results back to the dual-objective optimization module for iterative optimization, ultimately obtaining the Pareto optimal solution set, which is used to configure and optimize fixed charging stations and mobile battery swapping stations.

Images