Electric vehicle charging station recommendation method and device based on charging preference of driver

Abstract

The invention relates to the field of big data analysis, in particular to an electric vehicle charging station recommendation method and device based on driver charging preference. According to the method, personal information and historical charging behavior data of a current user are collected, and a mileage anxiety threshold L of the user is updated. Collecting state information of a vehicle driven by a current user and operation data of a charging station; when a user puts forward a charging demand, candidate charging stations are screened firstly; calculating the payment cost Cb, the parking time cost Ct, the mileage anxiety cost Ch and the charging utility Ui of the user arriving at any candidate charging station i according to the collected information and the L of the current user; calculating the selection probability Pi of each candidate charging station according to the Ui; and the Ct and the Ch are ranked in a low-to-high order to generate a recommendation list. The problem that the existing charging station recommendation service cannot meet the personalized preference of the user is solved.

Description

Technical Field

[0001] The present invention relates to the field of big data analysis, and in particular to a method for recommending electric vehicle charging stations based on drivers' charging preferences, and a corresponding computer program product, storage medium, and data processing device. Background Art

[0002] With the increasing popularity of electric vehicles, electric vehicle charging stations, as essential infrastructure for the development of the automotive and energy industries, are rapidly developing. As a key driver of electric vehicle adoption, convenient charging plays a crucial role in improving the driver experience and promoting the sustainable development of the industry. Unlike fuel-powered vehicles, the actual range of electric vehicles fluctuates depending on factors such as low temperatures and driving habits. In the early days, insufficient charging facilities, long charging times, and the difficulty in accurately estimating the vehicle's remaining range exacerbated user anxiety. This mental stress caused by concerns about low battery levels while driving an electric vehicle is known as range anxiety.

[0003] While driving, users typically choose to charge their vehicles when their range falls below their personal tolerance. Most users rely on charging station recommendations from their electric vehicle's onboard system or third-party navigation software. Given that each driver has individual preferences for charging stations, their perceived utility value of the same charging station in the same driving scenario may vary. However, current charging station recommendation solutions often fail to fully account for these individual preferences. Consequently, the system typically provides consistent charging station recommendations for all drivers in the same driving scenario. This can result in the recommended charging stations not meeting the driver's individual preferences and even exacerbating congestion at popular charging stations, impacting the user's charging experience. Therefore, in-depth analysis of drivers' charging preferences and recommending the most suitable charging stations based on these preferences are crucial for improving the driver's charging experience. Summary of the Invention

[0004] In order to solve the problem that existing charging station recommendation services fail to accurately analyze user personalized preferences, resulting in poor user experience of recommendation results, the present invention provides an electric vehicle charging station recommendation method based on driver charging preferences, and its corresponding computer program product, storage medium and data processing device.

[0005] The technical solutions provided by the present invention are as follows: A method for recommending electric vehicle charging stations based on driver charging preferences includes the following steps: Collect personal information representing the current user's income level and historical charging behavior data, and update the user's range anxiety threshold accordingly L .

[0006] Collect the status information of the vehicle currently driven by the user and the operating data of all charging stations; when the user makes a charging request, first combine the vehicle's mileage and vehicle location to select the charging stations that can be reached around the vehicle as candidate charging stations, and form a candidate station set I .

[0007] The following formula combines the collected information and the current user's mileage anxiety threshold L Calculate the time it takes for the user to reach any candidate charging station i Payment cost C b , parking time cost C t , mileage anxiety costs C h and charging utility U i : , In the above formula, is the number of sites in the candidate site set; D i Indicates that the current user's vehicle has arrived at i The remaining range of each charging station; C i Indicates that the current user's vehicle is in i The total charging cost of each charging station.

[0008] According to the following formula U i Calculating selection probabilities P i : , In the above formula, Indicates the driver's perception of the charging utility.

[0009] According to user preferences, each candidate charging station is P i From high to low and by C b 、 C t and C h Sort the lists from low to high and return the generated recommendation list to the user.

[0010] As a further improvement of the present invention, the personal information includes: the driver's daily income R , daily working hours H Daily commuting time T w .

[0011] As a further improvement of the present invention, the historical charging behavior data includes: the remaining mileage of the electric vehicle corresponding to each historical charging behavior of the user, the selected charging station, the charging time, the charging cost, the location of the vehicle and each candidate charging station, and the distribution density of the candidate charging stations.

[0012] As a further improvement of the present invention, the vehicle status information includes: real-time location information, current mileage, total mileage D , power consumption per kilometer e .

[0013] As a further improvement of the present invention, any candidate charging station i Operational data includes: location of charging stations, number of charging piles n i , congestion level β i , charging rate m i and charging power W i .

[0014] As a further improvement of the present invention, the payment cost of any charging behavior C b The calculation formula is as follows: .

[0015] As a further improvement of the present invention, the time cost of any charging behavior C t The calculation formula is as follows: , In the above formula, C r Indicates that the vehicle is at a candidate charging station i Actual charging time cost; C l represents the psychologically expected queuing waiting time cost generated according to the congestion level of the candidate charging station; TC Indicates the unit travel time value of the current user.

[0016] As a further improvement of the present invention, the operating status of the charging station is divided into three types: busy, medium and idle. The corresponding congestion levels are β i The values ​​of are 0.5, 0.35 and 0 respectively.

[0017] As a further improvement of the present invention, the range anxiety cost of any charging behavior C h The calculation formula is as follows: , In the above formula, x Indicates the driving range of the electric vehicle; α is a preset power parameter; F ( x ) is a power function that represents the degree of change in the current user's mileage anxiety; C n Represents the anxiety cost per unit mileage.

[0018] As a further improvement of the present invention, the unit mileage anxiety cost C n Cost of maximum range anxiety C p Divide by the preset total mileage anxiety value to get. D i The towing fee generated after zeroing is used as the maximum mileage anxiety cost of the current trip C p .

[0019] As a further improvement of the present invention, the unit travel time value of the current user is TC The calculation formula is as follows: , In the above formula, Represents the travel urgency coefficient obtained based on the current user's travel purpose.

[0020] As a further improvement of the present invention, the user's travel purposes are work commuting, entertainment and emergency tasks, and the corresponding urgency coefficients of the three are The values ​​of are 1, 0.7, and 1.5 respectively.

[0021] As a further improvement of the present invention, the mileage anxiety threshold of any user L The update formula is as follows: , In the above formula, Indicates the ratio of basic mileage anxiety threshold to total cruising range, =0.2; S Indicates the quantitative value of the driver's driving style; represents the influence coefficient of driving style on range anxiety threshold; The coefficient indicating the impact of charging station distribution density on range anxiety threshold; Z represents the distribution density of candidate charging stations; Z 0 represents the base charging station distribution density; As a further improvement of the present invention, S 、 and The value of is evaluated based on the user's historical charging behavior data. Represents a conservative driver. represents an aggressive driver; ; .

[0022] The present invention also includes a computer program product, which includes a computer program. When the computer program is executed by a processor, it implements the steps of the aforementioned method for recommending electric vehicle charging stations based on the driver's charging preferences, and then generates a recommended list of candidate charging stations that meet the user's preferences based on the user's current vehicle status and the operating data of the surrounding charging stations, and returns the list to the user.

[0023] The present invention also includes a storage medium storing a computer program. When the computer program is executed by a processor, the computer program implements the steps of the aforementioned method for recommending electric vehicle charging stations based on the driver's charging preferences, and then generates a recommended list of candidate charging stations that meet the user's preferences based on the user's current vehicle status and the operating data of the surrounding charging stations, and returns the list to the user.

[0024] The present invention also includes a data processing device, which includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the computer program is executed by the processor, it implements the steps of the aforementioned method for recommending electric vehicle charging stations based on the driver's charging preferences, and then generates a recommended list of candidate charging stations that meet the user's preferences based on the user's current vehicle status and the operating data of the surrounding charging stations, and returns the list to the user.

[0025] The present invention has the following beneficial effects: Unlike traditional technologies that focus solely on a single factor (such as distance or price), this invention innovatively integrates multiple factors, including distance, price, time, and user experience, to comprehensively and realistically describe drivers' charging station selection behavior in real-world scenarios. By incorporating key factors such as range anxiety threshold, charging station congestion, and the value of travel time per unit, this solution can flexibly adapt to a variety of complex real-world scenarios and accurately describe the charging choices of different drivers in various driving scenarios. This multi-dimensional comprehensive consideration provides drivers with personalized charging station recommendations that align with their charging preferences and behaviors, significantly improving their charging satisfaction.

[0026] This invention fully utilizes historical charging selection behavior data to comprehensively analyze drivers' charging behavior characteristics. Through in-depth analysis of historical data, key parameters of each driver's range anxiety threshold model are obtained, accurately quantifying the driver's range anxiety and significantly improving the accuracy and reliability of the model's personalized recommendations. Furthermore, this invention also uses historical data to obtain specific environmental data, such as towing fees on urban roads, rural roads, and highways, to accurately characterize the range anxiety costs in different scenarios and further optimize the charging selection model to make it more tailored to actual needs.

[0027] This invention provides drivers with a one-stop charging station recommendation service by comprehensively considering multiple key factors, including charging costs, parking time, and range anxiety. The system flexibly adjusts charging recommendations based on different driving scenarios (e.g., urban roads, highways), the driver's travel purpose (e.g., commuting, entertainment, emergency travel), and by combining real-time vehicle data with information on the status of surrounding charging stations. This personalized, comprehensive charging station recommendation approach overcomes the limitations of traditional single-dimensional recommendations, eliminating the need for drivers to repeatedly weigh multiple factors and significantly reducing decision-making time and effort. The system also provides a list of recommended charging stations based on each single dimension (lowest charging cost, shortest parking time, and lowest range anxiety), providing comprehensive decision-making support to drivers and meeting diverse user needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 This is a flowchart of a method for recommending electric vehicle charging stations based on driver charging preferences provided in Example 1 of the present invention.

[0029] Figure 2 A schematic diagram of the interactive interface of the data processing device provided in Example 2 of the present invention.

[0030] Figure 3 Schematic diagram of a typical research scenario in a simulation experiment.

[0031] Figure 4 This is the influence curve of different unit travel time values ​​on the selection probability of each charging station in this scenario in the simulation experiment.

[0032] Figure 5 This is the influence curve of different charging powers on the selection probability of each charging station in this scenario in the simulation experiment.

[0033] Figure 6 This is the influence curve of different congestion levels in the simulation experiment on the selection probability of each charging station in this scenario.

[0034] Figure 7 This is the influence curve of different maximum range anxiety costs in the simulation experiment on the selection probability of each charging station in this scenario.

[0035] Figure 8 This is the influence curve of different mileage anxiety thresholds on the selection probability of each charging station in this scenario in the simulation experiment.

[0036] Figure 9 This is the influence curve of different remaining mileages on the selection probability of each charging station in this scenario in the simulation experiment. DETAILED DESCRIPTION

[0037] 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.

[0038] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this invention pertains. The terms used herein in the specification of the present invention are for the purpose of describing specific embodiments only and are not intended to limit the present invention. The term "or / and" as used herein includes any and all combinations of one or more of the associated listed items.

[0039] Example 1 This embodiment provides a method for recommending electric vehicle charging stations based on the driver's charging preference. Typical application scenarios of this method are as follows: Figure 1 As shown, when a user raises a charging demand, the system background can combine the collected various data to evaluate the comprehensive cost and selection probability of the user's choice of each charging station, and then return at least one recommendation list that meets the user's personalized preferences to the user's driver terminal based on the evaluation results for the user's reference.

[0040] Specifically, the method for recommending electric vehicle charging stations based on the driver's charging preferences provided in this embodiment includes the following steps: 1. Create user profiles based on historical charging behavior data In this embodiment, the current user's personal information and historical charging behavior data can be collected and the user's range anxiety threshold can be updated accordingly. L , thereby accurately characterizing users' charging preferences and behavior patterns.

[0041] In this embodiment, the collected personal information mainly refers to various data that can help determine the user's income level. Such data can be used to evaluate the user's time value and to perform cost accounting for the subsequent evaluation of the user's decision-making behavior of choosing different charging stations for charging. For example, in actual applications, the personal information collected for each user may include: the driver's daily income R (Unit: Yuan), daily working hours H (Unit: hours) and daily commuting time T w (Unit: hour). Based on the above data, the current user’s hourly wage can be obtained as R / ( H + T w Generally speaking, users with higher hourly wages have higher corresponding time costs and are more willing to spend higher economic costs to save time.

[0042] Historical charging behavior data is mainly used to evaluate each person’s range anxiety threshold. In this embodiment, the range anxiety threshold L Refers to the shortest driving range of an electric vehicle that each user can tolerate. When the remaining range is lower than the range anxiety threshold L When the battery is running low, the driver will feel anxious about running out of power, which is called "range anxiety". At this time, the user will have a strong desire to choose to charge rather than continue driving. In fact, the range anxiety threshold L This parameter is related to multiple factors and can have a significant impact on users' charging decisions. In the same driving scenario, different drivers have different range anxiety thresholds due to their different driving styles. Therefore, the same remaining range will cause different range anxiety values ​​for different drivers, resulting in different range anxiety costs. In different driving scenarios, the same driver's range anxiety threshold will also change based on their observation and judgment of the environment.

[0043] In order to more accurately analyze the user's charging decision preference, this embodiment first needs to quantify the range anxiety threshold of each user. The technicians of this embodiment analyzed and found that the range anxiety threshold of each user LThe factors that affect the range anxiety threshold mainly include the following: First, the total range of the vehicle is an important factor affecting the range anxiety threshold. The higher the range of the vehicle, the higher the driver's range anxiety threshold is usually. This is mainly because when the electric vehicle's battery level is less than 20%, the dashboard indicator light will warn that the vehicle needs to be charged. Therefore, 20% of the total range is a key node for the driver's range anxiety. Secondly, the density and distribution of charging stations have a significant impact on the range anxiety threshold. If charging stations are widely distributed and convenient, the driver's range anxiety threshold will be reduced. Finally, the driver's driving experience and charging habits will also affect the range anxiety threshold. An aggressive driving style (such as extreme charging) will lead to a decrease in the range anxiety threshold, while a conservative driving style (avoiding risks as early as possible) will lead to an increase in the range anxiety threshold.

[0044] Taking the above three key factors into consideration, this embodiment uses 20% of the vehicle's range as the basic component of the range anxiety threshold. On this basis, the threshold is dynamically adjusted according to the density of charging stations and driving style to more accurately reflect the changes in the driver's range anxiety threshold. The following range anxiety threshold calculation model is constructed: , In the above formula, Indicates the ratio of the basic mileage anxiety threshold to the total cruising range. Generally, when the remaining mileage is less than 20%, the driving mileage anxiety will increase significantly. Therefore, in this embodiment, =0.2. S Indicates the quantitative value of the driver's driving style; this embodiment determines that the user's driving type includes two types: conservative and aggressive. The former has a more conservative driving style, so it needs to be charged when the vehicle's mileage is high, so the range anxiety threshold is higher. On the contrary, drivers with a more aggressive driving style are more willing to drive the vehicle until the battery is almost exhausted before choosing to charge, and their corresponding range anxiety threshold is lower. Under this condition, this embodiment will S The value range is set to -1 to 1. Represents a conservative driver. Considering that driving style is not the only factor affecting the range anxiety threshold, this embodiment sets a parameter , which represents the influence coefficient of driving style on the range anxiety threshold; in this embodiment, Accordingly, this embodiment also designs a parameter , which represents the influence coefficient of charging station distribution density on range anxiety threshold; in this embodiment, . Z represents the distribution density of candidate charging stations; Z 0 represents the base charging station distribution density, Z andZ The difference between 0 and multiply The corresponding weights can just reflect the degree of impact of the differences in charging station distribution density in different scenarios on users' charging decisions.

[0045] It should be noted that, in this embodiment, S 、 and The parameters are not fixed, but are evaluated through each driver's historical charging selection behavior data. That is, the driver's charging preference characteristics are extracted from the collected historical charging behavior data, and each driver's driving style or driving experience factor, the degree of trust in driving experience, and the sensitivity to the distribution density of charging stations are obtained, so that the mileage anxiety threshold of different drivers in different driving scenarios can be calculated.

[0046] In actual applications, the remaining mileage of the electric vehicle, the selected charging station, charging time, charging cost, the location of the vehicle and each candidate charging station, the distribution density of the candidate charging stations, etc. corresponding to each historical charging behavior of the user can be collected and recorded as the historical driving behavior data of the current user. Then, based on the user's historical charging behavior data, a comprehensive analysis of the user's driving style, as well as the degree to which the driving style and the distribution density of charging stations affect the user's charging behavior decision, is conducted, and the update is then made. S 、 and The specific values ​​of the three parameters. For example, if a user's historical charging behavior data shows that he always charges when the battery is about to run out, it means that the user S The value of is closer to 1. When a user always charges when the remaining power is higher when the distribution density of surrounding charging stations is low, it means that the corresponding The value of is larger.

[0047] In summary, the range anxiety threshold constructed in this embodiment is L In fact, it accurately captures the user's behavior patterns when making charging decisions. Therefore, introducing this parameter into the user's charging behavior recommendations will make the results more consistent with the user's charging preferences.

[0048] 2. Provide recommendations based on acquired data and known user profiles 2.1. Actual collection of status information of the vehicle currently driven by the user and operational data of all charging stations.

[0049] In this embodiment, the essence of the charging station recommendation service is to match all available charging stations with the user's current vehicle, then select the most suitable charging stations for the vehicle and recommend them to the user. To match charging stations with users, this embodiment first determines the relevant attributes of the vehicle and charging station, namely the vehicle's status information and the charging station's operational data. The degree of match between the two is then quantified using specified evaluation metrics.

[0050] Specifically, in this embodiment, the vehicle status information includes: real-time location information, current mileage, total mileage, D , power consumption per kilometer e . And any candidate charging station i The operational data includes: location of charging stations, number of charging piles n i , congestion level β i , charging rate m i and charging power W i .

[0051] For example, the location information of the vehicle and the charging station determines the distance between the two, and users are usually more willing to choose a closer charging station for charging. e Determines the accessibility between each charging pile and the vehicle. The vehicle can only receive the corresponding charging service if it can reach the charging station. n i , congestion level β i and charging power W i etc. will affect the user experience at each charging station, and the charging rate m i This will affect the payment cost of the charging services enjoyed by users.

[0052] 2.2. When a user requests charging, the vehicle's mileage and location are combined to select all available charging stations around the vehicle as candidate charging stations, and form a candidate station set. I .

[0053] The technicians of this embodiment believe that the user's ability to drive the vehicle to the charging station independently is a prerequisite for the user to choose the charging station and accept the charging service. Any user who has other options will generally not accept the situation where the vehicle runs out of power before reaching the charging station and then needs to pay additional costs to have the vehicle towed to the charging station. Therefore, in order to reduce the amount of data calculations in the recommendation process, the solution provided in this embodiment will first calculate the accessibility of the current vehicle to each charging pile based on the vehicle's status information and the charging station's operating data each time the user makes a charging request, and will select each charging station that can be reached as a candidate charging station, and then select the most suitable charging station from them.

[0054] In this embodiment, the criterion for determining whether any charging station is reachable is that the vehicle's range is greater than the distance between the vehicle's real-time location and the charging station's location. Note: The distance here refers to the shortest route along a municipal road between the two locations, not the geographical straight-line distance.

[0055] 2.3. Evaluate the decision cost of the user selecting each candidate charging station for charging, and rank each charging station according to the selection probability.

[0056] In this embodiment, electric car drivers usually consider two types of attributes when choosing a charging station: decision attributes and solution attributes. The decision attribute is the remaining mileage, and the solution attribute is the charging cost. Among them, the broad charging cost should usually include multiple levels that cannot be accurately calculated. In order to more accurately evaluate the differences in charging costs corresponding to the selection of different charging stations. This embodiment divides the charging cost into three levels. The first is the charging fee paid by the user's charging behavior. The second is the time cost spent on completing the charging action, which mainly includes the user's charging time and the waiting time for the user to receive the charging service after arriving at the charging station. The third is the user experience brought to the user by the choice of different charging stations, which is mainly manifested in the difference in the degree of anxiety of the user when choosing different charging stations. Based on the above three points, this embodiment defines the three as payment costs. C b , parking time cost C t , mileage anxiety costs C h Among them, the payment cost of any charging behavior is C b , parking time cost C t and the cost of range anxiety C h The calculation method is as follows: (2.3.1) Payment costs C b Each time a user charges their vehicle, they usually choose to fully charge it before leaving the charging station. Therefore, this embodiment uses the total cruising range as D Arriving at a charging station with an electric car i Remaining mileage D i The difference, multiplied by the power consumption per kilometer e Calculate the total amount of electricity required to charge and multiply it by the charging station i Unit charging price m i Get a charging station i Payment cost C b The calculation formula is as follows: .

[0057] (2.3.2) Parking time cost C t In paying the cost C b On the basis of the above, considering that the driver is not only affected by the charging cost but also by the charging time when choosing a charging station, the parking time cost is introduced. C t . Parking time cost C t First, electric vehicles are included in the charging station i Actual charging time cost C r , while also considering the driver's psychological expected waiting time cost caused by the congestion of the charging stations provided by the system C l Specifically, this embodiment uses the total cruising range D Arriving at a charging station with an electric car i Remaining mileage D i The difference is multiplied by the power consumption per kilometer e , calculate the amount of electricity required to charge the electric vehicle. Then divide the amount of electricity by the charging station i Charging power W i To get the actual charging time, multiply the actual charging time by the driver's unit travel time value. TC , get the actual charging time cost C r Furthermore, this embodiment is based on the charging station i The degree of congestion β i Multiply charging stations i Number of charging piles n i Calculate the driver's charging stationi The number of vehicles expected to queue. Assume that the required charging mileage for these vehicles is the total range. D , multiply the number of vehicles in the queue by the charging mileage, and then multiply by the power consumption per kilometer e Get the total charging capacity of the vehicles in the queue. Divide the total charging capacity by the charging power W i , multiplied by the driver's unit travel time value TC , get the driver's psychological expected waiting time cost C l Therefore, the parking time cost C t The calculation formula is as follows: , In this embodiment, β i For charging stations displayed in the navigation system i The congestion level is displayed in accordance with the three congestion levels of the navigation system, namely "red, yellow and green". In this embodiment, the charging station congestion feedback is divided into three types: busy, medium and idle, corresponding to three congestion colors. Among them, the busy state usually means that the number of vehicles queuing at the charging station reaches 50% of the number of charging piles; the medium state usually means that the charging station i The number of vehicles in the queue reaches 20-50% of the number of charging piles; the idle state usually refers to the charging station i The number of vehicles in the queue is less than 20% of the number of charging piles. i When the state is classified as busy, let β i = 0.5. When the navigation system will charge the station i When the congestion level is medium, β i =0.35. When the navigation system will charge the station i When the state of is divided into idle, let β i =0.

[0058] In this embodiment, the driver's unit travel time value is taken into account. TC It is mainly affected by the user's income level and travel purpose. In the same driving scenario, different drivers have different personal income levels and their unit travel time value is TC For example, when the user's income level is higher, the corresponding unit travel time value is higher, and vice versa. In addition, in different driving scenarios, the unit travel time value of the same driver will change due to the change of travel purpose. TC For example, the unit time value of recreational travel will be lower than that of commuting travel, and much lower than that of emergency travel.

[0059] Considering the above two key factors, the unit travel time value is constructed TC The mathematical model of the driver’s daily income has been introduced above. R , daily working hours H Daily commuting time T w The method for calculating the user's hourly wage is based on this. This embodiment further combines the user's travel purpose to construct a travel urgency coefficient. , and use this coefficient to adjust the driver's unit travel time value TC Value:

[0060] In actual applications of the solution of this embodiment, the urgency coefficient for commuting trips may be set to 1, the urgency coefficient for entertainment trips may be set to 0.7, and the urgency coefficient for emergency trips may be set to 1.5.

[0061] (2.3.3) Range anxiety costs C h Every electric car driver has a range anxiety threshold when traveling L In order to describe the driver's range anxiety, this embodiment introduces a range anxiety cost C h The larger the value of this parameter is, the greater the driver's mileage anxiety is, and the more he wants to charge earlier to alleviate travel risks.

[0062] To quantify C h The technicians in this embodiment first thought of: when the remaining mileage of the electric vehicle x Greater than the range anxiety threshold L When the remaining mileage of the electric vehicle is x Less than L When, with x As the remaining mileage decreases, the driver's mileage anxiety increases rapidly. x When the driver's mileage anxiety reaches its peak value, the electric vehicle can no longer drive, resulting in a rescue towing fee. This fee is the maximum mileage anxiety cost when the driver's mileage anxiety reaches its peak value, calculated as C p In order to ensure that the maximum anxiety value reached when the remaining mileage is 0 remains unchanged, this embodiment chooses to use a power function to describe mileage anxiety, and then constructs a power function that characterizes the degree of change in the current user's mileage anxiety. F ( x ): , In the above formula, α is the preset power parameter.

[0063] On this basis, the technicians of this embodiment continue to think that: during the driving process, as the remaining mileage of the electric vehicle continues to decrease, the driver's mileage anxiety continues to accumulate and increase. L Reduce to D i The mileage anxiety function integral can be used to calculate the driver's cumulative anxiety value during the remaining mileage change process. Then multiply it by the unit mileage anxiety cost C n , we can calculate the driver's mileage anxiety cost.

[0064] So far, this embodiment obtains the mileage anxiety cost C h The calculation formula is as follows: , In the above formula, x Indicates the driving range of an electric vehicle; C n represents the unit mileage anxiety cost, which can be obtained by the maximum mileage anxiety cost C p Divide by the preset total range anxiety value to calculate. In actual application, the maximum range anxiety cost C p The corresponding towing costs can also be calculated in detail based on the road types in actual scenarios (urban roads, rural roads and highways).

[0065] (2.3.4) Charging Utility U i and selection probability P i The charging station recommendation problem belongs to a typical discrete choice model. Based on the random utility theory, this embodiment will i Total charging cost C i As a utility indicator, it is used to evaluate the probability of users choosing each charging station. A multinomial logit model is constructed to describe the charging choice behavior of electric vehicle drivers. Generally speaking, rational users usually consider the generalized total charging cost when choosing a charging station. C i , the cost will be paid C b , parking time cost C t , mileage anxiety costs C h All taken into consideration.

[0066] In addition, considering the cost of range anxiety C h Usually only exists when the user expects that the vehicle may not be able to reach the charging station. Therefore, this embodiment further calculates the total charging cost according to different situations. C i There are two ways to calculate: (1) When the electric vehicle arrives at the charging station i Remaining mileage D i Greater than the range anxiety threshold L When the charging station is selected, the total charging cost is C i Include only C b and C t (2)When the electric car arrives at the charging station i Remaining mileage D i Less than the range anxiety threshold L When the charging station is selected, the total charging cost is C i include C b 、 C t and C h . That is: , On this basis, the charging utility U i The calculation formula is as follows: , In the above formula, is the number of sites in the candidate site set; D i Indicates that the current user's vehicle has arrived at i The remaining range of each charging station; C i Indicates that the current user's vehicle is in i The total charging cost of each charging station.

[0067] On this basis, assuming that the utility random term follows the Gumbel distribution, in any driving scenario, there are two cases for the selection probability function: when the electric car reaches the i Remaining mileage at charging stations D iWhen it is greater than 0, the charging station has the possibility of being selected, and the multinomial logit model is used to calculate the selection probability. Otherwise, when the remaining mileage is D i When it is less than 0, the electric vehicle cannot reach the charging station i , then the probability of being selected is 0. Therefore, this embodiment uses the following formula according to U i Calculating selection probabilities P i : , In the above formula, Indicates the driver's perception of the charging utility.

[0068] Of course, since the previous text of this case has excluded the candidate charging stations when screening D i <0, therefore, in practical applications, this embodiment only calculates each candidate charging station ( ) is the probability of selection.

[0069] (2.3.5) Generation of recommendation lists In this embodiment, each candidate charging station can be P i Sort the list from high to low and return it to the user. This list can actually reflect the charging utility of each candidate charging pile. U i , and therefore is more in line with the preferences of users who require lower overall costs for charging behavior.

[0070] Considering the charging effect U i The calculation process also realizes the payment cost C b , parking time cost C t and the cost of range anxiety C h Therefore, when generating the recommendation list, this embodiment can also be based on C b 、 C t and C h The candidate charging stations are sorted from lowest to highest, resulting in three additional recommendation lists, which are then returned to the user. These three recommendation lists can meet the needs of users who prioritize economic cost, time cost, and user experience.

[0071] In the actual application of this embodiment, the electric vehicle driver can provide his or her location, destination, and travel purpose in advance through the driver terminal. The system then calculates the driver's probability of selecting each charging station based on the driver's charging preferences, recommends the charging station with the highest probability to the driver through the driver terminal, and provides a list of recommended charging stations. When the driver finally selects one of the charging stations based on the recommended list, a new historical charging behavior data can be generated to facilitate the system's perception of the current user's behavior pattern in selecting charging stations and the user's range anxiety threshold. L Update to improve the system's adaptability.

[0072] Example 2 Based on the solution in Example 1, this embodiment further provides a computer program product, which includes a computer program. When the computer program is executed by a processor, it implements the steps of the electric vehicle charging station recommendation method based on the driver's charging preferences as in Example 1, and then generates a recommended list of candidate charging stations that meet the user's preferences based on the user's current vehicle status and the operating data of the surrounding charging stations, and returns the recommended list to the user.

[0073] This embodiment also provides a storage medium storing a computer program. When the computer program is executed by a processor, the computer program implements the steps of the method for recommending electric vehicle charging stations based on the driver's charging preferences as in Example 1, and then generates a recommended list of candidate charging stations that meet the user's preferences based on the user's current vehicle status and the operating data of the surrounding charging stations, and returns the list to the user.

[0074] This embodiment also provides a data processing device, which includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the computer program is executed by the processor, it implements the steps of the electric vehicle charging station recommendation method based on the driver's charging preferences as in Example 1, and then generates a recommended list of candidate charging stations that meet the user's preferences based on the user's current vehicle status and the operating data of the surrounding charging stations, and returns the list to the user.

[0075] The data processing device is essentially a computer device. In practical applications, it can be an embedded device deployed within the electric vehicle's onboard system to support data processing and interaction. It can also be used as a standalone computer device, for example, integrated into a computer device in a backend data center connected to the onboard system, thereby implementing the aforementioned processes during interaction between the onboard system and the backend data center. This non-embedded computer device used in the backend data center can be a laptop, tablet computer, desktop computer, or a medium-to-large computer device such as a rack-mounted server, blade server, tower server, or cabinet server (including standalone servers or server clusters consisting of multiple servers) capable of executing computer programs.

[0076] Specifically, the computer device of this embodiment includes at least, but is not limited to, a memory and a processor that are communicatively connected to each other via a system bus. In this embodiment, the memory (i.e., readable storage medium) includes flash memory, a hard disk, a multimedia card, card-type memory (e.g., SD or DX memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic storage, magnetic disks, optical disks, and the like. In some embodiments, the memory may be an internal storage unit of the computer device, such as the computer device's hard disk or internal memory. In other embodiments, the memory may also be an external storage device of the computer device, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a flash memory card, and the like. Of course, the memory may also include both the internal storage unit and external storage devices of the computer device. In this embodiment, the memory is typically used to store the operating system and various application software installed on the computer device. Furthermore, the memory may also be used to temporarily store various types of data that has been output or is about to be output.

[0077] In some embodiments, the processor may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chip. The processor is generally used to control the overall operation of a computer device.

[0078] Figure 2Typical interactive interfaces that can be used in practical applications of the data processing device of this embodiment are provided. As shown in the figure, in Interface 1, users can request charging, enter their location (which can also be actively obtained by the device), destination, and travel purpose, and set their preferences for overall utility, payment fees, time cost, and user experience when selecting a charging station. In Interface 2, the system intuitively presents the current location, the destination, and the spatial distribution of candidate charging stations. In Interface 3, the system can display each charging station's location, charging price, congestion level, and charging power, based on user instructions. In Interface 4, the system can simultaneously display the ranking results of the best charging stations corresponding to different user preferences. In Interface 5, the system calculates the best charging station based on the user's specified preferences and displays the charging station's location, charging price, congestion level, and charging power in real time.

[0079] Performance Testing In order to verify the performance of the solution of the present invention, technicians developed an experimental plan and conducted simulation tests on relevant solutions.

[0080] 1. Solution Verification 1. Experimental scenario This experiment uses Figure 3 The road network condition is used as the research scenario. When the current remaining mileage is 100km, the driver can choose any of the four candidate charging stations 1-4 to charge the car to ensure the normal driving of the electric vehicle.

[0081] In this research scenario, the driver sends his / her location, destination, purpose of travel, and desired charging station characteristics to the backend through the terminal. The backend device then calculates the driver's probability of selecting each charging station according to the following steps. In this scenario, the initial settings of each parameter are: total vehicle range D =250km, initial remaining mileage D 0=100km, power consumption per kilometer e =0.15kWh / km (based on the national average power consumption per kilometer for electric vehicles), unit charging price of charging stations m = 0.8 yuan / kWh (according to the charging details of the State Grid charging station), rescue towing cost C p = 500 yuan (based on the national average traction cost of cars), charging power of the charging station W =60kW (based on the common power of electric vehicle charging piles), the driver's range anxiety threshold L =50km, driver's unit travel time value TC=38 yuan / h (based on the average salary level in a certain place), mileage anxiety function power parameter α=-2, driver perception level =0.8, charging station congestion β =0.

[0082] 2. Evaluation and Analysis This experiment uses the method in Example 1 to calculate the remaining mileage of the vehicle after it reaches four charging stations for charging. D i , payment costs C b , parking time cost C t , mileage anxiety costs C h Total charging cost C i , charging effect U i and selection probability P i They are shown in the following table: Table 1: Calculation results of various evaluation indicators for users’ selection of different charging stations

[0083] 3. Generation of recommendation lists Combining the data in the above table, this experiment shows that the charging station with the highest selection probability based on charging utility is charging station 3, the optimal charging station based on payment cost is charging station 3, the optimal charging station based on parking time cost is charging station 3, and the optimal charging station based on range anxiety cost is charging station 1 or 3.

[0084] 2. Sensitivity Analysis of Core Parameters in the Scheme This experiment studies the six key parameters that affect the method of the present invention in evaluating the probability of users choosing each charging station (including: unit travel time value TC , charging power W , congestion level β , Maximum Range Anxiety Cost C p , mileage anxiety threshold L and remaining mileage D i ) to conduct a sensitivity analysis. During the analysis, the remaining range, maximum range anxiety cost (also known as rescue towing cost), charging power, range anxiety threshold, unit travel time value, and charging station congestion were adjusted in turn to observe the impact of parameter changes on drivers' charging preferences. The results are as follows: Figure 4-Figure 9 shown.

[0085] Analyzing the data in the figure, we can find the following patterns: (1) The remaining mileage has the greatest impact. The less the remaining mileage, the more likely the driver is to charge earlier and the more likely they are to choose a nearby charging station. The more the remaining mileage, the less range anxiety the driver has and the more likely they are to delay charging.

[0086] (2) Range anxiety thresholds of different drivers ( L ) are different, conservative drivers ( L Higher) like to charge in time to avoid risks, aggressive drivers ( L Smaller) tends to charge to the limit when the power is low enough to reduce the number of charging times.

[0087] (3) Drivers' choices vary in different scenarios. The higher the maximum range anxiety cost, the earlier the driver will charge to reduce the risk. For example, on urban roads, towing costs are low, so drivers are less likely to experience range anxiety and generally delay charging. However, on highways, towing costs are extremely high, so drivers' range anxiety rises rapidly and they generally charge in advance.

[0088] (4) The higher the charging power of the charging station, the shorter the charging time. The time cost of rescue towing caused by exhaustion of power is much higher than the actual charging time cost, so drivers are more inclined to charge in advance to avoid risks.

[0089] (5) When the charging station is moderately crowded or busy, drivers tend to wait for the congestion to clear and thus delay charging.

[0090] (6) The value of unit travel time will make drivers tend to delay charging, but the degree is weak and there is almost no change.

[0091] Combining the influence mechanisms of all the above factors, it can be proved that in the solution provided by the present invention, electric vehicle drivers are more inclined to choose charging stations with higher charging power, lower congestion and closer distance when the remaining mileage drops to their own mileage anxiety threshold; and this also happens to be in line with the actual needs of users.

[0092] The above-described embodiment merely represents one embodiment of the present invention. While the description is relatively specific and detailed, it should not be construed as limiting the scope of the invention. It should be noted that a person skilled in the art would be able to make various modifications and improvements without departing from the spirit of the present invention, and these modifications and improvements fall within the scope of the present invention. Therefore, the scope of the present invention shall be determined by the appended claims.

Claims

1. A method for recommending electric vehicle charging stations based on driver charging preferences, characterized in that: It includes: Collect personal information representing the current user's income level and historical charging behavior data, and update the user's range anxiety threshold accordingly L ; Collect the status information of the vehicle currently driven by the user and the operating data of all charging stations; when the user makes a charging request, first combine the vehicle's mileage and vehicle location to select the charging stations that can be reached around the vehicle as candidate charging stations, and form a candidate station set I ; The following formula combines the collected information and the current user's mileage anxiety threshold L Calculate the time it takes for the user to reach any candidate charging station i Payment cost C b , parking time cost C t , mileage anxiety costs C h and charging utility U i : ， In the above formula, is the number of sites in the candidate site set; D i Indicates that the current user's vehicle has arrived at i The remaining range of each charging station; C i Indicates that the current user's vehicle is in i The total charging cost of each charging station; According to the following formula U i Calculating selection probabilities P i : ， In the above formula, Indicates the driver's perception of the utility of charging; According to user preferences, each candidate charging station is P i From high to low and by C b 、 C t and C h Sort the recommendations from low to high and return the generated recommendation list to the user.

2. The method for recommending electric vehicle charging stations based on driver charging preferences according to claim 1, characterized in that: The personal information includes: the driver's daily income R , daily working hours H Daily commuting time T w ; The historical charging behavior data includes: the remaining mileage of the electric vehicle corresponding to each historical charging behavior of the user, the selected charging station, the charging time, the charging fee, the location of the vehicle and each candidate charging station, and the distribution density of the candidate charging stations; The vehicle status information includes: real-time location information, current mileage, total mileage D , power consumption per kilometer e ; Any candidate charging station i Operational data includes: location of charging stations, number of charging piles n i , congestion level β i , charging rate m i and charging power W i .

3. The method for recommending electric vehicle charging stations based on driver charging preferences according to claim 2, characterized in that: Payment cost for any charging behavior C b The calculation formula is as follows: 。 4. The method for recommending electric vehicle charging stations based on driver charging preferences according to claim 3, characterized in that: The time cost of any charging behavior C t The calculation formula is as follows: ， In the above formula, C r Indicates that the vehicle is at a candidate charging station i Actual charging time cost; C l represents the psychologically expected queuing waiting time cost generated according to the congestion level of the candidate charging station; TC Indicates the unit travel time value of the current user; and / or The operating status of charging stations is divided into three types: busy, medium and idle, and the corresponding congestion levels are β i The values ​​of are 0.5, 0.35 and 0 respectively.

5. The method for recommending electric vehicle charging stations based on driver charging preferences according to claim 4, characterized in that: The range anxiety cost of arbitrary charging behavior C h The calculation formula is as follows: ， In the above formula, x Indicates the driving range of the electric vehicle; α is a preset power parameter; F ( x ) is a power function that represents the degree of change in the current user's mileage anxiety; C n represents the anxiety cost per unit mileage; and / or The unit mileage anxiety cost C n Cost of maximum range anxiety C p Divide by the preset total mileage anxiety value to get, where D i The towing fee generated after zeroing is used as the maximum mileage anxiety cost of the current trip C p .

6. The method for recommending electric vehicle charging stations based on driver charging preferences according to claim 4, characterized in that: The unit travel time value of the current user TC The calculation formula is as follows: ， In the above formula, Represents the travel urgency coefficient obtained based on the current user's travel purpose; and / or When the travel purposes are work commuting, entertainment and emergency tasks, the corresponding urgency coefficients The values ​​of are 1, 0.7, and 1.5 respectively.

7. The method for recommending electric vehicle charging stations based on driver charging preferences according to claim 2, characterized in that: Range anxiety threshold for any user L The update formula is as follows: ， In the above formula, Indicates the ratio of basic mileage anxiety threshold to total cruising range, =0.2; S Indicates the quantitative value of the driver's driving style; represents the influence coefficient of driving style on range anxiety threshold; The coefficient indicating the impact of charging station distribution density on range anxiety threshold; Z represents the distribution density of candidate charging stations; Z 0 represents the base charging station distribution density; and / or S 、 and The value of is evaluated based on the user's historical charging behavior data. Represents a conservative driver. represents an aggressive driver; ; .

8. A computer program product comprising a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method for recommending electric vehicle charging stations based on the driver's charging preferences as described in any one of claims 1 to 7 are implemented, and then a recommendation list of candidate charging stations that meet the user's preferences is generated based on the user's current vehicle status and the operating data of the surrounding charging stations, and the list is returned to the user.

9. A storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method for recommending electric vehicle charging stations based on the driver's charging preferences as described in any one of claims 1 to 7 are implemented, and then a recommendation list of candidate charging stations that meet the user's preferences is generated based on the user's current vehicle status and the operating data of the surrounding charging stations, and the list is returned to the user.

10. A data processing device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the computer program is executed by a processor, the steps of the method for recommending electric vehicle charging stations based on the driver's charging preferences as described in any one of claims 1 to 7 are implemented, and then a recommendation list of candidate charging stations that meet the user's preferences is generated based on the user's current vehicle status and the operating data of the surrounding charging stations, and the list is returned to the user.

Images