Method and device for identifying differentiated preferences of rapid charging station users

Abstract

The invention provides a method and device for identifying differentiated preferences of users of a quick charging station, and belongs to the field of electric vehicle load flexibility evaluation and the field of electric vehicle load regulation and control. The method comprises the following steps: collecting multi-source data related to quick charging station selection, including quick charging station data, quick charging user data and quick charging order data; based on the multi-source data, obtaining an alternative charging station set corresponding to each quick charging order and an attribute of each quick charging station in the alternative charging station set, wherein the attribute affects user selection; and based on the attributes, establishing an optimization model for user differentiation preference identification and solving the optimization model to obtain attribute preference weights of the users so as to complete identification of the user differentiation preferences of the rapid charging station. According to the method, through accurate identification of user differentiation preferences, a charging station operator can be helped to optimize charging load distribution in urban-level rapid charging station load flexibility management.

Description

Technical Field

[0001] The present invention belongs to the field of electric vehicle load flexibility assessment and electric vehicle load regulation, and specifically relates to a method and device for identifying differentiated preferences of users of fast charging stations. Background Art

[0002] In densely populated urban areas, fast charging stations (FCS) have emerged as key infrastructure to address the growing demand for charging and alleviate users' range anxiety. With the rapid expansion of fast charging infrastructure, the spatial flexibility of electric vehicles (EVs) has become a research focus. Spatial flexibility refers to the redistribution of EV charging load among different charging stations by influencing users' choice of charging stations. This flexibility provides important opportunities for the power system to enable EV users to participate in grid ancillary services such as mitigating congestion and absorbing renewable energy.

[0003] As private investment in charging infrastructure increases, charging station operators (CSOs) as decision makers of price incentives have become key players in exploiting the spatial flexibility of electric vehicles. Initially, price incentives were mainly set by power system operators. For example, studies have obtained the spatial demand elasticity of charging loads through traffic assignment problems, and then redistributed charging loads to reduce safety risks in urban power-traffic coupled networks. Today, charging prices include not only the electricity costs set by the power system, but also the service prices set by charging station operators. These operators play an important role in influencing users' choice of charging stations to exploit the spatial flexibility of charging loads.

[0004] In order to fully utilize the spatial flexibility of electric vehicles, it is crucial to identify users' differentiated preferences, as these preferences directly affect the spatial distribution of charging station loads. However, many existing studies simplify user preferences and assume that users are completely rational or have uniform preferences. For example, some studies guide users to the station with the shortest charging time based on the total charging time. However, user preferences in reality are uncertain and are affected by multiple conflicting attributes. Although some studies have considered differentiated preferences and bounded rationality, they often rely on theoretical assumptions rather than real data. For example, some studies have modeled user decision-making behavior through mixed Gaussian distribution parameters, and other studies have explored the impact of user preferences on route and charging decisions. Despite the progress of these studies, identifying user preferences through real-world data is still in its early stages.

[0005] In recent years, data-driven models have been increasingly used to identify user preferences, usually using vehicle trajectory data. For commercial electric vehicle fleets, trajectory data is more readily available and charging behavior can be inferred. For example, studies have explored the attributes that affect the choice of charging stations for large fleets of electric taxis based on trajectory data, and other studies have analyzed user preferences using multi-source real-world data, and there is also research based on cumulative prospect theory to model the random charging behavior of electric taxis. Despite these advances, it remains a challenge to identify different user preferences based on real data from charging station operators. Summary of the invention

[0006] The purpose of the present invention is to overcome the shortcomings of the prior art and propose a method and device for identifying the differentiated preferences of users of fast charging stations. The present invention can help charging station operators optimize charging load distribution in the city-level fast charging station load flexibility management by accurately identifying the differentiated preferences of users.

[0007] The first embodiment of the present invention provides a method for identifying differentiated preferences of users of fast charging stations, comprising:

[0008] Collect multi-source data related to fast charging station selection, including: fast charging station data, fast charging user data and fast charging order data;

[0009] Based on the multi-source data, a set of candidate charging stations corresponding to each fast charging order and an attribute of each fast charging station in the set of candidate charging stations that affects user selection are obtained;

[0010] Based on the attributes, an optimization model for identifying differentiated user preferences is established and solved to obtain the user's attribute preference weights, so as to complete the identification of differentiated preferences of fast charging station users.

[0011] In a specific embodiment of the present invention, obtaining the candidate charging station set corresponding to each fast charging order and the attributes of each fast charging station in the candidate charging station set that affect the user's selection includes:

[0012] 1) Based on the multi-source data, each fast charging order is marked to obtain a set of candidate charging stations corresponding to each fast charging order;

[0013] If the sth fast charging order of user n is displayed in scenario s, the user selects fast charging station i, then the fast charging station j within the preset distance δ around fast charging station i is used as the alternative charging station for the user in this scenario to form the alternative charging station set. The candidate charging station set includes a fast charging station i;

[0014] Remember a sjIndicates whether the fast charging station j is marked as an alternative charging station in scenario s. If the fast charging station j belongs to Then a sj is 1, otherwise it is 0; record y sj Indicates whether the fast charging station j is a historically selected charging station in scenario s, where if j = i, then y sj is 1, otherwise it is 0;

[0015] 2) Based on the result of step 1), the attributes that affect the user's choice of each fast charging station in the set of candidate charging stations corresponding to each fast charging order are counted;

[0016] Note x nsj is the attribute of fast charging station j in scenario s for user n, x nsjk is the kth attribute of fast charging station j in scenario s for user n, and the value range of k is 1 to 7, where x nsj1 to x nsj7 The average charging fee of fast charging station j in scenario s for user n is Fast charging pile power of fast charging station j The number of supporting facilities Q of fast charging station j j , the number of office-related points of interest (POIs) within the set range around the fast charging station The number of entertainment and shopping POIs within the set range around the fast charging station Number of medical and educational POIs within the set range around the fast charging station The number of residential POIs within the set range around the fast charging station

[0017] In a specific embodiment of the present invention, the optimization model expression of the user differentiated preference identification is as follows:

[0018]

[0019]

[0020] Wherein, formula (1) is the objective function of maximizing the optimization model of user differentiated preference identification, formula (2)-formula (4) are the constraints of the optimization model of user differentiated preference identification; L n is the maximum likelihood function of user n, i.e., the objective function; P nsj is the probability of user n choosing fast charging station j in scenario s; β n is the preference weight of user n; V nsj is the utility of fast charging station j in scenario s for user n; V nsjk is the utility of the kth attribute of fast charging station j in scenario s for user n, β nkis the preference weight of the kth attribute of user n.

[0021] In a specific embodiment of the present invention, before establishing the optimization model for identifying user differentiated preferences, the method further includes:

[0022] Filter the attributes, including:

[0023] By performing correlation analysis on each attribute, highly correlated attribute groups are obtained, and the attributes with the highest sum of correlations with other attributes in the group are retained in the group.

[0024] In a specific embodiment of the present invention, before establishing the optimization model for identifying user differentiated preferences, the method further includes:

[0025] The attribute is subjected to a piecewise linear transformation, comprising:

[0026] The k-th attribute set {x nsjk |n,s,j} is segmented using k-Means clustering. The number of segments is determined by the elbow method, and the k-th attribute is obtained. The starting point of the l-th segment is b kl , the starting point of the l+1th segment of the kth attribute is b k(l+1) ; The lth segment of the kth attribute of fast charging station j in user n scenario s is x nsjkl , β nkl is the preference weight of the lth segment of the kth attribute of user n;

[0027]

[0028] Using the piecewise linear conversion of formula (5) nsjk Replace equation (3) to obtain the updated optimization model for identifying user differentiated preferences.

[0029] A second embodiment of the present invention provides a device for identifying differentiated preferences of users of fast charging stations, comprising:

[0030] A data collection module is used to collect multi-source data related to the selection of fast charging stations, including: fast charging station data, fast charging user data and fast charging order data;

[0031] an attribute statistics module, configured to obtain, based on the multi-source data, a set of candidate charging stations corresponding to each fast charging order and an attribute of each fast charging station in the set of candidate charging stations that affects the user's selection;

[0032] The preference identification module is used to establish and solve an optimization model for user differentiated preference identification based on the attributes to obtain the user's attribute preference weights to complete the identification of differentiated preferences of users of fast charging stations.

[0033] In a specific embodiment of the present invention, obtaining the candidate charging station set corresponding to each fast charging order and the attributes of each fast charging station in the candidate charging station set that affect the user's selection includes:

[0034] 1) Based on the multi-source data, each fast charging order is marked to obtain a set of candidate charging stations corresponding to each fast charging order;

[0035] If the sth fast charging order of user n is displayed in scenario s, the user selects fast charging station i, then the fast charging station j within the preset distance δ around fast charging station i is used as the alternative charging station for the user in this scenario to form the alternative charging station set. The candidate charging station set includes a fast charging station i;

[0036] Remember a sj Indicates whether the fast charging station j is marked as an alternative charging station in scenario s. If the fast charging station j belongs to Then a sj is 1, otherwise it is 0; record y sj Indicates whether the marked fast charging station j is a historically selected fast charging station in scenario s, where if j = i, then y sj is 1, otherwise it is 0;

[0037] 2) Based on the result of step 1), the attributes that affect the user's choice of each fast charging station in the set of candidate charging stations corresponding to each fast charging order are counted;

[0038] Note x nsj is the attribute of fast charging station j in scenario s for user n, x nsjk is the kth attribute of fast charging station j in scenario s for user n, and the value range of k is 1 to 7, where x nsj1 to x nsj7 The average charging fee of fast charging station j in scenario s for user n is Fast charging pile power of fast charging station j The number of supporting facilities Q of fast charging station j j 、Number of office-related POIs within the set range around the fast charging station The number of entertainment and shopping POIs within the set range around the fast charging station Number of medical and educational POIs within the set range around the fast charging station The number of residential POIs within the set range around the fast charging station

[0039] In a specific embodiment of the present invention, the optimization model expression of the user differentiated preference identification is as follows:

[0040]

[0041]

[0042] Wherein, formula (1) is the objective function of maximizing the optimization model of user differentiated preference identification, formula (2)-formula (4) are the constraints of the optimization model of user differentiated preference identification; L n is the maximum likelihood function of user n, i.e., the objective function; P nsj is the probability of user n choosing fast charging station j in scenario s; β n is the preference weight of user n; V nsj is the utility of fast charging station j in scenario s for user n; V nsjk is the utility of the kth attribute of fast charging station j in scenario s for user n, β nk is the preference weight of the kth attribute of user n.

[0043] In a specific embodiment of the present invention, before establishing the optimization model for identifying user differentiated preferences, the method further includes:

[0044] Filter the attributes, including:

[0045] By performing correlation analysis on each attribute, highly correlated attribute groups are obtained, and the attributes with the highest sum of correlations with other attributes in the group are retained in the group.

[0046] In a specific embodiment of the present invention, before establishing the optimization model for identifying user differentiated preferences, the method further includes:

[0047] The attribute is subjected to a piecewise linear transformation, comprising:

[0048] The k-th attribute set {x nsjk |n,s,j} is segmented using k-Means clustering. The number of segments is determined by the elbow method, and the k-th attribute is obtained. The starting point of the l-th segment is b kl , the starting point of the l+1th segment of the kth attribute is b k(l+1) ; The lth segment of the kth attribute of fast charging station j in user n scenario s is x nsjkl , β nkl is the preference weight of the lth segment of the kth attribute of user n;

[0049]

[0050] Using the piecewise linear conversion of formula (5) nsjk Replace equation (3) to obtain the updated optimization model for identifying user differentiated preferences.

[0051] A third aspect of the present invention provides an electronic device, including:

[0052] at least one processor; and a memory communicatively coupled to the at least one processor;

[0053] The memory stores instructions executable by the at least one processor, and the instructions are configured to execute the above-mentioned method for identifying differentiated preferences of users of fast charging stations.

[0054] A fourth aspect of the present invention provides a computer-readable storage medium storing computer instructions for causing the computer to execute the above-mentioned method for identifying differentiated preferences of users of fast charging stations.

[0055] The advantages and beneficial effects of the present invention are:

[0056] The present invention aims to help charging station operators (CSOs) analyze user selection behaviors based on real-world data, thereby optimizing the load distribution and management of charging stations. The present invention uses a data-driven discrete choice model, combined with a variety of factors that affect user selection, to accurately identify user preferences, with a prediction accuracy of over 80%. Using the present invention, charging station operators can more effectively explore the spatial flexibility of electric vehicle loads through price incentives, thereby promoting the optimal allocation and management of power resources in scenarios such as grid congestion. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 The present invention is an overall flow chart of a method for identifying differentiated preferences of users of fast charging stations according to an embodiment of the present invention. DETAILED DESCRIPTION

[0058] The present invention proposes a method and device for identifying differentiated preferences of users of fast charging stations, which are further described in detail below with reference to the accompanying drawings and specific embodiments.

[0059] The first embodiment of the present invention provides a method for identifying differentiated preferences of users of fast charging stations, comprising:

[0060] Collect multi-source data related to fast charging station selection, including: fast charging station data, fast charging user data and fast charging order data;

[0061] Based on the multi-source data, a set of candidate charging stations corresponding to each fast charging order and an attribute of each fast charging station in the set of candidate charging stations that affects user selection are obtained;

[0062] Based on the attributes, an optimization model for identifying differentiated user preferences is established and solved to obtain the user's attribute preference weights, so as to complete the identification of differentiated preferences of fast charging station users.

[0063] In a specific embodiment of the present invention, the method for identifying the differentiated preferences of users of fast charging stations has an overall process as follows: Figure 1 As shown, the following steps are included:

[0064] 1) Collect multi-source data related to fast charging station selection.

[0065] In this embodiment, the multi-source data includes: fast charging station data, fast charging user data and fast charging order data.

[0066] The fast charging station data needs to include static information of each fast charging station (charging station type, charging power, surrounding points of interest (POI), supporting facilities) and dynamic charging prices during the selected period. In this embodiment, the charging station type is used to determine whether it is a public fast charging station rather than a dedicated charging station. The fast charging stations in this embodiment are all public fast charging stations, and the selected period lasts at least three months.

[0067] The fast charging user data is user data displayed in the fast charging station data that uses the fast charging pile of the fast charging station to charge. In this embodiment, the fast charging user data includes the type of user, which is divided into private car users and commercial vehicle users in this embodiment.

[0068] The fast charging order data is the fast charging order data corresponding to the fast charging station data.

[0069] 2) Based on the data collected in step 1), the charging scene is reproduced to obtain the set of candidate charging stations corresponding to each fast charging order and the attributes of each fast charging station in the set of candidate charging stations that affect the user's choice; the specific steps are as follows:

[0070] 2-1) Based on the data collected in step 1), each fast charging order is marked to obtain a set of candidate charging stations corresponding to each fast charging order.

[0071] Specifically, if the sth fast charging order of user n (i.e., in scenario s) shows that the user selects fast charging station i, then all fast charging stations j (including fast charging station i) within a preset distance δ around fast charging station i are regarded as the alternative charging stations of the user in this scenario to form an alternative charging station set. In this embodiment, it is recommended that the value of the preset distance δ be less than or equal to 3. In a specific implementation of the present invention, the value of δ is 3.

[0072] Remember a sjIndicates whether the fast charging station j is marked as an alternative charging station in scenario s. If the fast charging station j belongs to Then a sj is 1, otherwise it is 0. sj Indicates whether the fast charging station j is a historically selected charging station in scenario s, where if j = i, then y sj is 1, otherwise it is 0. This embodiment assumes that user n selects the sth fast charging order by comparing All the fast charging stations in the area finally chose the fast charging station i.

[0073] 2-2) Based on the result of step 2-1), the attributes that affect the user's choice of each fast charging station in the set of candidate charging stations corresponding to each fast charging order are counted.

[0074] In this embodiment, x nsj is the attribute of fast charging station j in scenario s for user n, x nsjk is the kth attribute of the fast charging station j in the user n scenario s. In this embodiment, the value range of k is 1 to 7, where x nsj1 to x nsj7 The average charging fee of fast charging station j in scenario s for user n is Fast charging pile power of fast charging station j The number of supporting facilities Q of fast charging station j j , the number of office-related POIs (points of interest) within the set range around the fast charging station j The number of entertainment and shopping POIs within the set range around the fast charging station Number of medical and educational POIs within the set range around the fast charging station The number of residential POIs within the set range around the fast charging station In a specific embodiment of the present invention, the set range is 1 km.

[0075] 3) Based on the result of step 2), an optimization model for identifying differentiated user preferences is established. The optimization model is solved to obtain the user's attribute preference weights, and the identification of differentiated user preferences for fast charging stations is completed.

[0076] In this embodiment, for any user n, the optimization model for identifying user differentiated preferences is expressed as follows:

[0077]

[0078]

[0079] Among them, formula (1) is the objective function of maximizing the optimization model of user differentiated preference identification, and formula (2)-formula (4) (4) are the constraints of the optimization model of user differentiated preference identification. n is the maximum likelihood function of user n, i.e., the objective function. Formula (2) represents the probability P of user n selecting fast charging station j in scenario s: nsj (β n ,x nsj ). β n is the attribute preference weight of user n. Formula (3) represents the utility V of fast charging station j in scenario s for user n: nsj (β n ,x nsj ). nsjk is the utility of the kth attribute of fast charging station j in scenario s for user n, β nk is the preference weight of the kth attribute of user n. Formula (4) is β nk range constraints.

[0080] Specifically, when estimating the preference weights, this embodiment needs to constrain the preference weights according to prior knowledge. The three terms in formula (4) represent non-negative, non-positive or unconstrained, respectively. For example, if other attributes remain unchanged and the average price rises, the selection utility decreases, that is, the preference weight of the average price should be a non-positive constraint. According to prior knowledge, the preference range of charging price is non-positive, the preference range of charging power and supporting facilities is non-negative, and the preference range related to POI is not restricted.

[0081] In this embodiment, solving the model can obtain the attribute preference weight β of user n: n ,Then according to Equations (2) and (3), the charging station users’ choice of charging station can be predicted, helping the charging station operators to optimize the charging load distribution.

[0082] Furthermore, before establishing the optimization model for identifying user differentiated preferences, the method of this embodiment may also include:

[0083] Filter the attributes obtained in step 2).

[0084] Specifically, attribute screening is a common way to avoid high collinearity of input attributes. If two input attributes are highly positively correlated, there are countless feasible solutions to solve the model to obtain the weight coefficient. In this case, the physical meaning of the weight coefficient reflecting the user's preference for the attribute is no longer reliable.

[0085] In this embodiment, the attribute screening method is as follows:

[0086] By performing correlation analysis on each attribute, multiple groups of highly correlated attributes can be obtained. In each group of highly correlated attributes, only the attribute with the highest sum of correlations with other attributes in the group needs to be retained.

[0087] Furthermore, before establishing the optimization model for identifying user differentiated preferences, the method of this embodiment may also include:

[0088] Perform piecewise linear transformation on the attributes obtained in step 2).

[0089] Specifically, this embodiment can perform piecewise linear conversion on the attribute. Taking charging power as an example, its utility V nsjk There may be changes in preference weights in different segments, rather than a linear change trend as shown in formula (3). From a physical point of view, the utility of charging power may increase rapidly in the middle and low segments of fast charging, but increase slowly in the super fast charging segment, or have an opposite increase trend for some users. For attributes with this trend, this embodiment uses formula (5) to convert V nsjk The replacement formula (3) can better reflect the utility trend.

[0090] In this embodiment, the kth attribute set {x nsjk |n,s,j} is segmented using k-Means clustering. The number of segments is determined by the elbow method. The k-th attribute can be obtained. The starting point of the l-th segment is b kl , the starting point of the l+1th segment of the kth attribute is b k(l+1) The lth segment of the kth attribute of fast charging station j under user n scenario s is x nsjkl , β nkl is the preference weight of the lth segment of the kth attribute of user n.

[0091]

[0092] Furthermore, the method described in this embodiment is further described in detail below in conjunction with a specific embodiment.

[0093] In a specific implementation of the present invention, the method for identifying differentiated preferences of users of fast charging stations comprises the following steps:

[0094] 1) Collect multi-source data related to fast charging station selection.

[0095] In this embodiment, the data comes from the actual operation data of the charging station operator Star Charge, including fast charging station data, fast charging user data and fast charging order data. The data collection scope is Taiyuan City, China. To identify user preferences, the specific data used include the following:

[0096] The fast charging station data includes 223 fast charging stations with fast charging piles in Taiyuan City, Shanxi Province. According to the Star Charging Operator Operating Rules, charging piles with a charging power greater than 20kW are considered fast charging piles. The data content includes static information of the station, such as charging station type, charging power, surrounding points of interest (POI), supporting facilities, etc., as well as dynamic charging prices from July 25, 2023 to October 31, 2023.

[0097] The fast charging user data includes information on 2,746 users in Taiyuan City, Shanxi Province who used fast charging piles to charge their cars. The user types are divided into private car users and commercial vehicle users (such as taxis and online car-hailing services).

[0098] The fast charging order data refers to 114,781 fast charging order data generated by the 2,746 fast charging users mentioned above at 223 fast charging stations in Taiyuan City, Shanxi Province between July 25, 2023 and October 31, 2023, including 25,805 orders for private cars and 88,976 orders for commercial vehicles.

[0099] 2) Based on the data collected in step 1), the charging scene is reproduced to obtain the set of candidate charging stations corresponding to each fast charging order and the attributes of each fast charging station in the set of candidate charging stations that affect the user's choice; the specific steps are as follows:

[0100] 2-1) Based on the data collected in step 1), each fast charging order is marked to obtain a set of alternative charging stations corresponding to each fast charging order, so as to obtain a set of alternative charging stations corresponding to each fast charging order and attributes of each fast charging station in the set of alternative charging stations that affect user selection.

[0101] In this embodiment, it can be known from the order data that the sth fast charging order of user n (i.e., in scenario s) shows that the user selects fast charging station i, then all fast charging stations j (including fast charging station i) within 3 km around fast charging station i are regarded as the alternative charging stations of the user in scenario s to form the alternative charging station set.

[0102] Remember a sj Indicates whether the fast charging station j is marked as an alternative charging station in scenario s. If the fast charging station j belongs to Then a sj is 1, otherwise it is 0. sj Indicates whether the marked fast charging station j is a historically selected fast charging station in scenario s, where if j = i, then y sj is 1, otherwise it is 0. This means that user n compares All the fast charging stations in the final selection of fast charging station i. In this embodiment, the data collected from step 1) can obtain 223 fast charging stations The number of fast charging stations ranges from a minimum of 1 to a maximum of 31, with an average of 14.

[0103] 2-2) Based on the result of step 2-1), the attributes that affect the user's choice of each fast charging station in the set of candidate charging stations corresponding to each fast charging order are counted.

[0104] In this embodiment, x nsj is the attribute of fast charging station j in scenario s for user n, x nsjk is the kth attribute of fast charging station j in scenario s for user n, and the value range of k is 1 to 7, where x nsj1 to x nsj7 The average charging fee of fast charging station j during the charging period of user n in scenario s is Fast charging pile power of fast charging station j The number of supporting facilities Q of fast charging station j j , the number of office-related POIs (points of interest) within 1 km of the fast charging station j Number of entertainment and shopping POIs within 1 km of fast charging station j Number of medical and educational POIs within 1km of fast charging station j Number of residential POIs within 1km of fast charging station j

[0105] 3) Based on the result of step 2), an optimization model for identifying differentiated user preferences is established. The optimization model is solved to obtain the user's attribute preference weights, and the identification of differentiated user preferences for fast charging stations is completed.

[0106] In this embodiment, for any user n, the optimization model for identifying user differentiated preferences is expressed as follows:

[0107]

[0108]

[0109] Among them, formula (1) is the objective function of maximizing the optimization model of user differentiated preference identification, and formula (2)-formula (4) (4) are the constraints of the optimization model of user differentiated preference identification. n is the maximum likelihood function of user n, i.e., the objective function. Formula (2) represents the probability P of user n selecting fast charging station j in scenario s: nsj (β n ,x nsj ). βn is the attribute preference weight of user n. Formula (3) represents the utility V of fast charging station j in scenario s for user n: nsj (β n ,x nsj ). nsjk is the utility of the kth attribute of fast charging station j in scenario s for user n, β nk is the preference weight of the kth attribute of user n. Formula (4) is β nk range constraints.

[0110] Specifically, when estimating preference weights, this embodiment needs to constrain the preference weights based on prior knowledge. The three terms in formula (4) represent non-negative, non-positive, or unconstrained, respectively. For example, if other attributes remain unchanged and the average price rises, the choice utility decreases, that is, the preference weight of the average price should be a non-positive constraint. According to prior knowledge, the preference range of charging price is non-positive, the preference range of charging power and supporting facilities is non-negative, and the preference range related to POI is not restricted. The charging station operator obtains the preference weight β of each attribute of user n n Finally, according to equations (2) and (3), the charging station selection of charging station users can be predicted, helping charging station operators to optimize charging load distribution.

[0111] Furthermore, this embodiment can further improve the accuracy of preference identification through attribute screening and attribute conversion; the details are as follows:

[0112] In this embodiment, through correlation analysis, the attributes with high collinearity in this embodiment are concentrated in four types of POI attributes, namely The correlations with the other three POIs are 0.92, 0.91, and 0.89 respectively. It is more correlated with the other three POIs and can better reflect the changes of the other three POIs. The POIs are retained as input influencing attributes, and the other three POIs are eliminated. After screening, the number of input influencing attributes is determined to be 4.

[0113] In this embodiment, the fast charging pile power of fast charging station j is The number of supporting facilities Q of fast charging station j j , entertainment and shopping POI number A piecewise linear transformation was performed.

[0114] In this example, the number of private car users is 1411, the number of commercial vehicle users is 1335, the number of private car orders is 25805, and the number of commercial vehicle orders is 88976. By comparing the following four cases, this example shows the importance of different user preferences and attribute conversions in improving prediction accuracy.

[0115] Solution 1: All users randomly select from the set of alternative solutions, without model.

[0116] Scheme 2: All users have the same preference weights, and all attributes use the linear utility of the basic model.

[0117] Scheme 3: All users have differentiated preference weights, and all attributes use the linear utility of the basic model.

[0118] Solution 4: All users have differentiated preference weights, and all attributes use converted attributes (i.e., the method described in this embodiment).

[0119] In this example, 70% of each user's data is used as a training set and 30% is used as a test set. The prediction accuracy ACC of user n on the test set is n And top3 accuracy is regarded as the model performance indicator for user n. β obtained using the training set of user n n Calculate all P on the user n test set nsj (β n ,x nsj ). For scenario s of user n, if P nsj (β n ,x nsj ) The largest fast charging station is consistent with the fast charging station actually selected by the user, then the scenario prediction is correct. ACC n reflects the ratio of the correctly predicted scenarios in the test set of user n to the total number of scenarios in the test set. For scenario s of user n, if P nsj (β n ,x nsj ) contains the fast charging station actually selected by the user, then it is considered that the scenario prediction is correct under the condition of top3. It reflects the ratio of the correctly predicted scenarios under the top 3 conditions in the user n test set to the total number of test set scenarios.

[0120] Prediction accuracy ACC and top3 accuracy ACC on all user test sets 3 The overall model performance indicator is considered as n and ACC n 3 The weight is positively correlated with the number of scenarios in the user n test set, that is, the more charging records a user has, the greater the weight.

[0121] For the ACC index, the prediction accuracy rates of private car users in the four schemes of this embodiment are 11.6%, 27.5%, 61.1%, and 68.5% respectively, and the prediction accuracy rates of commercial vehicle users are 9.9%, 19.4%, 61.6%, and 71.7% respectively.

[0122] For ACC considering uncertainty 3 Indicators, the prediction accuracy rates of private car users in the four schemes are 27.6%, 48.3%, 76.8% and 81.9% respectively, and the prediction accuracy rates of commercial vehicle users are 25.1%, 38.1%, 79.8% and 86.6% respectively.

[0123] It can be found that users have different preference weights for both private cars and commercial vehicles. In addition, attribute conversion based on prior knowledge and attribute distribution helps to improve the performance index of the model. Overall, the ACC index of the method described in this embodiment is around 70%, and the ACC index adopted after considering the uncertainty of user selection is 3 The indicator remained above 80%, with the overall indicator for private car users at 81.9% and the overall indicator for commercial vehicle users at 86.6%.

[0124] Furthermore, according to the preference weights of each user obtained by the method described in this embodiment, it can be obtained:

[0125] Among the average charging fee preference weights during charging, the proportion of users with high and medium preference weights for commercial vehicles is 45.3%, which is much higher than the 18.6% for private cars. The preference weight of this attribute varies greatly between the two types of users. From the perspective of tapping the load flexibility of fast charging stations, commercial vehicles have greater flexibility potential.

[0126] The power preference weights of fast charging piles between private car users and commercial vehicle users are quite different. The proportion of commercial vehicle users with low and high preferences is greater than that of private cars, among which low preference users account for 61.2%, while the proportion of users with medium preference is lower than that of private cars. This shows that commercial vehicle users mainly prefer fast charging stations with low charging power, but some users also like high charging power fast charging stations and their proportion exceeds that of private car users. The proportion of private car users who prefer low and medium charging power fast charging stations is very close, and the sum of the two proportions is very large, reaching 88.2%.

[0127] Regarding the preference weight for the number of supporting facilities, the preference weight for private cars is slightly higher than that for commercial vehicles, indicating that fast charging stations with better supporting facilities are slightly more attractive to private car users overall.

[0128] As for the preference weights of entertainment and shopping POIs, since the difference from the aforementioned attributes is that the preference signs are unlimited, both the positive and negative parts are divided into high, medium and low preferences. For all users as a whole, the increase in the number of such POIs will help increase the probability of users choosing fast charging stations. Specifically for the two types of users, private car users have slightly more positive and negative high preferences than commercial vehicles, and private car users have more medium-high preferences and high-no preferences for POI entertainment fast charging stations.

[0129] This example uses the existing data of CSO operators (FCS data, fast charging user data, and fast charging order data) and a data-driven discrete choice model to identify differentiated user preferences. These conclusions lay a solid foundation for CSOs to effectively utilize the spatial flexibility of electric vehicles through price incentives.

[0130] To implement the above embodiment, a second aspect of the present invention provides a device for identifying differentiated preferences of users of fast charging stations, including:

[0131] A data collection module is used to collect multi-source data related to the selection of fast charging stations, including: fast charging station data, fast charging user data and fast charging order data;

[0132] an attribute statistics module, configured to obtain, based on the multi-source data, a set of candidate charging stations corresponding to each fast charging order and an attribute of each fast charging station in the set of candidate charging stations that affects the user's selection;

[0133] The preference identification module is used to establish and solve an optimization model for user differentiated preference identification based on the attributes to obtain the user's attribute preference weights to complete the identification of differentiated preferences of users of fast charging stations.

[0134] In a specific embodiment of the present invention, obtaining the candidate charging station set corresponding to each fast charging order and the attributes of each fast charging station in the candidate charging station set that affect the user's selection includes:

[0135] 1) Based on the multi-source data, each fast charging order is marked to obtain a set of candidate charging stations corresponding to each fast charging order;

[0136] If the sth fast charging order of user n is displayed in scenario s, the user selects fast charging station i, then the fast charging station j within the preset distance δ around fast charging station i is used as the alternative charging station for the user in this scenario to form the alternative charging station set. The candidate charging station set includes a fast charging station i;

[0137] Remember a sj Indicates whether the fast charging station j is marked as an alternative charging station in scenario s. If the charging station j belongs to Then a sj is 1, otherwise it is 0; record y sj Indicates whether the marked fast charging station j is a historically selected fast charging station in scenario s, where if j = i, then y sj is 1, otherwise it is 0;

[0138] 2) Based on the result of step 1), the attributes that affect the user's choice of each fast charging station in the set of candidate charging stations corresponding to each fast charging order are counted;

[0139] Note x nsj is the attribute of fast charging station j in scenario s for user n, x nsjk is the kth attribute of fast charging station j in scenario s for user n, and the value range of k is 1 to 7, where x nsj1 to x nsj7 The average charging fee of fast charging station j in scenario s for user n is Fast charging pile power of fast charging station j The number of supporting facilities Q of fast charging station j j 、Number of office-related POIs within the set range around the fast charging station The number of entertainment and shopping POIs within the set range around the fast charging station Number of medical and educational POIs within the set range around the fast charging station The number of residential POIs within the set range around the fast charging station j

[0140] In a specific embodiment of the present invention, the optimization model expression of the user differentiated preference identification is as follows:

[0141]

[0142]

[0143]

[0144] Wherein, formula (1) is the objective function of maximizing the optimization model of user differentiated preference identification, formula (2)-formula (4) are the constraints of the optimization model of user differentiated preference identification; L n is the maximum likelihood function of user n, i.e., the objective function; P nsj is the probability of user n choosing fast charging station j in scenario s; β n is the preference weight of user n; V nsj is the utility of fast charging station j in scenario s for user n; V nsjk is the utility of the kth attribute of fast charging station j in scenario s for user n, β nk is the preference weight of the kth attribute of user n.

[0145] In a specific embodiment of the present invention, before establishing the optimization model for identifying user differentiated preferences, the method further includes:

[0146] Filter the attributes, including:

[0147] By performing correlation analysis on each attribute, highly correlated attribute groups are obtained, and the attributes with the highest sum of correlations with other attributes in the group are retained in the group.

[0148] In a specific embodiment of the present invention, before establishing the optimization model for identifying user differentiated preferences, the method further includes:

[0149] The attribute is subjected to a piecewise linear transformation, comprising:

[0150] The k-th attribute set {x nsjk |n,s,j} is segmented using k-Means clustering. The number of segments is determined by the elbow method, and the k-th attribute is obtained. The starting point of the l-th segment is b kl , the starting point of the l+1th segment of the kth attribute is b k(l+1) ; The lth segment of the kth attribute of fast charging station j in user n scenario s is x nsjkl , β nkl is the preference weight of the lth segment of the kth attribute of user n;

[0151]

[0152] Using the piecewise linear conversion of formula (5) nsjk Replace equation (3) to obtain the updated optimization model for identifying user differentiated preferences.

[0153] To implement the above embodiment, a third aspect of the present invention provides an electronic device, including:

[0154] at least one processor; and a memory communicatively coupled to the at least one processor;

[0155] The memory stores instructions executable by the at least one processor, and the instructions are configured to execute the above-mentioned method for identifying differentiated preferences of users of fast charging stations.

[0156] To implement the above embodiment, a fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the above method for identifying differentiated preferences of users of fast charging stations.

[0157] It should be noted that the computer-readable medium disclosed above may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, device or device. In the present disclosure, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, in which a computer-readable program code is carried. This propagated data signal may take a variety of forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination of the above. The computer readable signal medium may also be any computer readable medium other than a computer readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0158] The computer-readable medium may be included in the electronic device, or may exist independently without being installed in the electronic device. The computer-readable medium carries one or more programs. When the one or more programs are executed by the electronic device, the electronic device executes a method for identifying differentiated preferences of users of fast charging stations according to the above embodiment.

[0159] Computer program code for performing the operations of the present disclosure may be written in one or more programming languages, or a combination thereof, including object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0160] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0161] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0162] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code that includes one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present application belong.

[0163] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute the instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purpose of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or otherwise processing in a suitable manner if necessary, and then stored in a computer memory.

[0164] It should be understood that the various parts of the present application can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0165] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0166] In addition, each functional unit in each embodiment of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0167] The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present application. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A method for identifying differentiated preferences of users of fast charging stations, characterized in that: include: Collect multi-source data related to fast charging station selection, including: fast charging station data, fast charging user data and fast charging order data; Based on the multi-source data, a set of candidate charging stations corresponding to each fast charging order and an attribute of each fast charging station in the set of candidate charging stations that affects user selection are obtained; Based on the attributes, an optimization model for identifying differentiated user preferences is established and solved to obtain the user's attribute preference weights, so as to complete the identification of differentiated preferences of fast charging station users.

2. The method according to claim 1, characterized in that The step of obtaining a set of candidate charging stations corresponding to each fast charging order and the attributes of each fast charging station in the set of candidate charging stations that affect the user's selection includes: 1) Based on the multi-source data, each fast charging order is marked to obtain a set of candidate charging stations corresponding to each fast charging order; If the sth fast charging order of user n is displayed in scenario s, the user selects fast charging station i, then the fast charging station j within the preset distance δ around fast charging station i is used as the alternative charging station for the user in this scenario to form the alternative charging station set. The candidate charging station set includes a fast charging station i; Remember a sj Indicates whether the fast charging station j is marked as an alternative charging station in scenario s. If the fast charging station j belongs to Then a sj is 1, otherwise it is 0; record y sj Indicates whether the fast charging station j is a historically selected charging station in scenario s, where if j = i, then y sj is 1, otherwise it is 0; 2) Based on the result of step 1), the attributes that affect the user's choice of each fast charging station in the set of candidate charging stations corresponding to each fast charging order are counted; Note x nsj is the attribute of fast charging station j in scenario s for user n, x nsjk is the kth attribute of fast charging station j in scenario s for user n, and the value range of k is 1 to 7, where x nsj1 to x nsj7 The average charging fee of fast charging station j in scenario s for user n is Fast charging pile power of fast charging station j The number of supporting facilities Q of fast charging station j j , the number of office-related points of interest (POIs) within the set range around the fast charging station The number of entertainment and shopping POIs within the set range around the fast charging station Number of medical and educational POIs within the set range around the fast charging station The number of residential POIs within the set range around the fast charging station 3. The method according to claim 2, characterized in that The optimization model expression of the user differentiated preference identification is as follows: Wherein, formula (1) is the objective function of maximizing the optimization model of user differentiated preference identification, formula (2)-formula (4) are the constraints of the optimization model of user differentiated preference identification; L n is the maximum likelihood function of user n, i.e., the objective function; P nsj is the probability of user n choosing fast charging station j in scenario s; β n is the preference weight of user n; V nsj is the utility of fast charging station j in scenario s for user n; V nsjk is the utility of the kth attribute of fast charging station j in scenario s for user n, β nk is the preference weight of the kth attribute of user n.

4. The method according to claim 3, characterized in that Before establishing the optimization model for identifying user differentiated preferences, the method further includes: Filter the attributes, including: By performing correlation analysis on each attribute, highly correlated attribute groups are obtained, and the attributes with the highest sum of correlations with other attributes in the group are retained in the group.

5. The method according to claim 3, characterized in that: Before establishing the optimization model for identifying user differentiated preferences, the method further includes: The attribute is subjected to a piecewise linear transformation, comprising: The k-th attribute set {x nsjk |n,s,j} is segmented using k-Means clustering. The number of segments is determined by the elbow method, and the k-th attribute is obtained. The starting point of the l-th segment is b kl , the starting point of the l+1th segment of the kth attribute is b k(l+1) ; The lth segment of the kth attribute of fast charging station j in user n scenario s is x nsjkl , β nkl is the preference weight of the lth segment of the kth attribute of user n; Using the piecewise linear conversion of formula (5) nsjk Replace equation (3) to obtain the updated optimization model for identifying user differentiated preferences.

6. A device for identifying differentiated preferences of users of fast charging stations, characterized in that: include: A data collection module is used to collect multi-source data related to the selection of fast charging stations, including: fast charging station data, fast charging user data and fast charging order data; an attribute statistics module, configured to obtain, based on the multi-source data, a set of candidate charging stations corresponding to each fast charging order and an attribute of each fast charging station in the set of candidate charging stations that affects the user's selection; The preference identification module is used to establish and solve an optimization model for user differentiated preference identification based on the attributes to obtain the user's attribute preference weights to complete the identification of differentiated preferences of users of fast charging stations.

7. The device according to claim 6, characterized in that The step of obtaining a set of candidate charging stations corresponding to each fast charging order and the attributes of each fast charging station in the set of candidate charging stations that affect the user's selection includes: 1) Based on the multi-source data, each fast charging order is marked to obtain a set of candidate charging stations corresponding to each fast charging order; If the sth fast charging order of user n is displayed in scenario s, the user selects fast charging station i, then the fast charging station j within the preset distance δ around fast charging station i is used as the alternative charging station for the user in this scenario to form the alternative charging station set. The candidate charging station set includes a fast charging station i; Remember a sj Indicates whether the fast charging station j is marked as an alternative charging station in scenario s. If the fast charging station j belongs to Then a sj is 1, otherwise it is 0; record y sj Indicates whether the marked fast charging station j is a historically selected fast charging station in scenario s, where if j = i, then y sj is 1, otherwise it is 0; 2) Based on the result of step 1), the attributes that affect the user's choice of each fast charging station in the set of candidate charging stations corresponding to each fast charging order are counted; Note x nsj is the attribute of fast charging station j in scenario s for user n, x nsjk is the kth attribute of fast charging station j in scenario s for user n, and the value range of k is 1 to 7, where x nsj1 to x nsj7 The average charging fee of fast charging station j in scenario s for user n is Fast charging pile power of fast charging station j The number of supporting facilities Q of fast charging station j j 、Number of office-related POIs within the set range around the fast charging station The number of entertainment and shopping POIs within the set range around the fast charging station Number of medical and educational POIs within the set range around the fast charging station The number of residential POIs within the set range around the fast charging station 8. The device according to claim 7, characterized in that The optimization model expression of the user differentiated preference identification is as follows: Wherein, formula (1) is the objective function of maximizing the optimization model of user differentiated preference identification, formula (2)-formula (4) are the constraints of the optimization model of user differentiated preference identification; L n is the maximum likelihood function of user n, i.e., the objective function; P nsj is the probability of user n choosing fast charging station j in scenario s; β n is the preference weight of user n; V nsj is the utility of fast charging station j in scenario s for user n; V nsjk is the utility of the kth attribute of fast charging station j in scenario s for user n, β nk is the preference weight of the kth attribute of user n.

9. The device according to claim 8, characterized in that Before establishing the optimization model for identifying user differentiated preferences, the method further includes: Filter the attributes, including: By performing correlation analysis on each attribute, highly correlated attribute groups are obtained, and the attributes with the highest sum of correlations with other attributes in the group are retained in the group.

10. The device according to claim 8, characterized in that Before establishing the optimization model for identifying user differentiated preferences, the method further includes: The attribute is subjected to a piecewise linear transformation, comprising: The k-th attribute set {x nsjk |n,s,j} is segmented using k-Means clustering. The number of segments is determined by the elbow method, and the k-th attribute is obtained. The starting point of the l-th segment is b kl , the starting point of the l+1th segment of the kth attribute is b k(l+1) ; The lth segment of the kth attribute of fast charging station j in user n scenario s is x nsjkl , β nkl is the preference weight of the lth segment of the kth attribute of user n; Using the piecewise linear conversion of formula (5) nsjk Replace equation (3) to obtain the updated optimization model for identifying user differentiated preferences.

Images