Energy replenishment resource recommendation method, electronic device, and storage medium

Abstract

The present invention relates to the technical field of electric vehicles, and specifically provides an energy replenishment resource recommendation method, an electronic device, and a storage medium. The present invention aims to solve the problem of inaccurate recommendation results due to most of existing energy replenishment resource recommendation methods being limited to a single station dimension. To achieve this objective, the method of the present invention comprises: obtaining user information and resource information, the resource information comprising information of a plurality of candidate energy replenishment stations in an energy replenishment resource list; inputting the information of the plurality of candidate energy replenishment stations into a trained resource fusion model to obtain comparison information between the plurality of candidate energy replenishment stations, and obtaining fusion attribute information of each candidate energy replenishment station on the basis of the comparison information; and inputting the user information and the fusion attribute information of each candidate energy replenishment station into a trained energy replenishment resource recommendation model to obtain an energy replenishment resource recommendation result. According to the described embodiment, the information of the plurality of candidate energy supplementing stations in the list can be fused, to provide more accurate and comprehensive results, and recommend the best resources matching users.

Description

Energy replenishment resource recommendation method, electronic device and storage medium

[0001] This application claims priority to Chinese patent application No. 202311028951.5 filed on August 14, 2023, with the invention name “Energy Supplementary Resource Recommendation Method, Electronic Device and Storage Medium”. The entire contents of the above Chinese patent application are incorporated into this application by reference. Technical Field

[0002] The present invention relates to the technical field of electric vehicles, and in particular to an energy replenishment resource recommendation method, electronic equipment, and storage medium. Background Art

[0003] As the number of electric vehicles continues to grow, the availability of various charging resources for electric vehicles is becoming increasingly dense. Traditional EV charging resource recommendation methods are mostly limited to a single site dimension. Specifically, they rank candidate resources based on the degree to which resource attributes match user preferences, with higher-ranked resources indicating a better match for the user. These methods typically only rank individual resources in a list, ignoring the correlations and differences between different resources. This can lead to inaccurate and one-sided ranking results, failing to fully reflect the relative importance of resources in the list.

[0004] In resource recommendation scenarios, user choices are highly exclusive. After placing an order for one recommended item, users generally won't order any other resources on the list. Therefore, accurate resource recommendations are crucial. Simply measuring the degree of match between resource attributes and user preferences cannot accurately determine whether a candidate resource is optimal. Often, after the match exceeds a certain threshold, resource attributes are mostly consistent with the user's needs, with only a small difference. Therefore, selecting the optimal resource based on a single resource dimension is infeasible.

[0005] Accordingly, a new energy resource recommendation method is needed in this field to solve the above problems.

[0006] Summary of the Invention

[0007] In order to overcome the above-mentioned defects, the present invention is proposed to provide a method, electronic device and storage medium for recommending energy resources, which solves or at least partially solves the technical problem that most existing energy resource recommendation methods are limited to a single site dimension, resulting in inaccurate recommendation results.

[0008] In a first aspect, a method for recommending energy replenishment resources is provided, the method comprising:

[0009] Acquiring user information and resource information, wherein the resource information includes information of multiple candidate energy replenishment sites in the energy replenishment resource list;

[0010] Inputting information of the plurality of candidate energy replenishment stations into a trained resource fusion model, obtaining comparison information between the plurality of candidate energy replenishment stations, and obtaining fusion attribute information of each candidate energy replenishment station based on the comparison information;

[0011] The user information and the fused attribute information of each candidate energy replenishment site are input into a trained energy replenishment resource recommendation model to obtain an energy replenishment resource recommendation result.

[0012] In one technical solution of the above-mentioned energy replenishment resource recommendation method, obtaining user information includes:

[0013] Obtain at least one of user profile information, context information, and target object information to be replenished;

[0014] The context information includes at least one of location information, time information, scene information, and weather information.

[0015] In one technical solution of the above-mentioned energy replenishment resource recommendation method, obtaining resource information includes:

[0016] Obtain at least one of the following: site portrait information, battery portrait information, and charging pile portrait information of the plurality of candidate energy replenishment sites in the energy replenishment resource list;

[0017] The station portrait information includes the station ID and station attributes.

[0018] In one technical solution of the above-mentioned method for recommending energy replenishment resources, the station attribute includes at least one of distance, price, and energy replenishment speed; the comparison information includes association information and difference information between the multiple candidate energy replenishment stations; and obtaining fused attribute information of each candidate energy replenishment station based on the comparison information includes:

[0019] Obtaining a weight of the station attribute of each candidate refueling station based on the importance of the station attribute and the comparison information;

[0020] The station attributes are weighted based on the weights to obtain the fusion attribute information of each candidate energy replenishment station.

[0021] In one technical solution of the above energy replenishment resource recommendation method, the method further includes training the resource fusion model based on at least the following steps:

[0022] Obtaining a first training set, where the first training set is energy replenishment stations marked with fusion attribute information labels;

[0023] Training the resource fusion model based on the first training set;

[0024] When the resource fusion model converges to a first preset error value, the resource fusion model training is completed.

[0025] In one technical solution of the above-mentioned energy replenishment resource recommendation method, the energy replenishment resource recommendation model includes a prediction module and / or a ranking module; the method further includes training the energy replenishment resource recommendation model based on at least the following steps:

[0026] Obtaining historical user information and historical resource information from at least a portion of historical orders as a second training set;

[0027] Training the prediction module and / or the sorting module of the energy replenishment resource recommendation model based on the second training set;

[0028] When the energy replenishment resource recommendation model converges to a second preset error value, the energy replenishment site recommendation model training is completed.

[0029] In one technical solution of the above-mentioned energy replenishment resource recommendation method, the training of the prediction module of the energy replenishment resource recommendation model based on the second training set includes:

[0030] Based on the prediction module, the historical candidate refueling sites in the second training set are predicted to obtain the conversion probability of each of the historical candidate refueling sites.

[0031] In a technical solution of the above energy replenishment resource recommendation method, the method further includes:

[0032] Based on the ranking module and the conversion probability of each of the historical candidate energy replenishment sites, all the historical candidate energy replenishment sites are ranked to obtain the energy replenishment resource recommendation result.

[0033] In a second aspect, an electronic device is provided, which includes a processor and a storage device, wherein the storage device is suitable for storing multiple program codes, and the program codes are suitable for being loaded and run by the processor to execute the energy replenishment resource recommendation method described in any one of the technical solutions of the above-mentioned energy replenishment resource recommendation method.

[0034] In a third aspect, a computer-readable storage medium is provided, which stores a plurality of program codes, wherein the program codes are suitable for being loaded and run by a processor to execute the energy replenishment resource recommendation method described in any one of the technical solutions of the above-mentioned energy replenishment resource recommendation method.

[0035] The above one or more technical solutions of the present invention have at least one or more of the following beneficial effects:

[0036] In implementing the technical solution of the present invention, user information and resource information are first obtained, wherein the resource information includes information of multiple candidate energy replenishment stations in the energy replenishment resource list; then, the information of the multiple candidate energy replenishment stations is input into a trained resource fusion model to obtain comparison information between the multiple candidate energy replenishment stations, and based on the comparison information, the fusion attribute information of each candidate energy replenishment station is obtained; finally, the user information and the fusion attribute information of each candidate energy replenishment station are input into a trained energy replenishment resource recommendation model to obtain the energy replenishment resource recommendation result. Through the above implementation, the information of multiple candidate energy replenishment stations in the list can be integrated, and the relevance and difference between multiple resources can be comprehensively considered. It can provide more accurate and comprehensive results based on the user information and the fusion attribute information of each candidate energy replenishment station, and recommend the best resources that match the user, thereby improving the user's energy replenishment experience and satisfaction, and increasing user stickiness. This solution has broad application prospects and commercial value. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The disclosure of the present invention will become more easily understood with reference to the accompanying drawings. Those skilled in the art will readily appreciate that these drawings are for illustrative purposes only and are not intended to limit the scope of protection of the present invention. Among them:

[0038] FIG1 is a flow chart showing the main steps of a method for recommending energy replenishment resources according to an embodiment of the present invention;

[0039] FIG2 is a flow chart showing the main steps of training a resource fusion model according to an embodiment of the present invention;

[0040] FIG3 is a flowchart illustrating the main steps of obtaining fusion attribute information of each candidate energy replenishment station based on comparison information according to one embodiment of the present invention;

[0041] FIG4 is a flowchart illustrating main steps of training an energy replenishment resource recommendation model according to an embodiment of the present invention.

[0042] FIG5 is a schematic diagram of the main flow of a method for recommending energy replenishment resources according to an embodiment of the present invention.

[0043] FIG6 is a schematic diagram of the main structure of an electronic device according to an embodiment of the present invention.

[0044] Reference Signs List: 601: processor; 602: storage device. DETAILED DESCRIPTION

[0045] Some embodiments of the present invention are described below with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0046] In the description of the present invention, "module" and "processor" may include hardware, software, or a combination of both. A module may include hardware circuitry, various suitable sensors, communication ports, and memory. It may also include software components, such as program code, or a combination of software and hardware. A processor may be a central processing unit, a microprocessor, an image processor, a digital signal processor, or any other suitable processor. A processor has data and / or signal processing capabilities. A processor may be implemented in software, hardware, or a combination of both. Non-transitory computer-readable storage media include any suitable medium capable of storing program code, such as magnetic disks, hard disks, optical disks, flash memory, read-only memory, random access memory, and the like. The term "A and / or B" refers to all possible combinations of A and B, such as only A, only B, or both A and B. The terms "at least one of A or B" or "at least one of A and B" have similar meanings to "A and / or B" and may include only A, only B, or both A and B. The singular forms "one" and "the" may also include the plural forms.

[0047] Here we first explain some terms involved in the present invention.

[0048] Feature crossover is a feature engineering technique used to combine different features to generate new features. It can generate new features with higher-level correlations by combining, multiplying, adding, and subtracting original features. Feature crossover can improve the expressiveness of the model, enabling it to better capture interactions and nonlinear relationships between features. Feature crossover is commonly used in traditional machine learning and deep learning models to extract more expressive features.

[0049] Attention mechanism: This mechanism adjusts the weights of different parts of a model to better focus on and process important information in the input. The attention mechanism calculates an attention weight for each position in the input sequence and then applies these weights to different parts of the model to adjust their importance. This allows the model to more flexibly process information at different positions and achieve more accurate modeling. The attention mechanism is primarily used in deep learning models for tasks such as processing sequence data and image data, and has important applications in fields such as natural language processing and computer vision.

[0050] Machine learning algorithms are algorithms that learn patterns and patterns from data to perform tasks such as prediction, classification, and clustering. Based on mathematical and statistical principles, they extract features from large amounts of data and train models, enabling machines to automatically learn and improve, making accurate predictions or effective decisions based on unknown data.

[0051] As described in the background, with the increasing number of electric vehicles, the availability of various resources for charging electric vehicles is becoming increasingly dense. Traditional methods for recommending electric vehicle charging resources are mostly limited to a single site dimension. This involves ranking candidate resources based on the degree of match between resource attributes and user preferences, with the higher the ranking, the more compatible the resource is with the user. These methods typically only rank individual resources in the list, ignoring the correlations and differences between different resources. This can result in inaccurate and one-sided ranking results that fail to fully reflect the relative importance of resources in the list.

[0052] In resource recommendation scenarios, user choices are highly exclusive. After placing an order for one recommended item, users generally won't order any other resources on the list. Therefore, accurate resource recommendations are crucial. Simply measuring the degree of match between resource attributes and user preferences cannot accurately determine whether a candidate resource is optimal. Often, after the match exceeds a certain threshold, resource attributes are mostly consistent with the user's needs, with only a small difference. Therefore, selecting the optimal resource based on a single resource dimension is infeasible.

[0053] To address the aforementioned issue—that existing methods for recommending rechargeable resources are mostly limited to a single site, resulting in inaccurate recommendations—the present invention provides a rechargeable resource recommendation method, electronic device, and storage medium. By integrating information from multiple candidate rechargeable sites within a rechargeable resource list and comprehensively considering the relevance and differences between the candidate resources, more accurate and comprehensive recommendations can be provided.

[0054] Specifically, referring to FIG1 , FIG1 is a flow chart showing the main steps of a method for recommending energy replenishment resources according to an embodiment of the present invention. As shown in FIG1 , the method for recommending energy replenishment resources in the embodiment of the present invention mainly includes the following steps S101 to S103 .

[0055] Step S101: Acquire user information and resource information.

[0056] The resource information includes information of multiple candidate energy replenishment sites in the energy replenishment resource list.

[0057] Step S102: Input information of multiple candidate charging stations into the trained resource fusion model, obtain comparison information between the multiple candidate charging stations, and obtain fusion attribute information of each candidate charging station based on the comparison information.

[0058] Step S103: Input the user information and the fused attribute information of each candidate charging station into the trained charging resource recommendation model to obtain the charging resource recommendation result.

[0059] Based on the method described in steps S101 to S103 above, the information of multiple candidate charging stations in the list can be integrated, and the correlation and differences between multiple resources can be comprehensively considered. It can provide more accurate and comprehensive results based on user information and the integrated attribute information of each candidate charging station, and recommend the optimal resources that match the user, thereby improving the user's charging experience and satisfaction and increasing user stickiness. This solution has broad application prospects and commercial value.

[0060] The above steps S101 to S103 are further explained below.

[0061] In some implementations of the above step S101, the acquired user information includes at least one of user portrait information, context information, and information of a target object to be charged.

[0062] Among them, user portrait information is used to analyze users' charging behavior, such as users' preferences for distance, price, and charging speed.

[0063] The target object information to be charged is the portrait information of the vehicle to be charged, which mainly includes the vehicle's battery capacity, charging interface type, cruising range, charging speed, battery power, etc.

[0064] Contextual information is user-related information, such as user location information, time information when charging is required, scene information, weather information, environmental information, etc.

[0065] In some implementations of the above step S101, the acquired resource information includes at least one of site portrait information, battery portrait information, and charging pile portrait information of multiple candidate energy replenishment sites in the energy replenishment resource list.

[0066] Among them, the station portrait information is used to describe the basic attributes of the station resources, mainly including the station ID, station attributes, business policies and business rules, etc.

[0067] Among them, the station attributes mainly include distance, price, charging speed, brand type, availability, working status, geographical location, number of people in queue, number of available charging resources, user ratings, etc.

[0068] Specifically, the distance is the distance between the target object to be recharged and the candidate recharge station.

[0069] Availability refers to the degree to which a candidate charging station can function properly according to its design and intended functions. It is a key indicator of how well a charging station meets charging / swapping needs and provides a positive user experience. Specifically, the reliability of the charging equipment, including charging equipment and charging piles, determines the availability of the charging station. Frequent failures or downtime of charging equipment will impact the user experience and service stability.

[0070] Working status refers to the current operating state of a charging station. It is one of the key indicators for measuring the normal operation of a charging station. It typically includes: Charging pile status, that is, whether the charging pile is working properly and whether there are any abnormalities such as power outages or short circuits; Charging port status, that is, whether the charging port is working properly and whether there are any abnormalities such as poor connection or contact; Service status, that is, whether the charging / battery swapping service is being provided normally; Charging station status, that is, whether the charging station is operating normally, including equipment maintenance and troubleshooting.

[0071] Battery portrait information comes from charging stations and big data platforms, including basic battery information, power level, service time, number of uses, etc.

[0072] Furthermore, the battery portrait information may include long-term stable characteristic information such as battery basics, and may also include information with real-time characteristics such as the current battery power.

[0073] The charging pile portrait information comes from charging stations and big data platforms, including basic information of the charging pile, power, service time, number of uses, jump gun rate, etc.

[0074] Furthermore, the charging pile portrait information may include long-term stable characteristic information such as the number of uses and basic information, and may also include information with real-time characteristics such as the current available status of the charging pile.

[0075] It should be pointed out that the above examples of obtained user information and resource information are only illustrative. In actual applications, those skilled in the art can choose to obtain user information and resource information as needed according to specific scenarios. The relevant obtainable user information and resource information are all within the scope of protection of the present invention and are not limited here.

[0076] The above is a further description of step S101 , and the following further describes step S102 .

[0077] In some embodiments of the above step S102, when a user has a need for energy replenishment, a plurality of candidate energy replenishment stations in the energy replenishment resource list may be obtained first. Specifically, a preset number (e.g., 10) of energy replenishment stations are selected according to the following conditions to form an energy replenishment resource list for the user to select:

[0078] Location and distance: Choose a charging station that is close to the user's location to facilitate user access and reduce charging time and driving distance.

[0079] Recharging speed and power: Based on your charging needs and schedule, choose a charging station with the appropriate charging speed and power. For example, a fast charging station can complete charging needs in a short time, while a slow charging station is suitable for longer stays.

[0080] Recharging costs: Consider the costs of the recharging station, including service fees, electricity fees, and parking fees, and choose a suitable recharging station to control recharging costs.

[0081] Quality of charging equipment and service: Choose a charging station with good equipment quality, high service quality, and well-maintained charging equipment to ensure the safety and reliability of the charging process.

[0082] Idleness of charging facilities: Understand the usage of charging facilities and choose charging stations with higher idle rates to avoid inconveniences such as waiting or queuing.

[0083] Evaluation and reputation of charging stations: Consider the evaluation and reputation of other users on the charging stations, choose charging stations with better evaluations, and improve the charging experience and service quality.

[0084] It should be pointed out that the above method for obtaining multiple candidate charging stations in the charging resource list is only a schematic illustration. In actual application, those skilled in the art can choose the above method or other methods to screen out suitable charging stations for users to choose according to the actual needs and personal preferences of users. No limitation is made here.

[0085] Furthermore, in some implementations, information of multiple candidate energy replenishment stations may be input into a trained resource fusion model to obtain comparative information between the multiple candidate energy replenishment stations.

[0086] The comparison information between multiple candidate charging stations includes correlation information and difference information.

[0087] Specifically, the association information refers to the similar features or attributes among multiple candidate charging stations, and the connection between them can be established through these similarities.

[0088] For example, two candidate charging stations may be located in the same area, close to each other, have the same equipment type, and have a similar number of charging piles. These similarities can help us understand the relationship between them, such as whether they belong to the same service provider or are part of the same charging network.

[0089] Difference information refers to the different characteristics or attributes among multiple candidate charging stations, and these differences can be used to distinguish the differences between them.

[0090] For example, two charging stations may be located in different areas, be close to each other or far apart, have different equipment types, or have different numbers of charging piles. These differences can help us understand the differences between them, for example, they may belong to different service providers or serve different user groups.

[0091] By analyzing correlation information and difference information, we can better understand the relationship between multiple candidate refueling sites, and make further analysis and decisions based on this information, such as comparing their performance and service quality.

[0092] By inputting the information of multiple candidate energy replenishment stations in the energy replenishment resource list into the trained resource fusion model, the association information and difference information between the multiple candidate energy replenishment stations can be obtained.

[0093] Specifically, in the resource fusion model, feature cross-talk, attention mechanism, and other feature enhancement technologies can be used to enhance the expressive power of the resource fusion model and obtain more accurate output results.

[0094] Among them, feature crossover can be used to generate new features, which usually includes the following steps: selecting some features, randomly combining or linearly transforming these features, and then adding these new features to the model.

[0095] In some embodiments, feature intersection technology can be used to intersect the site attributes of multiple candidate charging sites, such as considering their distance, device type, number of charging piles, service fees, etc. By fusing these site attributes, a feature combination with correlation and difference can be obtained.

[0096] The attention mechanism can be used to adjust the weights of different parts of the model, which usually includes the following steps: first use an attention mechanism to calculate the attention weights for each position in the input sequence, and then apply these weights to different parts of the model to adjust their importance.

[0097] In some embodiments, the attention mechanism can determine association information by calculating the similarity or correlation between two sites, and determine difference information by calculating the contribution of each feature to the association information.

[0098] Further, referring to FIG2 , FIG2 is a flow chart showing the main steps of training a resource fusion model according to an embodiment of the present invention. As shown in FIG2 , the training process includes the following steps S201 to S203 .

[0099] Step S201: Obtain a first training set.

[0100] Among them, the first training set is the information of multiple charging stations marked with fusion attribute information labels.

[0101] Furthermore, the first training set can be obtained through an open source database, or data can be collected and processed by the user. For example, relevant data about multiple charging stations can be collected, including the number of charging piles, charging speed, service quality and other information, and these data can be pre-processed, such as normalized or standardized, to ensure the comparability and consistency of the data. Finally, the data can be labeled to obtain the first training set.

[0102] Step S202: training the resource fusion model based on the first training set.

[0103] The resource fusion model includes a fusion module, which is used to obtain comparative information between multiple candidate charging stations. Therefore, the fusion module in the resource fusion model can be trained based on the first training set to learn how to accurately calculate the association information and difference information between multiple charging stations.

[0104] Step S203: When the resource fusion model converges to a first preset error value, the resource fusion model training is completed.

[0105] During model training, appropriate loss functions and optimization algorithms can be used to minimize the gap between the predicted results and the true labels.

[0106] Specifically, the resource fusion model will continuously obtain information on multiple candidate charging stations, fuse and output the correlation information and difference information between multiple candidate charging stations, and continuously train the model based on the degree of similarity between the model's estimated results and the actual labels, iterating until the resource fusion model learns to the optimal level.

[0107] Furthermore, by inputting information of multiple candidate energy replenishment stations in the energy replenishment resource list into the trained resource fusion model, comparative information between the multiple candidate energy replenishment stations can be obtained.

[0108] In some other implementations of the above step S102, the fusion attribute information of each candidate charging station may be obtained based on the comparison information.

[0109] 3 , which is a flowchart illustrating the main steps of obtaining fused attribute information of each candidate charging station based on comparison information according to an embodiment of the present invention. As shown in FIG3 , the process includes the following steps S301 to S302 .

[0110] Step S301: Obtain the weight of the station attributes of each candidate refueling station based on the importance and comparison information of the station attributes.

[0111] The importance of station attributes indicates the degree of influence of different station attributes on the weight. The same station attribute will be assigned different weights in different charging stations.

[0112] For example, for charging station A, distance (i.e., the distance between the target object to be charged and A) is the most important station attribute, so a larger weight is assigned to distance, while a smaller weight is assigned to other station attributes (such as price); for charging station B, price is the most important station attribute, so a larger weight is assigned to price, while a smaller weight is assigned to other station attributes (such as distance).

[0113] The comparison information indicates the degree of influence of the association information and difference information between multiple charging stations on the attributes of different stations.

[0114] For example, based on the comparative information between charging stations A and B, charging station A has a faster charging speed, while charging station B has a smaller queue. Therefore, for charging station A, a larger weight may be assigned to charging speed, while other station attributes (such as the number of people in the queue) are assigned smaller weights; for charging station B, a larger weight may be assigned to the number of people in the queue, while other station attributes (such as charging speed) are assigned smaller weights.

[0115] Therefore, the weight of the station attributes of each candidate refueling station can be obtained based on the importance and comparison information of the station attributes.

[0116] Step S302: Perform weighted processing on the station attributes based on the weights to obtain fused attribute information of each candidate charging station.

[0117] Specifically, after obtaining the weight of the station attributes of each candidate recharging station, a machine learning algorithm can be used to perform weighted processing on the station attributes to obtain fused attribute information of each candidate recharging station.

[0118] Machine learning algorithms are algorithms that learn patterns and patterns from data to perform tasks such as prediction, classification, and clustering. Based on mathematical and statistical principles, they extract features from large amounts of data and train models, enabling machines to automatically learn and improve, making accurate predictions or effective decisions based on unknown data.

[0119] The fused attribute information is the information obtained by weighted processing of all station attributes of each candidate charging station.

[0120] The above is a further description of step S102 , and the following further describes step S103 .

[0121] In some implementations of the above step S103, the energy replenishment resource recommendation model includes a prediction module and / or a ranking module.

[0122] Among them, the prediction module is used to predict the conversion probability of each candidate energy replenishment site; the sorting module is used to sort all candidate energy replenishment sites to obtain the energy replenishment resource recommendation results.

[0123] Further, referring to FIG4 , FIG4 is a flowchart illustrating the main steps of training the energy replenishment resource recommendation model according to an embodiment of the present invention. As shown in FIG4 , the training process includes the following steps S401 to S403 .

[0124] Step S401: Obtain historical user information and historical resource information from at least part of historical orders as a second training set.

[0125] Specifically, feature information, namely, historical user information and historical resource information, may be extracted from the energy replenishment process of at least part of the historical orders.

[0126] Among them, historical user information includes historical user portrait information, historical portrait information of target objects to be replenished, and historical context information. Historical context information includes historical scene information, historical weather information, historical environmental information, etc.

[0127] Historical resource information includes historical energy replenishment station portrait information, historical battery portrait information, historical charging pile portrait information, etc.

[0128] By connecting the above information together, a set of all feature vectors of historical orders can be formed. Furthermore, the set of feature vectors can be used as a second training set.

[0129] Step S402: training the prediction module and / or ranking module of the energy replenishment resource recommendation model based on the second training set.

[0130] In some implementations, the historical candidate recharging stations in the second training set may be predicted based on the prediction module to obtain the conversion probability of each historical candidate recharging station.

[0131] Specifically, the prediction module can predict the input historical user information and historical resource information and output the user's conversion probability at the historical candidate charging station. The conversion probability is the predicted probability of being able to go to the charging station.

[0132] In some implementations, all historical candidate recharging stations may be ranked based on the ranking module and the conversion probability of each historical candidate recharging station to obtain a recharging resource recommendation result.

[0133] Specifically, the sorting module will sort the conversion probabilities of all input historical candidate recharging sites in descending order, and finally output a recommended list of recharging resources to the user.

[0134] Step S403: When the energy replenishment resource recommendation model converges to a second preset error value, the energy replenishment station recommendation model training is completed.

[0135] During model training, appropriate loss functions and optimization algorithms can be used to minimize the gap between the predicted results and the true labels.

[0136] Specifically, the refueling resource recommendation model continuously collects user feedback from historical orders, estimates and outputs the user's conversion probability at each candidate refueling station, and continuously trains the model based on the degree of similarity between the model's output and the actual refueling situation. This process is repeated until the refueling resource recommendation model is optimal.

[0137] Furthermore, after the energy replenishment resource recommendation model is trained, energy replenishment resource recommendation results can be provided to users.

[0138] The above is a further explanation of step S103.

[0139] In some implementations, referring to FIG5 , FIG5 is a schematic diagram of the main flow of a method for recommending energy replenishment resources according to an embodiment of the present invention.

[0140] As shown in FIG5 , when a user has a need for energy replenishment, user information and resource information, that is, information of multiple candidate energy replenishment sites in the energy replenishment resource list, can be obtained.

[0141] Furthermore, the information of multiple candidate energy replenishment stations is input into the trained resource fusion model to obtain the fusion attribute information of each candidate energy replenishment station.

[0142] Furthermore, the user information and the fused attribute information of each candidate charging station are input into the trained charging resource recommendation model to obtain the charging resource recommendation results.

[0143] Through the above implementation, the information of multiple candidate charging stations in the list can be integrated, and the correlation and differences between multiple resources can be comprehensively considered. It can provide more accurate and comprehensive results based on user information and the integrated attribute information of each candidate charging station, and recommend the optimal resources that match the user, thereby improving the user's charging experience and satisfaction and increasing user stickiness. This solution has broad application prospects and commercial value.

[0144] It should be pointed out that although the various steps in the above embodiments are described in a specific order, those skilled in the art will understand that in order to achieve the effects of the present invention, different steps do not have to be performed in such an order. They can be performed simultaneously (in parallel) or in other orders. These changes are within the scope of protection of the present invention.

[0145] It will be understood by those skilled in the art that all or part of the processes in the method for implementing the above embodiment of the present invention may also be completed by instructing the relevant hardware through a computer program. The computer program may be stored in a computer-readable storage medium. When the computer program is executed by a processor, it may implement the steps of each of the above method embodiments. The computer program includes computer program code, which may be in source code form, object code form, executable file, or some intermediate form. The computer-readable storage medium may include: any entity or device, medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory, random access memory, electric carrier signal, telecommunication signal, and software distribution medium, etc., which can carry the computer program code.

[0146] Furthermore, the present invention also provides an electronic device. Refer to Figure 6, which is a schematic diagram of the main structure of an electronic device according to an embodiment of the present invention. As shown in Figure 6, the electronic device in the embodiment of the present invention mainly includes a processor 601 and a storage device 602. The storage device 602 can be configured to store a program for executing the energy replenishment resource recommendation method of the above-mentioned method embodiment, and the processor 601 can be configured to execute the program in the storage device 602, which includes but is not limited to a program for executing the energy replenishment resource recommendation method of the above-mentioned method embodiment. For ease of explanation, only the parts related to the embodiment of the present invention are shown. For specific technical details not disclosed, please refer to the method part of the embodiment of the present invention.

[0147] In some possible implementations of the present invention, the electronic device may include multiple processors 601 and multiple storage devices 602. The program for executing the energy replenishment resource recommendation method of the above-mentioned method embodiment may be divided into multiple subprograms, and each subprogram may be loaded and run by the processor 601 to execute different steps of the energy replenishment resource recommendation method of the above-mentioned method embodiment. Specifically, each subprogram may be stored in different storage devices 602, and each processor 601 may be configured to execute the program in one or more storage devices 602 to jointly implement the energy replenishment resource recommendation method of the above-mentioned method embodiment, that is, each processor 601 executes different steps of the energy replenishment resource recommendation method of the above-mentioned method embodiment to jointly implement the energy replenishment resource recommendation method of the above-mentioned method embodiment.

[0148] The multiple processors 601 may be processors deployed on the same device. For example, the electronic device may be a high-performance device composed of multiple processors, and the multiple processors 601 may be processors configured on the high-performance device. Furthermore, the multiple processors 601 may be processors deployed on different devices. For example, the electronic device may be a server cluster, and the multiple processors 601 may be processors on different servers in the server cluster.

[0149] Furthermore, the present invention also provides a computer-readable storage medium. In a computer-readable storage medium embodiment according to the present invention, the computer-readable storage medium can be configured to store a program for executing the energy replenishment resource recommendation method of the above-mentioned method embodiment, and the program can be loaded and run by the processor to implement the above-mentioned energy replenishment resource recommendation method. For ease of explanation, only the parts related to the embodiment of the present invention are shown. For specific technical details not disclosed, please refer to the method part of the embodiment of the present invention. The computer-readable storage medium can be a storage device formed by various electronic devices. Optionally, the computer-readable storage medium in the embodiment of the present invention is a non-transitory computer-readable storage medium.

[0150] It should be noted that the relevant user personal information that may be involved in the various embodiments of this application is strictly in accordance with the requirements of laws and regulations, follows the principles of legality, legitimacy and necessity, and is based on the reasonable purposes of business scenarios to process personal information that users actively provide during the use of products / services or generated due to the use of products / services, as well as personal information obtained with the user's authorization.

[0151] The personal information processed by the Applicant will vary depending on the specific product / service scenario and will be based on the specific scenario in which the user uses the product / service. This may involve the user's account information, device information, driving information, vehicle information, or other related information. The Applicant will treat the user's personal information and its processing with a high degree of diligence.

[0152] The Applicant attaches great importance to the security of user personal information and has taken reasonable and feasible security measures that comply with industry standards to protect user information and prevent personal information from being accessed, disclosed, used, modified, damaged or lost without authorization.

[0153] Thus far, the technical solution of the present invention has been described in conjunction with an embodiment shown in the accompanying drawings. However, it is readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art may make equivalent changes or substitutions to the relevant technical features, and the technical solutions after such changes or substitutions will fall within the scope of protection of the present invention.

Claims

1. A method for recommending energy replenishment resources, characterized in that: The method comprises: Acquire user information and resource information, wherein the resource information includes information of a plurality of candidate energy replenishment sites in the energy replenishment resource list; Inputting the information of the plurality of candidate energy replenishment stations into a trained resource fusion model, obtaining comparison information between the plurality of candidate energy replenishment stations, and obtaining fusion attribute information of each candidate energy replenishment station based on the comparison information; The user information and the fused attribute information of each candidate energy replenishment site are input into a trained energy replenishment resource recommendation model to obtain an energy replenishment resource recommendation result.

2. The energy supplement resource recommendation method according to claim 1, characterized in that: The obtaining of user information comprises: Obtain at least one of user portrait information, context information, and target object information to be supplemented; The context information includes at least one of location information, time information, scene information, and weather information.

3. The energy supplement resource recommendation method according to claim 1, characterized in that: The obtaining of resource information includes: Obtain at least one of the station portrait information, battery portrait information, and charging pile portrait information of the plurality of candidate energy replenishment stations in the energy replenishment resource list; The station portrait information includes the station ID and station attributes.

4. The energy supplement resource recommendation method according to claim 3, characterized in that: The station attribute includes at least one of distance, price and charging speed, the comparison information includes association information and difference information between the multiple candidate charging stations; and obtaining fusion attribute information of each candidate charging station based on the comparison information includes: Obtaining a weight of the station attribute of each of the candidate energy replenishment stations based on the importance of the station attribute and the comparison information; The station attributes are weighted based on the weights to obtain each of the candidate supplementary The fusion attribute information of the energy station.

5. The energy supplement resource recommendation method according to claim 1, characterized in that: The method further comprises training the resource fusion model based at least on the following steps: Acquire a first training set, where the first training set is energy replenishment stations marked with fusion attribute information labels; Training the resource fusion model based on the first training set; When the resource fusion model converges to a first preset error value, the resource fusion model training is completed.

6. The energy supplement resource recommendation method according to claim 1, characterized in that: The energy replenishment resource recommendation model includes a prediction module and / or a sorting module; the method further includes training the energy replenishment resource recommendation model based at least on the following steps: Acquire historical user information and historical resource information in at least part of the historical orders as a second training set; Training a prediction module and / or a sorting module of the energy replenishment resource recommendation model based on the second training set; When the energy replenishment resource recommendation model converges to a second preset error value, the energy replenishment site recommendation model training is completed.

7. The energy supplement resource recommendation method according to claim 6, characterized in that: The training of the prediction module of the energy replenishment resource recommendation model based on the second training set includes: Based on the prediction module, the historical candidate energy replenishment sites in the second training set are predicted to obtain the conversion probability of each of the historical candidate energy replenishment sites.

8. The energy supplement resource recommendation method according to claim 7, characterized in that: The method further comprises: Based on the sorting module and the conversion probability of each of the historical candidate energy replenishment sites, all the historical candidate energy replenishment sites are sorted to obtain the energy replenishment resource recommendation result.

9. An electronic device comprising a processor and a storage device, wherein the storage device is suitable for storing a plurality of program codes, characterized in that: The program code is suitable for being loaded and run by the processor to execute the energy replenishment resource recommendation method according to any one of claims 1 to 8.

10. A computer-readable storage medium storing a plurality of program codes, characterized in that: The program code is suitable for being loaded and run by a processor to execute the energy replenishment resource recommendation method according to any one of claims 1 to 8.

Images