Path generation method for two-wheeled electric vehicle, electronic equipment and storage medium

Abstract

The invention relates to a path generation method for a two-wheeled electric vehicle, electronic equipment and a storage medium, and the method comprises the steps: carrying out the data collection of a user terminal, and obtaining the actual driving data and destination information of a user; analyzing the actual driving data through a pre-trained trajectory generation model to obtain a plurality of candidate point locations; determining a priority destination from the destination information according to a travel control strategy of the current user, performing distance analysis on a plurality of candidate point locations and the position information of the priority destination, and selecting the candidate point location closest to the priority destination as a next movement point location of the user; and performing information complementation on the next motion point location through the point location generation control quantity to obtain predicted trajectory data, and sending the predicted trajectory data to the user. The optimal road navigation information is provided for the user through the deep learning model, optimization is performed in combination with the destination of the user, and finally the nearest point is obtained as the optimal next point driving position for the user to go to the destination.

Description

Technical Field

[0001] The present application relates to the technical field of electric vehicle travel analysis, and in particular to a path generation method, electronic device and storage medium for two-wheeled electric vehicles. Background Art

[0002] As a light, economical, low-carbon and environmentally friendly way of traveling on urban roads, electric two-wheelers are widely accepted by the public, and their market share is increasing day by day. Compared with cars, electric two-wheelers are gradually showing vigorous vitality in the fields of instant delivery, short-distance travel, logistics express, shared riding, etc. Therefore, the riding efficiency of users on urban roads has gradually become the focus of attention of these industries. In terms of improving user riding efficiency, navigation algorithms are extremely important tools that can help users make quick path decisions in unfamiliar environments and select the best route to achieve their riding goals.

[0003] At present, most of the existing path navigation technologies are based on the relevant information of the map road, which is characterized as a vector or graph, and is calculated and guided by algorithms such as Dijkstra, A-star, and D-star. However, these methods are all based on the established map road information. Although they can guarantee arrival, they are not the optimal solution. Because the characteristic of electric two-wheeled vehicles is that they have a wider road driving area than cars, they can travel through streets and alleys such as buildings and communities, and even some inaccessible places marked on maps. Therefore, the traditional regularized path navigation method based on road information is not satisfactory when providing navigation services for electric two-wheeled vehicle users. The current fast and efficient travel mode still depends on the rider's experience, which is particularly evident in the instant delivery industry. For example, some passable areas can quickly reach the destination, but traditional navigation uses a longer-distance route to guide users; some complex road conditions cannot be captured by traditional algorithms, which will reduce the user's riding efficiency.

[0004] Therefore, the existing route recommendation algorithm is only applicable to users of four-wheeled vehicles, and there is a lack of a more effective route navigation method for users of electric two-wheeled vehicles. Summary of the invention

[0005] The embodiments of the present application provide a path generation method, an electronic device and a storage medium for a two-wheeled electric vehicle, so as to at least solve the problem that the related art lacks a path navigation method for electric two-wheeled vehicle users.

[0006] In a first aspect, an embodiment of the present application provides a path generation method for a two-wheeled electric vehicle, comprising: Collecting data from the user terminal to obtain the user's actual driving data and destination information, wherein the destination information includes location information of at least one destination; The actual driving data is analyzed by a pre-trained trajectory generation model to obtain a plurality of candidate points; wherein the trajectory generation model is trained by a large amount of riding trajectory data of two-wheeled electric vehicles; Determine a priority destination from the destination information according to the current user's travel control strategy, perform distance analysis between multiple candidate points and the priority destination location information, and select the candidate point closest to the priority destination as the user's next movement point; The control amount is generated by the point to complete the information of the next moving point, obtain the predicted trajectory data and send it to the user.

[0007] In one embodiment, the method further comprises: When the user drives according to the predicted trajectory data, data is collected in real time to obtain new actual driving data; Determine whether the user has reached all the destinations in the destination information. If so, determine that navigation is finished; if not, splice the predicted trajectory data and the new actual driving data and input them into the trajectory generation model to regenerate the predicted trajectory data of the next destination.

[0008] In one embodiment, after obtaining a plurality of candidate points, the method further includes: The movement probability of each candidate point is calculated through the pre-trained trajectory generation model; A preset probability threshold is obtained to determine whether the movement probability is greater than the probability threshold. If so, the corresponding candidate point is retained; if not, the corresponding candidate point is deleted.

[0009] In one embodiment, determining the priority destination from the destination information according to the travel control strategy of the current user includes: Obtaining the current location of the user, calculating the distance between each destination in the destination information and the current location, sorting the distances in descending order, and taking the destination ranked first as a priority destination; Alternatively, the limited arrival time of each destination in the destination information is obtained, and the destinations are sorted in the order of the limited arrival times, and the destination ranked first is taken as the priority destination.

[0010] In one embodiment, the training process of the trajectory generation model includes: Build a model framework and determine the training strategy to obtain the initial trajectory generation model; Acquire historical driving data of the user, and divide the historical driving data into a training set, a validation set, and a test set; The initial trajectory generation model is trained according to the training strategy and the training set, and the parameters of the initial trajectory generation model are adjusted through the validation set to obtain an intermediate trajectory generation model; The test set is used to evaluate the performance of the intermediate trajectory generation model, and when the evaluation result meets the preset user requirements, the trajectory generation model is obtained.

[0011] In one embodiment, the model framework includes: A data embedding layer is used to map multi-dimensional actual driving data or historical driving data into a high-dimensional feature vector; A position encoding module is used to sequentially label the high-dimensional feature vector in the length direction of the trajectory sequence; Multiple encoding modules are used to perform in-depth trajectory feature mining on high-dimensional feature vectors containing sequential tags to obtain multi-dimensional feature data; A linear layer, used for performing data conversion according to the multi-dimensional characterization to obtain relative motion positions of multiple candidate points; The probability distribution calculation module is used to calculate the movement probability of the user to the candidate point according to the relative movement position.

[0012] In one embodiment, the encoding module includes a first normalization layer, a multi-head attention layer, a second normalization layer and a feedforward layer, wherein: Normalizing the high-dimensional feature vector through the first normalization layer to obtain first feature data; Capturing the dependency between the first feature data by the multi-head attention layer; The dependency relationship is merged with the first feature data to obtain second feature data, and the second feature data is normalized by the second normalization layer to obtain third feature data; Performing further feature extraction on the third feature data through the feed-forward layer to obtain fourth feature data; The second feature data and the fourth feature data are fused to obtain the multi-dimensional feature data.

[0013] In one embodiment, the actual driving data includes trajectory information of multiple points and electric two-wheeled vehicle information; The trajectory information includes the generation time, distance and geocoding of each point; the electric two-wheeled vehicle includes battery type, battery health, battery charge and battery power.

[0014] In a second aspect, an embodiment of the present application provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the path generation method for a two-wheeled electric vehicle as described in the first aspect above is implemented.

[0015] In a third aspect, an embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the path generation method for a two-wheeled electric vehicle as described in the first aspect above.

[0016] The path generation method, system, device and storage medium for two-wheeled electric vehicles provided in the embodiments of the present application have at least the following technical effects: The path generation method for two-wheeled electric vehicles provided in this application is a new navigation method that combines generative deep learning with travel control strategies. In this application method, the spatiotemporal trajectory large model that has learned the riding experience of a large number of users can provide users with the best road navigation information. Through deep learning, trajectory data that can take shortcuts (such as paths between office buildings) can be provided to two-wheeled electric vehicle users; at the same time, the destination that the user expects to reach is combined as the guidance target, and the candidate points generated by the large model are assisted in the optimization, and finally the nearest point is obtained as the optimal next point driving position for the user to the destination, thereby providing two-wheeled electric vehicle users with a short distance and short time navigation route.

[0017] Details of one or more embodiments of the present application are set forth in the following drawings and description to make other features, objects, and advantages of the present application more readily apparent. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings: Figure 1 is a flow chart of a path generation method for a two-wheeled electric vehicle in one embodiment of the present application; Figure 2 This is a basic architecture diagram based on a shared battery trajectory model in an embodiment of the present application; Figure 3 is a flow chart of a path generation method for a two-wheeled electric vehicle in another embodiment of the present application; Figure 4 It is a basic architecture diagram of an electronic device in one embodiment of the present application. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application is described and illustrated below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not intended to limit the present application. Based on the embodiments provided in the present application, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present application.

[0020] Obviously, the drawings described below are only some examples or embodiments of the present application. For ordinary technicians in this field, the present application can also be applied to other similar scenarios based on these drawings without creative work. In addition, it can also be understood that although the efforts made in this development process may be complicated and lengthy, for ordinary technicians in this field related to the content disclosed in this application, some changes in design, manufacturing or production based on the technical content disclosed in this application are just conventional technical means, and should not be understood as insufficient content disclosed in this application.

[0021] Reference to "embodiments" in this application means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those of ordinary skill in the art that the embodiments described in this application may be combined with other embodiments without conflict.

[0022] Unless otherwise defined, the technical terms or scientific terms involved in this application should be understood by people with ordinary skills in the technical field to which this application belongs. The words "one", "a", "a", "the" and the like involved in this application do not indicate a quantity limitation, and may indicate the singular or plural. The terms "include", "comprise", "have" and any of their variations involved in this application are intended to cover non-exclusive inclusions; for example, a process, method, system, product or device that includes a series of steps or modules (units) is not limited to the listed steps or units, but may also include steps or units that are not listed, or may also include other steps or units inherent to these processes, methods, products or devices. The words "connect", "connected", "coupled" and the like involved in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The "multiple" involved in this application refers to two or more. "And / or" describes the association relationship of associated objects, indicating that there may be three relationships, for example, "A and / or B" can mean: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the objects before and after are in an "or" relationship. The terms "first", "second", "third", etc. involved in this application are only used to distinguish similar objects and do not represent a specific ordering of the objects.

[0023] In recent years, with the rapid development of deep learning, a large number of algorithm models based on deep learning have emerged. By using deep learning algorithm models to learn and mine massive data for a long time, we can learn output results that are more in line with business scenarios. At present, it has played a significant role in finance, medical care, meteorology, social analysis and other fields. In the field of path generation and navigation, the path navigation method of automobiles has been very mature. A large number of rules or AI-based navigation algorithms have not only achieved excellent results in path navigation, but also produced great success in the automatic guidance of unmanned vehicles. However, these algorithms are all for car users. Compared with car users, electric two-wheeled vehicles are very different in many aspects such as driving roads, mileage, driving habits, etc.

[0024] For users of two-wheeled electric vehicles, this application proposes a new navigation method for electric two-wheeled vehicle users that combines generative deep learning with control strategies. In this method, a large spatiotemporal trajectory model that has learned a large number of user riding experiences can provide users with the best road navigation information; at the same time, the user's desired destination is combined as the guidance target to assist in the optimization of candidate points generated by the large model, and finally the nearest point is obtained as the optimal next point driving position for the user to the destination.

[0025] In a first aspect, the present application provides a path generation method for a two-wheeled electric vehicle. Figure 1is a flow chart of a path generation method for two-wheeled electric vehicles, such as Figure 1 As shown, the method includes the following steps.

[0026] Step S1, collect data from the user terminal to obtain the user's actual driving data and destination information, wherein the destination information includes at least one location information of the destination. In this embodiment, the actual driving data includes track information of multiple points and electric two-wheeled vehicle information.

[0027] Specifically, each actual trajectory data is composed of tens to hundreds of multi-dimensional trajectory point information, which constitutes the user's riding habits including various time periods and situations spanning several days. Each trajectory point information contains two major dimensions, namely trajectory information and electric two-wheeled vehicle information. The trajectory information reflects the spatiotemporal attributes of the trajectory and is the result of the trajectory presentation, including: time, distance, geocoding, next point movement point, etc. The electric two-wheeled vehicle information reflects the dynamic characteristics of the driving process and is the basis of the trajectory generation process, including: battery type, battery health, battery charge, battery power and other information. Before inputting the data into the model, all the data will be labeled to facilitate the feature extraction and identification of the model.

[0028] Step S2, analyzing the actual driving data through a pre-trained trajectory generation model to obtain multiple candidate points, and the candidate points are used to guide the user to move to the destination; wherein the trajectory generation model is trained by a large amount of riding trajectory data of two-wheeled electric vehicles.

[0029] In the embodiment of the present application, the training process of the trajectory generation model is as follows: A model framework is constructed and a training strategy is determined to obtain an initial trajectory generation model; user historical driving data is obtained, and the historical driving data is divided into a training set, a validation set, and a test set; the initial trajectory generation model is trained according to the training strategy and the training set, and the parameters of the initial trajectory generation model are adjusted through the validation set to obtain an intermediate trajectory generation model; the test set is used to evaluate the performance of the intermediate trajectory generation model, and when the evaluation result meets the preset user requirements, the trajectory generation model is obtained.

[0030] In the embodiment of the present application, the model framework includes a data embedding layer, a position encoding module, N encoding modules (N is a natural number greater than 0), a linear layer, and a probability distribution calculation module. In the training process, the historical driving data (historical trajectory data) is used as the training input, and then the data embedding layer is used to map the multi-dimensional historical driving data into a high-dimensional feature vector; the position encoding module is used to apply sequential markings to the high-dimensional feature vector in the length direction of the trajectory sequence; multiple encoding modules are used to perform in-depth trajectory feature mining on the high-dimensional feature vector containing sequential markings to obtain multi-dimensional feature data; the linear layer is used to perform data conversion according to the multi-dimensional characterization to obtain the relative motion positions of multiple candidate points; the probability distribution calculation module is used to calculate the probability of the user moving to the candidate point according to the relative motion position.

[0031] More specifically, the encoding module includes a first normalization layer, a multi-head attention layer, a second normalization layer and a feedforward layer, wherein the high-dimensional feature vector is normalized by the first normalization layer to obtain first feature data; the dependency relationship between the first feature data is captured by the multi-head attention layer; the dependency relationship is fused with the first feature data to obtain second feature data, and the second feature data is normalized by the second normalization layer to obtain third feature data; the third feature data is further extracted by the feedforward layer to obtain fourth feature data; the second feature data and the fourth feature data are fused to obtain the multi-dimensional feature data.

[0032] After the trajectory generation model is obtained, the actual riding trajectory data of the current user's two-wheeled electric vehicle (that is, all trajectory data before the current moment) is input into the trajectory generation model to obtain multiple candidate points and corresponding movement probabilities.

[0033] In a preferred embodiment, the main framework of the trajectory generation model of this embodiment is as follows: Figure 2 As shown in the figure. The model framework mainly consists of a data embedding layer, position encoding, multiple encoding modules, a linear layer, and probability distribution calculation. During the training process, the historical trajectory data is used as the training input. The data embedding layer is the first processing link of the data input. The labeled multi-dimensional sequence trajectory information will be mapped to a high-dimensional feature vector through this layer, which is convenient for the subsequent analysis of the network structure. The position encoding process applies sequential labels to the characterized vectors in the length direction of the trajectory sequence, which facilitates the model to identify the trajectory sequence. The encoding module is mainly used to deeply mine the high-dimensional trajectory feature information. The specific encoding modules include normalization layer 1, Transformer multi-head attention layer, normalization layer 2, and feedforward layer.

[0034] The data processed by a single encoding module has the same dimension as the input feature data. The data output after the encoding module is processed will continue to be input into the encoding module of the same structure in sequence. After a series of encoding modules have deeply characterized the data, it is output to a linear layer, which is used to convert the multi-dimensional feature data into a set of feature vectors that simply reflect the relative motion position of the next point. The dimension of this feature vector is the same as the number of labels of the relative motion position of the next point. Finally, this set of vector information is input into the probability distribution calculation module to analyze the probability distribution of the possibility of each nearby geographical location. The training output is the next point result, including the coordinates of the location where the user may go and the possibility of going to that location (i.e., the movement probability).

[0035] During the model training process, the Transformer multi-head attention layer of the encoding module uses the target mask to cover the length of the original trajectory sequence in turn, making full use of the data in the original trajectory sequence to improve the training efficiency of the next point prediction result. The maximum probability results of a series of next points obtained through training and the actual position labels of the next points are subjected to cross entropy loss calculation, and the result of this loss function is used as the basis for updating the parameters of the model forward propagation.

[0036] Step S3, determining a priority destination from the destination information according to the current user's travel control strategy, performing distance analysis between multiple candidate points and the priority destination location information, and selecting the candidate point closest to the priority destination as the user's next movement point.

[0037] After obtaining multiple candidate points, the movement probability of each candidate point is calculated through the pre-trained trajectory generation model; a preset probability threshold is obtained to determine whether the movement probability is greater than the probability threshold, if so, the corresponding candidate point is retained; if not, the corresponding candidate point is deleted.

[0038] For determining the priority destination from the destination information according to the travel control strategy of the current user, the user's current location can be obtained, the distance between each destination in the destination information and the current location can be calculated, the distances can be sorted in order from large to small, and the destination ranked first can be used as the priority destination; or, the limited arrival time of each destination in the destination information can be obtained, the limited arrival times can be sorted in order of chronological order, and the destination ranked first can be used as the priority destination.

[0039] In a preferred embodiment, for a trajectory generation model that has been trained, a historical trajectory is input, and the model will generate the motion position information of the next point. However, the next point information alone cannot guide the user to move to a specific target location. In order to establish a connection between this position information and the user's navigation towards the destination, we use a set of control strategies to assist in the generation of navigation trajectories.

[0040] Step S4, the next moving point is supplemented with information through point generation control, and the predicted trajectory data is obtained and sent to the user. In the embodiment of the present application, the model generation method combined with the control only obtains the position information of the next point. In order to supplement other dimensional information of the trajectory information, we combine the point generation control quantity for supplementation. For example, the complete trajectory and multi-dimensional information of the electric two-wheeled vehicle are calculated through the control quantities such as the power consumption during the movement and the average movement time.

[0041] Step S5, while the user is driving according to the predicted trajectory data, data is collected in real time to obtain new actual driving data. Finally, it is determined whether the user has reached all the destinations in the destination information. If so, the navigation is terminated; if not, the predicted trajectory data and the new actual driving data are spliced ​​and input into the trajectory generation model to regenerate the predicted trajectory data of the next destination.

[0042] Specifically, the completed trajectory information will be judged to see whether all destination points have been reached. If all the set destinations have been reached, the navigation is judged to be over. If not all the set destinations have been reached, the inferred trajectory point information (i.e., the new user trajectory data) and the user's current trajectory data are spliced, combined with the remaining destination information and then input into the trajectory generation model for the next inference cycle. In this iterative process, the destination information will delete the information of each point it arrives at, until all destinations have been reached and the entire path inference process is over.

[0043] The implementation process of the control strategy is as follows Figure 3As shown, in this embodiment, the current trajectory data and destination of the user are used as input information, and the current trajectory data is predicted and analyzed by the trajectory generation model (i.e., the pre-trained model) to obtain multiple candidate points. At the same time, when there are multiple input destinations, the destination to be visited first is selected from the multiple destinations. For selecting the destination to be visited first, the current geographical location of the current two-wheeled electric vehicle can be obtained from the current trajectory data, and then the destination closest to the geographical location can be selected from multiple destinations as the destination to be visited first. Then, multiple candidate points are combined and analyzed with the priority destination, and the candidate point closest to the priority destination can be selected as the next movement point of the user, and then the trajectory information of the next movement point and the current geographical location of the user is completed by the point generation control amount to obtain the next segment of trajectory data. Then, it is judged that all the input destinations have been arrived. If so, the path generation is terminated; if not, the new user trajectory data and the user's current trajectory data are spliced, and the destination that has not been reached (i.e., the remaining destination) is combined as the input information for the next round of path generation, so as to obtain the trajectory data of the remaining destinations, and this cycle is repeated until all destinations have been arrived.

[0044] Specifically, in the actual navigation process, the user's current travel trajectory data and destination information are input into the entire algorithm. The destination can be the geographical longitude and latitude points that multiple users want to reach. The trajectory sequence information will be analyzed by the pre-trained model (i.e., the trajectory generation module). After the model is calculated, a series of candidate points will be generated. These candidate points indicate the probability of the user moving to the next point. Since there are many next point movements, we use a specific probability threshold to filter these candidate points here, and only retain the next candidate movement points with a higher probability recommended by the model. There are usually multiple destination information entered by the user, so it is also necessary to first determine the order of walking to these destinations. Here, we combine the user's current location in the user's current trajectory data as a reference point, and first select the point closer to the user among many destinations as the user's priority destination. This priority destination is then combined with multiple candidate points to determine which candidate point is closest to this destination, and the nearest next candidate point is used as the movement position to guide the user.

[0045] When using the path generation method of the present application, the user can select multiple target locations on the APP or web platform and start to confirm the navigation generation process. At the same time, the APP or terminal hardware provides the user with the current GPS or location information and battery-related information, which is combined with the user's target point information and transmitted back to the embedded computing terminal or cloud for data characterization and path generation. After the predicted path is generated, the embedded edge calculator or cloud will return the navigation path results to the user's interactive platform to display the navigation results to the user. In the process of providing this service, the trajectory generation model will continue to learn and update the version in combination with the user's personalized information, so as to better provide users with more matching navigation results and improve the user's riding efficiency.

[0046] Compared with traditional navigation methods, this application makes full use of the large model to deeply mine the rider's riding experience in real scenes. Therefore, during the navigation process, it will output an effect that is completely different from the commercial map-based road framework navigation. The road navigation information generated by the present invention has a faster path than conventional navigation. Most of the generated roads pass through some shortcuts and paths between communities, and there are fewer traffic lights, so it can reach the target location more quickly and conveniently. In addition, since the model learns the behavior patterns of riders in different time periods and riding strategies under different battery conditions during the training process, the routes planned for users in different time periods are different, which can automatically reduce the areas passing through poor road conditions; at the same time, when the rider is riding for a longer distance, the power information of the trajectory generated by the model will also prompt the user to change the battery immediately, reducing the user's "mileage anxiety", which is more advantageous for delivery personnel who need to quickly plan roads on complex roads.

[0047] In summary, the path generation method for two-wheeled electric vehicles provided in the embodiment of the present application uses the powerful learning ability of deep neural networks, and with the help of the spatiotemporal trajectory data of the historical movements of a large number of electric two-wheeled vehicle users, it fully learns the user's relevant habits in terms of electric two-wheeled vehicle trajectories and battery performance, thereby providing users with generative path navigation capabilities that are more in line with real scenarios. This method solves the problem that traditional rule-based navigation algorithms cannot learn the riding experience of riders in complex scenarios and are difficult to adapt to the navigation of electric two-wheeled vehicle users with greater flexibility. The new deep learning model empowers users with the implicit riding experience of a large number of users, providing more efficient riding efficiency for electric two-wheeled vehicle users and related industry employees.

[0048] In a second aspect, an embodiment of the present application provides an electronic device, Figure 4 FIG. 1 is a block diagram of an electronic device according to an exemplary embodiment. Figure 4 As shown, the electronic device may include a processor 11 and a memory 12 storing computer program instructions.

[0049] Specifically, the processor 11 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

[0050] Among them, the memory 12 may include a large-capacity memory for data or instructions. By way of example and not limitation, the memory 12 may include a hard disk drive (HDD), a floppy disk drive, a solid-state drive (SSD), a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive, or a combination of two or more of these. Where appropriate, the memory 12 may include a removable or non-removable (or fixed) medium. Where appropriate, the memory 12 may be inside or outside the data processing device. In a specific embodiment, the memory 12 is a non-volatile memory. In a specific embodiment, the memory 12 includes a read-only memory (ROM) and a random access memory (RAM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically alterable ROM (EAROM) or a flash memory (FLASH), or a combination of two or more of these. Under appropriate circumstances, the RAM can be a static random access memory (SRAM) or a dynamic random access memory (DRAM), wherein the DRAM can be a fast page mode dynamic random access memory (FPMDRAM), an extended data output dynamic random access memory (EDODRAM), a synchronous dynamic random access memory (SDRAM), etc.

[0051] The memory 12 may be used to store or cache various data files that need to be processed and / or used for communication, as well as possible computer program instructions executed by the processor 11 .

[0052] The processor 11 reads and executes computer program instructions stored in the memory 12 to implement any one of the path generation methods for the two-wheeled electric vehicle in the above embodiments.

[0053] In one embodiment, the electronic device may further include a communication interface 13 and a bus 10. Figure 4 As shown, the processor 11, the memory 12, and the communication interface 13 are connected via a bus 10 and communicate with each other.

[0054] The communication interface 13 is used to realize the communication between the modules, devices, units and / or equipment in the embodiment of the present application. The communication port 13 can also realize data communication with other components such as: external devices, image / data acquisition equipment, databases, external storage and image / data processing workstations.

[0055] The bus 10 includes hardware, software or both, and couples the components of the electronic device to each other. The bus 10 includes but is not limited to at least one of the following: a data bus, an address bus, a control bus, an expansion bus, and a local bus. By way of example and not limitation, bus 10 may include an Accelerated Graphics Port (AGP) or other graphics bus, an Extended Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a Hyper Transport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an InfiniBand interconnect, a Low Pin Count (LPC) bus, a memory bus, a Micro Channel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local Bus (VLB) bus, or other suitable buses or a combination of two or more of these. Where appropriate, bus 10 may include one or more buses. Although embodiments of the present application describe and illustrate a particular bus, the present application contemplates any suitable bus or interconnect.

[0056] In a third aspect, an embodiment of the present application provides a computer-readable storage medium having a program stored thereon, and when the program is executed by a processor, the path generation method for the two-wheeled electric vehicle provided in the first aspect is implemented.

[0057] The readable storage medium may include but is not limited to: a portable disk, a hard disk, a random access memory, a read-only memory, an erasable programmable read-only memory, an optical storage device, a magnetic storage device or any suitable combination of the above.

[0058] In a possible implementation, the present invention can also be implemented in the form of a program product, which includes a program code. When the program product is run on a terminal device, the program code is used to enable the terminal device to execute the steps of the path generation method for a two-wheeled electric vehicle provided in the first aspect.

[0059] The program code for executing the present invention may be written in any combination of one or more programming languages, and may be executed entirely on a user device, partially on a user device, as an independent software package, partially on a user device and partially on a remote device, or entirely on a remote device.

[0060] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0061] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A path generation method for a two-wheeled electric vehicle, characterized in that: include: Collecting data from the user terminal to obtain the user's actual driving data and destination information, wherein the destination information includes location information of at least one destination; The actual driving data is analyzed by a pre-trained trajectory generation model to obtain a plurality of candidate points; wherein the trajectory generation model is trained by a large amount of riding trajectory data of two-wheeled electric vehicles; Determine a priority destination from the destination information according to the current user's travel control strategy, perform distance analysis between multiple candidate points and the priority destination location information, and select the candidate point closest to the priority destination as the user's next movement point; The control amount is generated by the point to complete the information of the next moving point, obtain the predicted trajectory data and send it to the user.

2. The method according to claim 1, characterized in that The method further comprises: When the user drives according to the predicted trajectory data, data is collected in real time to obtain new actual driving data; Determine whether the user has reached all the destinations in the destination information. If so, determine that navigation is finished; if not, splice the predicted trajectory data and the new actual driving data and input them into the trajectory generation model to regenerate the predicted trajectory data of the next destination.

3. The method according to claim 1, characterized in that After obtaining multiple candidate points, the method further includes: The movement probability of each candidate point is calculated through the pre-trained trajectory generation model; A preset probability threshold is obtained to determine whether the movement probability is greater than the probability threshold. If so, the corresponding candidate point is retained; if not, the corresponding candidate point is deleted.

4. The method according to claim 1, characterized in that The determining a priority destination from the destination information according to the travel control strategy of the current user includes: Obtaining the current location of the user, calculating the distance between each destination in the destination information and the current location, sorting the distances in descending order, and taking the destination ranked first as a priority destination; Alternatively, the limited arrival time of each destination in the destination information is obtained, and the destinations are sorted in the order of the limited arrival times, and the destination ranked first is taken as the priority destination.

5. The method according to claim 1, characterized in that The training process of the trajectory generation model includes: Build a model framework and determine the training strategy to obtain the initial trajectory generation model; Acquire historical driving data of the user, and divide the historical driving data into a training set, a validation set, and a test set; The initial trajectory generation model is trained according to the training strategy and the training set, and the parameters of the initial trajectory generation model are adjusted through the validation set to obtain an intermediate trajectory generation model; The test set is used to evaluate the performance of the intermediate trajectory generation model, and when the evaluation result meets the preset user requirements, the trajectory generation model is obtained.

6. The method according to claim 5, characterized in that The model framework includes: A data embedding layer is used to map multi-dimensional actual driving data or historical driving data into a high-dimensional feature vector; A position encoding module is used to sequentially label the high-dimensional feature vector in the length direction of the trajectory sequence; Multiple encoding modules are used to perform in-depth trajectory feature mining on high-dimensional feature vectors containing sequential tags to obtain multi-dimensional feature data; A linear layer, used for performing data conversion according to the multi-dimensional characterization to obtain relative motion positions of multiple candidate points; The probability distribution calculation module is used to calculate the movement probability of the user to the candidate point according to the relative movement position.

7. The method according to claim 6, characterized in that The encoding module includes a first normalization layer, a multi-head attention layer, a second normalization layer and a feedforward layer, wherein: Normalizing the high-dimensional feature vector by the first normalization layer to obtain first feature data; Capturing the dependency relationship between the first feature data by the multi-head attention layer; The dependency relationship is merged with the first feature data to obtain second feature data, and the second feature data is normalized by the second normalization layer to obtain third feature data; Performing further feature extraction on the third feature data through the feedforward layer to obtain fourth feature data; The second feature data and the fourth feature data are fused to obtain the multi-dimensional feature data.

8. The method according to claim 1, characterized in that The actual driving data includes track information of multiple points and electric two-wheeled vehicle information; The trajectory information includes the generation time, distance and geocode of each point; the electric two-wheeled vehicle includes battery type, battery health, battery charge and battery power.

9. An electronic device, characterized in that: The invention comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the path generation method for a two-wheeled electric vehicle as claimed in any one of claims 1 to 8 when executing the computer program.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, a path generation method for a two-wheeled electric vehicle according to any one of claims 1 to 8 is implemented.

Images