Vehicle control instruction recommendation method, server and computer readable storage medium

Abstract

The invention discloses a vehicle control instruction recommendation method, a server and a computer readable storage medium. The vehicle control instruction recommendation method comprises the steps of obtaining current event information of a target vehicle and user identification information of the target vehicle; based on a pre-configured event-vehicle control instruction database, a first vehicle control instruction set is determined according to the current event information and the user identification information, and the first vehicle control instruction set is configured as a vehicle control instruction which is recommended to the target vehicle and possibly executed by the target vehicle under the current event. Thus, the server can intelligently provide the vehicle control instruction suggestions so that the user can determine the vehicle control instruction needing to be executed, the user does not need to memorize various vehicle control instructions, and therefore the intelligent level of the vehicle is enhanced. Moreover, compared with vehicle control instruction recommendation based on previous operation of the user, richer vehicle control instructions can be recommended to the user based on the pre-configured event-vehicle control instruction database.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and more specifically, to a vehicle control instruction recommendation method, a server, and a computer-readable storage medium. Background Art

[0002] The vehicle system can usually only control the vehicle to perform corresponding operations based on the clear instructions given by the user. However, in this case, the user needs to memorize various vehicle control instructions and judge the vehicle control instructions to be used in different scenarios. The intelligence is poor and the user experience is poor. Summary of the invention

[0003] The present application provides a vehicle control instruction recommendation method, a server and a computer-readable storage medium.

[0004] The present application provides a vehicle control instruction recommendation method, the method comprising:

[0005] Acquire current event information of a target vehicle and user identification information of the target vehicle;

[0006] Based on a preconfigured event-vehicle control instruction database, a first vehicle control instruction set is determined according to the current event information and the user identification information, and the first vehicle control instruction set is configured to be recommended to the target vehicle as vehicle control instructions that the target vehicle may execute under the current event.

[0007] In this way, in order to provide users with a better driving experience, the server obtains the current event information of the target vehicle and the user identification information of the target vehicle. Then, based on the preconfigured event-vehicle control instruction database, the server analyzes the acquired current event information and user identification information to obtain the first vehicle control instruction set that the target vehicle may execute under the current event, and recommends the first vehicle control instruction set to the target vehicle. In this way, the server can intelligently provide vehicle control instruction suggestions for users to decide which vehicle control instructions need to be executed, without the need for users to memorize various vehicle control instructions, thereby enhancing the intelligence level of the vehicle. Moreover, compared with recommending vehicle control instructions based on previous user operations, the preconfigured event-vehicle control instruction database can recommend more abundant vehicle control instructions to users.

[0008] In some embodiments, the determining the first vehicle control instruction set based on the preconfigured event-vehicle control instruction database according to the current event information and the user identification information includes:

[0009] Acquire a first matrix corresponding to the user identification information from the event-vehicle control instruction database according to the user identification information, wherein the first matrix is ​​an event-vehicle control instruction probability matrix of the target vehicle;

[0010] Based on the event-vehicle control instruction database, determining a similar matrix set according to the first matrix, the similar matrix set including at least one event-operation probability matrix similar to the first matrix;

[0011] The first vehicle control instruction set is determined according to the current event information and the similarity matrix set.

[0012] In this way, the server obtains the first matrix corresponding to the user identification information from the event-vehicle control instruction database according to the user identification information, and the first matrix is ​​the event-vehicle control instruction probability matrix of the target vehicle. Then, based on the event-vehicle control instruction database, the server determines a similar matrix set according to the first matrix, and the similar matrix set includes at least one event-operation probability matrix similar to the first matrix. Finally, the server determines the first vehicle control instruction set according to the current event information and the similar matrix set. In this way, the server can analyze the first matrix obtained according to the user identification information, and determine the similar matrix set similar to the first matrix from the event-vehicle control instruction database, and then determine the first vehicle control instruction set to be recommended according to the event executed by the current target vehicle and the similar matrix set, so as to provide intelligent, rich and efficient vehicle control instruction suggestions.

[0013] In some embodiments, the event-vehicle control instruction database includes a plurality of second matrices, the second matrix is ​​an event-vehicle control instruction probability matrix corresponding to a non-target vehicle, and the similarity matrix set is determined based on the event-vehicle control instruction database and the first matrix, including:

[0014] respectively solving the similarity between the first matrix and each of the second matrices;

[0015] A predetermined number of the second matrices is determined according to the similarity to determine the similarity matrix set.

[0016] In this way, the server solves the similarity of the first matrix and each second matrix respectively, and the second matrix is ​​the event-vehicle control instruction probability matrix corresponding to each non-target vehicle in the event-vehicle control instruction database. Then, the server determines a predetermined number of second matrices based on the similarity to determine a similar matrix set. In this way, by calculating and comparing the similarity between the first matrix and each second matrix, other user groups with similar behavior patterns to the users on the target vehicle are found, so that in the subsequent process, based on the similar matrix sets corresponding to these user groups, intelligent, rich and efficient vehicle control instruction recommendations are provided to users.

[0017] In some embodiments, determining the first vehicle control instruction set according to the current event information and the similarity matrix set includes:

[0018] Determine a first recommended vehicle control instruction set according to the current event information and the similarity matrix set;

[0019] The first vehicle control instruction set is determined based on the first recommended vehicle control instruction set.

[0020] In this way, the server can determine the first recommended vehicle control instruction set that all target vehicles may execute under the current event from the similarity matrix set according to the current event information. Then the server determines the first vehicle control instruction set according to the first recommended vehicle control instruction set. In this way, the first recommended vehicle control instruction set can be obtained according to the current event information and the similarity matrix set, and the first vehicle control instruction set is determined according to the first recommended vehicle control instruction set, thereby providing users with intelligent, rich and efficient vehicle control instruction suggestions.

[0021] In some embodiments, determining the first vehicle control instruction set according to the first recommended vehicle control instruction set includes:

[0022] In a current event, if the target vehicle does not execute the vehicle control instructions in the first recommended vehicle control instruction set, the first vehicle control instruction set is determined according to the first recommended vehicle control instruction set.

[0023] Thus, in the current event, if the target vehicle does not execute the vehicle control instructions in the first recommended vehicle control instruction set, the server determines the first vehicle control instruction set according to the first recommended vehicle control instruction set. Thus, in the current event, if the target vehicle does not execute the vehicle control instructions in the first recommended vehicle control instruction set, the server does not perform other operations on the first recommended vehicle control instruction set, but directly determines the first vehicle control instruction set according to the first recommended vehicle control instruction set, thereby reducing the computing burden of the server, reducing the recommendation delay, and improving the response speed of the system.

[0024] In some embodiments, determining the first vehicle control instruction set according to the first recommended vehicle control instruction set includes:

[0025] Under the current event, if the target vehicle executes at least one second vehicle control instruction in the first recommended vehicle control instruction set, the second vehicle control instruction is removed from the first recommended vehicle control instruction set to obtain a second recommended vehicle control instruction set;

[0026] The first vehicle control instruction set is determined based on the second recommended vehicle control instruction set.

[0027] Thus, in the current event, if the target vehicle executes at least one second vehicle control instruction in the first recommended vehicle control instruction set, the server removes the second vehicle control instruction from the first recommended vehicle control instruction set to obtain the second recommended vehicle control instruction set. Then, the server determines the first vehicle control instruction set based on the second recommended vehicle control instruction set. In this way, updating the recommended vehicle control instruction set based on the real-time behavior of the target vehicle helps to improve the accuracy and adaptability of the vehicle control instruction recommendation.

[0028] In some embodiments, determining the first vehicle control instruction set according to the first recommended vehicle control instruction set includes:

[0029] determining a third recommended vehicle control instruction set according to the first recommended vehicle control instruction set based on a preset instruction quantity and a preset instruction probability;

[0030] Acquire environmental information of the environment where the target vehicle is located;

[0031] Based on a pre-trained large language model, the first vehicle control instruction set is determined according to the environmental information and the third recommended vehicle control instruction set.

[0032] In this way, based on the preset number of instructions and the preset instruction probability, the server determines the third recommended vehicle control instruction set according to the first recommended vehicle control instruction set. Next, the server obtains environmental information of the environment where the target vehicle is located. Finally, the server determines the first vehicle control instruction set based on the pre-trained large language model and the environmental information and the third recommended vehicle control instruction set. In this way, by using the large language model and real-time environmental information, intelligent, rich and safe vehicle control instruction suggestions are provided for the target vehicle, which helps to improve the driving experience, reduce the decision-making burden of the user, and thus improve driving safety.

[0033] In some embodiments, the event-vehicle control command probability matrix database includes a plurality of event-vehicle control command probability matrices, each of which is associated with a vehicle, and the event-vehicle control command probability matrix is ​​constructed by the following steps:

[0034] defining an event set, wherein the event set includes a plurality of events;

[0035] defining a vehicle control instruction set, wherein the vehicle control instruction set includes a plurality of vehicle control instructions;

[0036] Determine a first occurrence number of each event and a second occurrence number of each vehicle control instruction under each event within a preset statistical period;

[0037] According to the first number of occurrences and the second number of occurrences, the occurrence probability of each of the vehicle control instructions under each of the events is determined to obtain an event-vehicle control instruction probability matrix.

[0038] In this way, the event-vehicle control instruction probability matrix database includes multiple event-vehicle control instruction probability matrices, wherein each event-vehicle control instruction probability matrix is ​​associated with a vehicle. The server defines a complete set of events, which includes multiple events. Next, the server defines a complete set of vehicle control instructions, which includes multiple vehicle control instructions. Then, the server determines the first number of occurrences of each event and the second number of occurrences of each vehicle control instruction under each event within a preset statistical period. The server then determines the probability of occurrence of each vehicle control instruction under each event based on the first number of occurrences and the second number of occurrences to obtain the event-vehicle control instruction probability matrix. In this way, by performing offline analysis and mining of user logs of multiple users, there is no need for a large number of high-quality data sets and a large amount of resources for fine-tuning training, thereby reducing computing costs and time. In addition, by counting and analyzing historical data, the server can more accurately predict and recommend vehicle control instructions, improve the accuracy of recommendations, and thus improve the user's driving experience.

[0039] An embodiment of the present application provides a server, the server comprising a memory and a processor, the memory storing a computer program, and implementing the above method when the computer program is executed by the processor.

[0040] An embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium memory stores a computer program, and when the computer program is executed by one or more processors, the above method is implemented.

[0041] Additional aspects and advantages of the embodiments of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the embodiments of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0043] Figure 1 It is one of the flowcharts of the communication method of certain embodiments of the present application;

[0044] Figure 2 This is a second flow chart of a communication method according to some embodiments of the present application;

[0045] Figure 3 This is a third flow chart of a communication method according to some embodiments of the present application;

[0046] Figure 4 This is a fourth flow chart of a communication method according to some embodiments of the present application;

[0047] Figure 5 This is a fifth flow chart of a communication method according to some embodiments of the present application;

[0048] Figure 6 This is the sixth flow chart of the communication method of certain embodiments of the present application;

[0049] Figure 7 This is the seventh flow chart of the communication method of certain implementation modes of the present application;

[0050] Figure 8 This is the eighth flow chart of the communication method of certain embodiments of the present application. DETAILED DESCRIPTION

[0051] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals represent the same or similar elements or elements having the same or similar functions from beginning to end. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the embodiments of the present application, and cannot be understood as limiting the embodiments of the present application.

[0052] The development of vehicle-mounted systems has evolved from simple mechanical control to complex electronic control and then to today's intelligent control. However, in the relevant technologies, intelligent vehicle-mounted systems can usually only control the vehicle to perform corresponding operations based on clear instructions given by the user, which has some obvious limitations.

[0053] For example, users need to memorize various vehicle control commands, which is particularly evident in complex vehicle control systems. As the number of vehicle functions increases, the burden on users to remember and understand these commands becomes heavier, resulting in poor user experience.

[0054] In addition, in different driving scenarios, users need to judge and select appropriate vehicle control instructions by themselves, which is less intelligent. This not only requires users to have certain driving experience, but also requires in-depth understanding of vehicle performance and driving environment, which increases the complexity and difficulty of driving.

[0055] Based on the above questions, please refer to Figure 1 The present application embodiment provides a vehicle control instruction recommendation method, the method comprising:

[0056] 01: Obtain the current event information of the target vehicle and the user identification information of the target vehicle;

[0057] 02: Based on the preconfigured event-vehicle control instruction database, determine the first vehicle control instruction set according to the current event information and user identification information.

[0058] The embodiment of the present application also provides a server, including a memory and a processor. The resource access method of the embodiment of the present application can be implemented by the server of the embodiment of the present application. Specifically, a computer program is stored in the memory, and the processor is used to obtain the current event information of the target vehicle and the user identification information of the target vehicle. And based on the preconfigured event-vehicle control instruction database, the first vehicle control instruction set is determined according to the current event information and the user identification information.

[0059] The embodiment of the present application also provides a vehicle control instruction recommendation device. The resource access method of the embodiment of the present application can be implemented by the vehicle control instruction recommendation device of the embodiment of the present application. Specifically, the vehicle control instruction recommendation device includes an acquisition module and a determination module. The acquisition module is used to acquire the current event information of the target vehicle and the user identification information of the target vehicle. The determination module is used to determine the first vehicle control instruction set according to the current event information and the user identification information based on the preconfigured event-vehicle control instruction database.

[0060] Specifically, the current event information can be regarded as the event currently executed by the target vehicle.

[0061] User identification information can be regarded as data information that can uniquely identify the user and is stored in the event-vehicle control instruction database. Based on the user identification information, the corresponding event-vehicle control instruction probability matrix can be found from the event-vehicle control instruction database, such as the user's registered account.

[0062] The event-vehicle control instruction database can be regarded as a database storing multiple event-vehicle control instruction probability matrices, and the corresponding event-vehicle control instruction probability matrix can be found from the event-vehicle control instruction database according to the user identification information.

[0063] The first vehicle control instruction set may be regarded as a vehicle control instruction recommended to the target vehicle and possibly executed by the target vehicle under the current event.

[0064] Specifically, in order to provide users with a better driving experience, the server obtains the current event information of the target vehicle and the user identification information of the target vehicle. Then, based on the preconfigured event-vehicle control instruction database, the server analyzes the obtained current event information and user identification information to obtain the first vehicle control instruction set that the target vehicle may execute under the current event, and recommends the first vehicle control instruction set to the target vehicle.

[0065] The following is an example to illustrate the method of the implementation method of the present application. In the embodiment of the present application, vehicle A is used as the target vehicle. "The driver gets on the car" is used as the current event information. User identification A1 is used as user identification information. The vehicle-mounted system is used as the server.

[0066] It should be noted that vehicle A may be used by different users, and the driver may be user A1 or user A2, which is not limited here. When user A2 drives vehicle A but logs in the user ID of user A1, the vehicle system obtains user ID A1. That is, the user ID information obtained by the vehicle system is consistent with the user account logged in. In the implementation of the present application, vehicle A logs in user ID A1.

[0067] When the driver opens the main driver's door of vehicle A and gets in, the vehicle system can immediately obtain the event information executed by vehicle A, that is, "the main driver gets in the car", and obtain the user ID A1 currently logged in to vehicle A. Then, based on the pre-configured event-vehicle control instruction database, the vehicle system analyzes the obtained "the main driver gets in the car" information and user ID A1, thereby obtaining the first vehicle control instruction set K that vehicle A may execute after the execution of "the main driver gets in the car", and recommends the first vehicle control instruction set K to vehicle A.

[0068] In summary, in the vehicle control instruction recommendation method and server of the implementation mode of the present application, in order to provide users with a better driving experience, the server obtains the current event information of the target vehicle and the user identification information of the target vehicle. Then, based on the preconfigured event-vehicle control instruction database, the server analyzes the acquired current event information and user identification information to obtain the first vehicle control instruction set that the target vehicle may execute under the current event, and recommends the first vehicle control instruction set to the target vehicle. In this way, the server can intelligently provide vehicle control instruction suggestions for users to decide the vehicle control instructions that need to be executed, without the need for users to memorize various vehicle control instructions, thereby enhancing the intelligence level of the vehicle. In addition, compared with recommending vehicle control instructions based on previous user operations, the preconfigured event-vehicle control instruction database can recommend more abundant vehicle control instructions to users.

[0069] See also Figure 2 In some embodiments, step 02 (determining a first vehicle control instruction set based on a preconfigured event-vehicle control instruction database according to current event information and user identification information) includes:

[0070] 021: acquiring a first matrix corresponding to the user identification information from an event-vehicle control instruction database according to the user identification information;

[0071] 022: Based on the event-vehicle control instruction database, determine a similarity matrix set according to the first matrix;

[0072] 023: Determine a first vehicle control instruction set according to current event information and a similarity matrix set.

[0073] In some embodiments, the acquisition module is further configured to acquire a first matrix corresponding to the user identification information from the event-vehicle control instruction database according to the user identification information. The determination module is further configured to determine a similar matrix set according to the first matrix based on the event-vehicle control instruction database. The determination module is further configured to determine the first vehicle control instruction set according to the current event information and the similar matrix set.

[0074] In some embodiments, the processor is further configured to obtain a first matrix corresponding to the user identification information from an event-vehicle control instruction database according to the user identification information, and determine a similar matrix set according to the first matrix based on the event-vehicle control instruction database, and determine a first vehicle control instruction set according to the current event information and the similar matrix set.

[0075] Specifically, the first matrix can be regarded as an event-vehicle control command probability matrix of the target vehicle.

[0076] The similar matrix set includes at least one event-action probability matrix similar to the first matrix.

[0077] The server obtains a first matrix corresponding to the user identification information from the event-vehicle control instruction database according to the user identification information. Then, the server determines a similar matrix set according to the first matrix based on the event-vehicle control instruction database. Finally, the server determines a first vehicle control instruction set according to the current event information and the similar matrix set.

[0078] Continuing with the above example, the first matrix can be regarded as an event-vehicle control command probability matrix obtained based on the previous operating behavior habits of user A1. In the implementation manner of the present application, the first matrix serves as the event-vehicle control command probability matrix of the target vehicle.

[0079] After obtaining the user identification A1, the vehicle-mounted system obtains a first matrix corresponding to the user identification A1 from the event-vehicle control instruction database according to the user identification A1.

[0080] Please refer to Table 1. The first matrix is ​​shown in Table 1. The first column in the table represents the event that occurred. The first column in the table represents the vehicle control command executed after the event occurs. The numbers in the table represent the probability of executing different vehicle control commands after each event occurs. It should be noted that if there is no special explanation in the table in the implementation mode of this application, the meaning of each part in the table is the same as above.

[0081] Table 1 is as follows:

[0082] Table 1

[0083]

[0084] Next, after acquiring the first matrix, the vehicle-mounted system determines a similar matrix set B according to the first matrix, where the similar matrix set B includes a plurality of event-vehicle control command probability matrices that are similar to the first matrix.

[0085] Finally, the vehicle system determines the first vehicle control instruction set K based on the acquired "driver getting on the vehicle" event information and the similarity matrix set B.

[0086] In this way, the vehicle-mounted system can analyze the first matrix obtained according to the user identifier A1, and determine the similarity matrix set B similar to the first matrix from the event-vehicle control instruction database, and then determine the recommended first vehicle control instruction set K based on the "driver getting on the vehicle" event executed by the current vehicle A and the similarity matrix set B, so as to provide intelligent, rich and efficient vehicle control instruction recommendations.

[0087] See also Figure 3In some embodiments, the event-vehicle control instruction database includes a plurality of second matrices, the second matrix being an event-vehicle control instruction probability matrix corresponding to the non-target vehicle, and step 022 (determining a similarity matrix set according to the first matrix based on the event-vehicle control instruction database) includes:

[0088] 0221: Solve the similarity between the first matrix and each second matrix respectively;

[0089] 0222: Determine a predetermined number of second matrices according to the similarity to determine a similar matrix set.

[0090] In some embodiments, the determination module includes a calculation submodule. The calculation submodule is used to solve the similarity between the first matrix and each second matrix respectively. The determination module is also used to determine a predetermined number of second matrices according to the similarity to determine a similar matrix set.

[0091] In some embodiments, the processor is further configured to respectively calculate the similarity between the first matrix and each second matrix, and determine a predetermined number of second matrices according to the similarity to determine a similar matrix set.

[0092] Specifically, the event-vehicle control instruction database includes a plurality of second matrices, and the second matrix is ​​an event-vehicle control instruction probability matrix corresponding to the non-target vehicle.

[0093] The server solves the similarity of the first matrix and each second matrix respectively, and the second matrix is ​​an event-vehicle control command probability matrix corresponding to the non-target vehicle in the event-vehicle control command database.

[0094] Next, the server determines a predetermined number of second matrices according to the similarity to determine a similar matrix set.

[0095] Continuing with the above example, the second matrix can be regarded as an event-vehicle control instruction probability matrix corresponding to other users except the user identifier A1 in the event-vehicle control instruction database. The preset number is 3.

[0096] The vehicle-mounted system uses the first matrix and each second matrix to perform calculations to obtain multiple similarities. In some embodiments, the method of similarity calculation includes cosine similarity, Euclidean distance, Pearson correlation coefficient, etc. The embodiment of the present application uses cosine similarity, and the solution steps are as follows: First, convert the matrix into a column vector, that is, treat each column of the matrix as a vector. Next, calculate the dot product between the column vectors corresponding to the two matrices, and the dot product is the sum of the products of the corresponding elements of the two vectors. Then, calculate the Euclidean norm, that is, the length, of the column vectors of the two matrices respectively. Finally, use the following formula to calculate the cosine similarity

[0097]

[0098] Where X·Y is the dot product of the corresponding column vectors of the two matrices, and ||X|| and ||Y|| are the norms of the column vectors of the two matrices, respectively.

[0099] Next, the vehicle-mounted system compares the obtained multiple similarities, determines the three second matrices with the highest similarities, and obtains a similarity matrix set B.

[0100] Please refer to Tables 2-4, Table 2 is the similarity matrix B1 in the similarity matrix set, Table 3 is the similarity matrix B2 in the similarity matrix set, and Table 4 is the similarity matrix B3 in the similarity matrix set.

[0101] Table 2 is as follows:

[0102] Table 2

[0103]

[0104] Table 3 is as follows:

[0105] Table 3

[0106]

[0107] Table 4 is as follows:

[0108] Table 4

[0109]

[0110] In this way, by calculating and comparing the similarity between the first matrix and each second matrix, other user groups with similar behavior patterns to user A1 are found, so that in the subsequent process, based on the similarity matrix set B corresponding to these user groups, intelligent, rich and efficient vehicle control instruction recommendations can be provided to user A1.

[0111] See also Figure 4 In some embodiments, step 23 (determining a first vehicle control instruction set based on current event information and a similarity matrix set) includes:

[0112] 0231: Determine a first recommended vehicle control instruction set according to current event information and a similarity matrix set;

[0113] 0232: Determine a first vehicle control instruction set based on a first recommended vehicle control instruction set.

[0114] In some embodiments, the determination module is further configured to determine a first recommended vehicle control instruction set based on the current event information and the similarity matrix set. The determination module is further configured to determine the first vehicle control instruction set based on the first recommended vehicle control instruction set.

[0115] In some embodiments, the processor is further configured to determine a first recommended vehicle control instruction set based on the current event information and the similarity matrix set, and determine a first vehicle control instruction set based on the first recommended vehicle control instruction set.

[0116] Specifically, the server can determine, from a similarity matrix set according to current event information, all first recommended vehicle control instruction sets that may be executed by the target vehicle under the current event.

[0117] Next, the server determines the first vehicle control instruction set according to the first recommended vehicle control instruction set.

[0118] Continuing with the above example, after the vehicle-mounted system obtains the similarity matrix set B, it determines the first recommended vehicle control instruction set C from the similarity matrix based on the "driver gets on the vehicle" event information. The first recommended vehicle control instruction set C includes the first recommended vehicle control instruction C1, the first recommended vehicle control instruction C2 and the first recommended vehicle control instruction C3.

[0119] The first recommended vehicle control instruction C1 is part of the content of the similarity matrix B1, the first recommended vehicle control instruction C2 is part of the content of the similarity matrix B2, and the first recommended vehicle control instruction C3 is part of the content of the similarity matrix B3.

[0120] Please refer to Tables 5-7, Table 5 is the first recommended vehicle control instruction C1, Table 6 is the first recommended vehicle control instruction C2, and Table 7 is the first recommended vehicle control instruction C3.

[0121] Table 5 is as follows:

[0122] Table 5

[0123]

[0124] Table 6 is as follows:

[0125] Table 6

[0126]

[0127] Table 7 is as follows:

[0128] Table 7

[0129]

[0130] Next, the vehicle-mounted system determines the first vehicle control instruction set K according to the first recommended vehicle control instruction set C obtained above.

[0131] In this way, the vehicle-mounted system can obtain the first recommended vehicle control instruction set C based on the "driver getting on the vehicle" event information and the similarity matrix set B, and determine the first vehicle control instruction set K based on the first recommended vehicle control instruction set C, thereby providing users with intelligent, rich and efficient vehicle control instruction recommendations.

[0132] See also Figure 5 In some embodiments, step 0232 (determining a first vehicle control instruction set according to a first recommended vehicle control instruction set) includes:

[0133] 02321: Under the current event, if the target vehicle does not execute the vehicle control instructions in the first recommended vehicle control instruction set, determine the first vehicle control instruction set according to the first recommended vehicle control instruction set.

[0134] In some embodiments, the determination module is further configured to determine the first vehicle control instruction set according to the first recommended vehicle control instruction set if the target vehicle does not execute the vehicle control instructions in the first recommended vehicle control instruction set in the current event.

[0135] In certain embodiments, the processor is further configured to determine the first vehicle control instruction set based on the first recommended vehicle control instruction set if the target vehicle does not execute the vehicle control instructions in the first recommended vehicle control instruction set in the current event.

[0136] Specifically, in the current event, if the target vehicle does not execute the vehicle control instructions in the first recommended vehicle control instruction set, the server determines the first vehicle control instruction set according to the first recommended vehicle control instruction set.

[0137] Continuing with the above example, in the current event, if vehicle A does not execute the vehicle control instructions in the first recommended vehicle control instruction set C, that is, vehicle A does not execute the above-mentioned vehicle control instructions such as "turn on the air conditioner", "turn off the air conditioner" and "driving mode: comfort", then the vehicle system will not perform any operations on the first recommended vehicle control instruction set C, and directly determine the first vehicle control instruction set K based on the first recommended vehicle control instruction set C.

[0138] In this way, under the current event, when the target vehicle does not execute the vehicle control instructions in the first recommended vehicle control instruction set, the server does not perform other operations on the first recommended vehicle control instruction set, but directly determines the first vehicle control instruction set based on the first recommended vehicle control instruction set, thereby reducing the computing burden of the server, reducing the recommendation delay, and improving the response speed of the system.

[0139] See also Figure 6 In some embodiments, step 0232 (determining a first vehicle control instruction set according to a first recommended vehicle control instruction set) includes:

[0140] 02322: Under the current event, if the target vehicle executes at least one second vehicle control instruction in the first recommended vehicle control instruction set, remove the second vehicle control instruction from the first recommended vehicle control instruction set to obtain a second recommended vehicle control instruction set;

[0141] 02323: Determine a first vehicle control instruction set based on a second recommended vehicle control instruction set.

[0142] In some embodiments, the determination module includes a removal submodule. The removal submodule is used to remove the second vehicle control instruction from the first recommended vehicle control instruction set to obtain a second recommended vehicle control instruction set if the target vehicle executes at least one second vehicle control instruction in the first recommended vehicle control instruction set under the current event. The determination module is also used to determine the first vehicle control instruction set based on the second recommended vehicle control instruction set.

[0143] In some embodiments, the processor is further configured to, in the current event, if the target vehicle executes at least one second vehicle control instruction in the first recommended vehicle control instruction set, remove the second vehicle control instruction from the first recommended vehicle control instruction set to obtain a second recommended vehicle control instruction set, and determine the first vehicle control instruction set based on the second recommended vehicle control instruction set.

[0144] Specifically, the second vehicle control instruction may be regarded as a vehicle control instruction that has been executed in the first recommended vehicle control instruction set.

[0145] In the current event, if the target vehicle executes at least one second vehicle control instruction in the first recommended vehicle control instruction set, the server removes the second vehicle control instruction from the first recommended vehicle control instruction set to obtain a second recommended vehicle control instruction set.

[0146] Next, the server determines the first vehicle control instruction set according to the second recommended vehicle control instruction set. In some embodiments, after obtaining the final first vehicle control instruction set, the second vehicle control instructions that have been executed in the first vehicle control instruction set may be removed.

[0147] Continuing with the above example, in the current event, if vehicle A executes at least one vehicle control instruction in the first recommended vehicle control instruction set C, that is, vehicle A executes at least one of the above-mentioned vehicle control instructions such as "turn on the air conditioner", "turn off the air conditioner" and "driving mode: comfort".

[0148] In the example given in the implementation mode of the present application, it is taken that vehicle A executes the vehicle control instruction of "driving mode: comfort".

[0149] In the current event, after vehicle A executes the vehicle control instruction "Driving Mode: Comfort", the vehicle system will remove the "Driving Mode: Comfort" vehicle control instruction from the first recommended vehicle control instruction set, and obtain the second recommended vehicle control instruction set D. The second recommended vehicle control instruction set D includes the second recommended vehicle control instruction D1, the second recommended vehicle control instruction D2 and the second recommended vehicle control instruction D3.

[0150] The second recommended vehicle control instruction D1 is obtained based on the first recommended vehicle control instruction C1 , the second recommended vehicle control instruction D2 is obtained based on the first recommended vehicle control instruction C2 , and the second recommended vehicle control instruction D3 is obtained based on the first recommended vehicle control instruction C3 .

[0151] Please refer to Tables 8-10, Table 8 is the second recommended vehicle control instruction D1, Table 9 is the second recommended vehicle control instruction D2, and Table 10 is the second recommended vehicle control instruction D3.

[0152] Table 8 is as follows:

[0153] Table 8

[0154]

[0155] Table 9 is as follows:

[0156] Table 9

[0157]

[0158] Table 10 is as follows:

[0159] Table 10

[0160]

[0161] After obtaining the second recommended vehicle control instruction set D, the vehicle-mounted system uses the second recommended vehicle control instruction set D as a new first recommended vehicle control instruction set C to determine the first vehicle control instruction set K.

[0162] In this way, updating the recommended vehicle control instruction set according to the real-time behavior of vehicle A helps to improve the accuracy and adaptability of vehicle control instruction recommendations.

[0163] See also Figure 7 In some embodiments, step 0232 (determining a first vehicle control instruction set according to a first recommended vehicle control instruction set) includes:

[0164] 02324: Determine a third recommended vehicle control instruction set according to the first recommended vehicle control instruction set based on the preset instruction quantity and the preset instruction probability;

[0165] 02325: Obtain environmental information of the target vehicle's environment;

[0166] 02326: Based on a pre-trained large language model, determine a first vehicle control instruction set according to environmental information and a third recommended vehicle control instruction set.

[0167] In some embodiments, the determination module is further configured to determine a third recommended vehicle control instruction set according to the first recommended vehicle control instruction set based on a preset instruction quantity and a preset instruction probability. The acquisition module is further configured to acquire environmental information of an environment where the target vehicle is located. The determination module is further configured to determine the first vehicle control instruction set according to the environmental information and the third recommended vehicle control instruction set based on a pre-trained large language model.

[0168] In some embodiments, the processor is further configured to determine a third recommended vehicle control instruction set based on the first recommended vehicle control instruction set based on the preset instruction quantity and the preset instruction probability, and obtain environmental information of the environment where the target vehicle is located, and determine the first vehicle control instruction set based on the environmental information and the third recommended vehicle control instruction set based on the pre-trained large language model.

[0169] Specifically, the preset number of instructions may be regarded as the maximum number of vehicle control instructions in the obtained first vehicle control instruction set.

[0170] The preset instruction probability may be regarded as the minimum probability of the vehicle control instructions in the obtained first vehicle control instruction set.

[0171] The pre-trained large language model can be regarded as a model specially trained to perform a specific task. In other words, the large language model has stronger processing capabilities and better processing effects in building customized solutions in the direction of this specific task compared to other general models or basic models.

[0172] The reference model in the implementation mode of the present application can be obtained by preliminary training based on the basic model. It can be understood that the reference model has more knowledge in a specific field than the basic model. However, the ability to handle the tasks that need to be solved for this application is still lacking. Therefore, it is necessary to make adjustments based on the reference model to obtain a large language model that can meet the task processing requirements.

[0173] Specifically, training data is constructed based on the collected voice request samples and the vehicle system function customization solution samples constructed based on the voice request samples.

[0174] Sample data can come from collected user logs, such as the user's history of successfully building a customized solution for in-vehicle system functions based on voice requests. Of course, training data can also be built as needed. For example, a set of training data can be:

[0175] Event: Driver gets on the car

[0176] Optional operations: Turn on the air conditioner, turn on the driver's seat ventilation, turn on the driver's seat heating

[0177] Environmental conditions: In-car temperature: 40°C; Out-car temperature: 35°C; Air conditioning: Off; Driver seat ventilation: Off; Driver seat heating: Off

[0178] Output: {"Vehicle control command":["Air conditioning: on","Seat ventilation: on"]}

[0179] Furthermore, in order to guide the large model to meet the task requirements, the prompt information template needs to be configured. Specifically, the prompt information template is used to guide the model to generate output, providing the background of the problem and relevant guidance. It usually includes a description and instructions of the problem to tell the model how to process the input data. The prompt information template can provide additional auxiliary information to facilitate the model to better understand and generate output. It can also include a specific format, style, theme that the model is required to follow, or instruct the model to pay attention to certain specific aspects and details. As a result, the model will understand the requirements of the task based on the content of the prompt information template, thereby generating answers and texts that meet expectations. Generally, the configuration process of the prompt information template may include clarifying the role that the model needs to play (optional), clarifying the input text content and structure, and clarifying the output text content and structure.

[0180] In an example, the content of the prompt message template is as follows:

[0181] Suppose you are an in-car voice assistant named "Xiao P". Please choose specific actions based on the events that occurred and the current status of the device.

[0182] Example:

[0183] Event: Driver gets on the car

[0184] Optional operations: Turn on the air conditioner, turn on the driver's seat ventilation, turn on the driver's seat heating

[0185] Environmental conditions: In-car temperature: 40°C; Out-car temperature: 35°C; Air conditioning: Off; Driver seat ventilation: Off; Driver seat heating: Off

[0186] Output: {"Vehicle control command":["Air conditioning: on","Seat ventilation: on"]}

[0187] Furthermore, the reference model is subjected to supervised fine tuning (SFT) using the prompt information template and training data configured as above, so that the reference model has corresponding task processing capabilities.

[0188] Specifically, the implementation method of the present application is based on the acquired current event information and user identification information, the vehicle system, and the configured prompt information template, so that the reference model can have the vehicle control instruction recommendation function after adjustment.

[0189] It should be noted that the above prompt information template is only for illustrative purposes.

[0190] Based on the preset instruction quantity and the preset instruction probability, the server determines the second recommended vehicle control instruction according to the first recommended vehicle control instruction set. In some embodiments, if the second vehicle control instruction is removed from the first recommended vehicle control instruction set to obtain the second recommended vehicle control instruction set, the server determines the second recommended vehicle control instruction according to the second recommended vehicle control instruction set.

[0191] Next, the server obtains environmental information of the environment where the target vehicle is located.

[0192] Finally, the server determines a first vehicle control instruction set based on the pre-trained large language model and according to the environmental information and the second recommended vehicle control instruction.

[0193] Continuing with the above example, the preset number of instructions is 3 and the preset instruction probability is 0.7.

[0194] In the example given in the implementation mode of the present application, it is taken as an example that vehicle A does not execute the vehicle control instructions in the first recommended vehicle control instruction set C after the "driver gets on the vehicle" event.

[0195] Please refer to Tables 11-13. After obtaining the first recommended vehicle control instruction set C, the vehicle system processes the first recommended vehicle control instruction set C based on the preset instruction quantity and the preset instruction probability to obtain the third recommended vehicle control instruction set E.

[0196] After the vehicle system processes the first recommended vehicle control instruction C1, it obtains the third recommended vehicle control instruction E1, which includes "turn on the air conditioner", "driver's air outlet: open" and "all vehicle door locks: close", as shown in Table 11.

[0197] Table 11 is as follows:

[0198] Table 11

[0199]

[0200] After the vehicle-mounted system processes the first recommended vehicle control instruction C2, it obtains the third recommended vehicle control instruction E2, which includes "turn on the air conditioner", "driving mode: comfort" and "all vehicle door locks: closed", as shown in Table 12.

[0201] Table 12 is as follows:

[0202] Table 12

[0203]

[0204] After the vehicle system processes the first recommended vehicle control instruction C3, it obtains the third recommended vehicle control instruction E3, which includes "driving mode: comfort", "main driver air outlet: open" and "all vehicle door locks: closed", as shown in Table 13.

[0205] Table 13 is as follows:

[0206] Table 13

[0207]

[0208] In this way, the obtained third recommended vehicle control instruction set E includes "turn on the air conditioner", "driver's air outlet: open", "driving mode: comfort" and "all vehicle door locks: close".

[0209] Next, the vehicle system obtains environmental information of the environment in which vehicle A is located, including "inside temperature: 40", "outside temperature: 35", "air conditioning: off", "driving mode: standard", "main driver's air outlet: open" and "all door locks: open".

[0210] The vehicle-mounted system uses the pre-trained large language model to process the environmental information obtained above and the third recommended vehicle control instruction set E, and determines that the first vehicle control instruction set K is "turn on the air conditioner", "driving mode: comfort" and "all vehicle door locks: close". Subsequently, the vehicle-mounted system sends the first vehicle control instruction set K to the display screen of vehicle A for user A1 to select. After user A1 completes the selection, vehicle A1 executes the corresponding vehicle control instruction. In some embodiments, the vehicle-mounted system provides the first vehicle control instruction set K to user A1 in the form of voice broadcast, which is not limited here.

[0211] In this way, by using a large language model and real-time environmental information, intelligent, rich and safe vehicle control instruction suggestions are provided to vehicle A, which helps to improve the driving experience and reduce the decision-making burden of user A1, thereby improving driving safety.

[0212] See also Figure 8 In some embodiments, the event-vehicle control command probability matrix database includes multiple event-vehicle control command probability matrices, each event-vehicle control command probability matrix is ​​associated with a vehicle, and the event-vehicle control command probability matrix is ​​constructed by the following steps:

[0213] 03: Define the complete set of events;

[0214] 04: Define the complete set of vehicle control instructions;

[0215] 05: Determine the first occurrence number of each event and the second occurrence number of each vehicle control instruction under each event within a preset statistical period;

[0216] 06: According to the first number of occurrences and the second number of occurrences, the occurrence probability of each vehicle control instruction under each event is determined to obtain an event-vehicle control instruction probability matrix.

[0217] In some embodiments, the vehicle control instruction recommendation device further includes a definition module and a processing module. The definition module is used to define a full set of events. The definition module is also used to define a full set of vehicle control instructions. The determination module is also used to determine a first occurrence number of each event and a second occurrence number of each vehicle control instruction under each event within a preset statistical period. The processing module is used to determine the occurrence probability of each vehicle control instruction under each event based on the first occurrence number and the second occurrence number, so as to obtain an event-vehicle control instruction probability matrix.

[0218] In some embodiments, the processor is further configured to define a complete set of events. And define a complete set of vehicle control instructions. The processor is further configured to determine a first occurrence number of each event and a second occurrence number of each vehicle control instruction under each event within a preset statistical period. And determine the occurrence probability of each vehicle control instruction under each event based on the first occurrence number and the second occurrence number, so as to obtain an event-vehicle control instruction probability matrix.

[0219] Specifically, the event-vehicle control command probability matrix database includes a plurality of event-vehicle control command probability matrices, wherein each event-vehicle control command probability matrix is ​​associated with a vehicle.

[0220] The preset statistical period can be regarded as a time range used by the server for analyzing and collecting statistical data when constructing the event-vehicle control command probability matrix database.

[0221] The server defines a complete set of events, which includes multiple events. Next, the server defines a complete set of vehicle control instructions, which includes multiple vehicle control instructions. Then, the server determines the first number of occurrences of each event and the second number of occurrences of each vehicle control instruction under each event within a preset statistical period. The server then determines the probability of occurrence of each vehicle control instruction under each event based on the first number of occurrences and the second number of occurrences to obtain an event-vehicle control instruction probability matrix.

[0222] Continuing with the above example, the preset statistical period is two weeks. The first occurrence count is the occurrence count (m i ). The second occurrence count is the occurrence count (n i ).

[0223] First, the vehicle system defines an event set M, which includes "volume set to: multimedia volume", "getting on and off the vehicle: main driver (when getting on the vehicle)", "main driver (when getting off the vehicle)", "double flash lights: when off()" and "seat belt: third row right seat (when fastened)".

[0224] Next, the vehicle system defines a complete set of vehicle control instructions N, which includes "air conditioning: air conditioning (off)", "air conditioning: air conditioning (on)", "AQS: AQS (off)" and "AQS: AQS (on)" and the like.

[0225] Based on the operation log of vehicle A in the past two weeks, calculate each event m in the event set M i The number of occurrences Count(m i ), and calculate at event m i Within a period of time after the occurrence, each vehicle control instruction n in the vehicle control instruction set N j The number of occurrences Count(n ij ).

[0226] Then, according to the number of occurrences Count(m i ) and the number of occurrences Count(n ij ), determine that vehicle A is executing event m i After that, each vehicle control command n j The probability of occurrence P ij , the calculation formula is:

[0227]

[0228] For example, in the past two weeks, the number of events "driver gets in the car" is 10 times. After "driver gets in the car", the number of vehicle control instructions "turn on the air conditioner" is 8 times. Then, under the "driver gets in the car" event, the probability of the vehicle control instruction "turn on the air conditioner" is 0.8. All occurrence probabilities are counted to obtain the event-vehicle control instruction probability matrix.

[0229] Finally, the same process is performed on other vehicles to obtain event-vehicle control command probability matrices of different vehicles, and these event-vehicle control command probability matrices are stored in the vehicle system to obtain an event-vehicle control command probability matrix database.

[0230] In this way, by analyzing and mining the user logs of multiple users offline, there is no need for a large number of high-quality data sets and a large amount of resources for fine-tuning training, thereby reducing computing costs and time. In addition, by counting and analyzing historical data, the server can more accurately predict and recommend vehicle control instructions, improve the accuracy of recommendations, and thus improve the user's driving experience.

[0231] The present application also provides a computer-readable storage medium containing a computer program. When the computer program is executed by one or more processors, the one or more processors execute the method of the present application.

[0232] It is understood that a computer program includes computer program code. The computer program code may be in source code form, object code form, executable file or some intermediate form. Computer readable storage media may include: any entity or device capable of carrying computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), and software distribution medium.

[0233] In the description of this specification, the descriptions with reference to the terms "specifically", "further", "particularly", "understandably", etc., mean that the specific features, structures, materials or characteristics described in conjunction with the embodiments or examples are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms are not intended to be considered as the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are contradictory.

[0234] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code that includes one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present application belong.

[0235] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A vehicle control instruction recommendation method, characterized in that: The method comprises: Acquire current event information of a target vehicle and user identification information of the target vehicle; Based on a preconfigured event-vehicle control instruction database, a first vehicle control instruction set is determined according to the current event information and the user identification information, and the first vehicle control instruction set is configured to be recommended to the target vehicle as vehicle control instructions that the target vehicle may execute under the current event.

2. The method according to claim 1, characterized in that The determining, based on the preconfigured event-vehicle control instruction database, of the first vehicle control instruction set according to the current event information and the user identification information comprises: Acquire a first matrix corresponding to the user identification information from the event-vehicle control instruction database according to the user identification information, wherein the first matrix is ​​an event-vehicle control instruction probability matrix of the target vehicle; Based on the event-vehicle control instruction database, determining a similar matrix set according to the first matrix, the similar matrix set including at least one event-operation probability matrix similar to the first matrix; The first vehicle control instruction set is determined according to the current event information and the similarity matrix set.

3. The method according to claim 2, characterized in that The event-vehicle control instruction database includes a plurality of second matrices, wherein the second matrix is ​​an event-vehicle control instruction probability matrix corresponding to a non-target vehicle, and the similarity matrix set is determined based on the event-vehicle control instruction database and the first matrix, including: respectively solving the similarity between the first matrix and each of the second matrices; A predetermined number of the second matrices is determined according to the similarity to determine the similarity matrix set.

4. The method according to claim 2, characterized in that: The determining the first vehicle control instruction set according to the current event information and the similarity matrix set includes: Determine a first recommended vehicle control instruction set according to the current event information and the similarity matrix set; The first vehicle control instruction set is determined based on the first recommended vehicle control instruction set.

5. The method according to claim 4, characterized in that The determining the first vehicle control instruction set according to the first recommended vehicle control instruction set includes: In a current event, if the target vehicle does not execute the vehicle control instructions in the first recommended vehicle control instruction set, the first vehicle control instruction set is determined according to the first recommended vehicle control instruction set.

6. The method according to claim 4, characterized in that The determining the first vehicle control instruction set according to the first recommended vehicle control instruction set includes: Under the current event, if the target vehicle executes at least one second vehicle control instruction in the first recommended vehicle control instruction set, the second vehicle control instruction is removed from the first recommended vehicle control instruction set to obtain a second recommended vehicle control instruction set; The first vehicle control instruction set is determined based on the second recommended vehicle control instruction set.

7. The method according to claim 4, characterized in that The determining the first vehicle control instruction set according to the first recommended vehicle control instruction set includes: determining a third recommended vehicle control instruction set according to the first recommended vehicle control instruction set based on a preset instruction quantity and a preset instruction probability; Acquire environmental information of the environment where the target vehicle is located; Based on a pre-trained large language model, the first vehicle control instruction set is determined according to the environmental information and the third recommended vehicle control instruction set.

8. The method according to claim 1, characterized in that The event-vehicle control command probability matrix database includes a plurality of event-vehicle control command probability matrices, each of which is associated with a vehicle, and the event-vehicle control command probability matrix is ​​constructed by the following steps: defining an event set, wherein the event set includes a plurality of events; defining a vehicle control instruction set, wherein the vehicle control instruction set includes a plurality of vehicle control instructions; Determine a first occurrence number of each event and a second occurrence number of each vehicle control instruction under each event within a preset statistical period; According to the first number of occurrences and the second number of occurrences, the occurrence probability of each of the vehicle control instructions under each of the events is determined to obtain an event-vehicle control instruction probability matrix.

9. A server, characterized in that: The server includes a memory and a processor, the memory stores a computer program, and when the computer program is executed by the processor, the method according to any one of claims 1 to 8 is implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium memory stores a computer program, and when the computer program is executed by one or more processors, the method according to any one of claims 1 to 8 is implemented.

Images