Vehicle control method, server and computer readable storage medium

Abstract

The invention discloses a vehicle control method, a server and a computer readable storage medium. The method comprises the steps that current perception information is acquired, the current perception information comprises user perception information and vehicle perception information, and the user perception information comprises user voice and / or user behaviors; according to the current perception information, determining an intention label, a user intention and candidate function points; and determining a vehicle control instruction according to the intention label, the user intention and the candidate function points. And issuing a vehicle control instruction to the vehicle to complete vehicle control. Thus, by combining the multi-mode current perception information, distinguishing the intention label of the user intention and understanding and reasoning the user intention, the service meeting the user expectation can be actively provided, and the driving experience of the user is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle control, and particularly relates to a vehicle control method, a server, and a computer-readable storage medium. Background Art

[0002] In the related art, an intelligent cockpit system can reason about clear and definite user instructions to meet user requirements. However, when dealing with ambiguous user instructions, the intelligent cockpit system has insufficient reasoning ability, making it difficult to generate vehicle control instructions that meet user requirements and affecting the user experience. Summary of the Invention

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

[0004] An embodiment of the present application provides a vehicle control method, the method including:

[0005] Obtain current perception information, where the current perception information includes user perception information and vehicle perception information, and the user perception information includes user voice and / or user behavior;

[0006] Determine an intent label, a user intent, and candidate function points according to the current perception information;

[0007] Determine a vehicle control instruction according to the intent label, the user intent, and the candidate function points;

[0008] Send the vehicle control instruction to the vehicle to complete the vehicle control.

[0009] In this way, the server obtains the current perception information, which includes user perception information and vehicle perception information, and the user perception information includes user voice and / or user behavior. Then, the server determines the intent label, the user intent, and the candidate function points according to the current perception information. Next, the server determines the vehicle control instruction according to the intent label, the user intent, and the candidate function points. Finally, the server sends the vehicle control instruction to the vehicle to complete the vehicle control. In this way, by combining multi-modal current perception information, differentiating the intent label of the user intent, and understanding and reasoning the user intent, it is possible to actively provide services that meet user expectations, thereby improving the user's driving experience.

[0010] In some embodiments, the determining the intent label, the user intent, and the candidate function points according to the current perception information includes:

[0011] Based on a first preset model, determine the intent label, the user intent, and the candidate function points according to the current perception information, and the first preset model is configured with a first prompt.

[0012] In this way, based on the first preset model, the server determines the intent label, the user intent, and the candidate function points according to the current perception information. The first preset model is configured with a first prompt. In this way, the first preset model can help the server quickly and accurately judge the intent label, determine the user intent, and select appropriate candidate function points, thereby improving the inference efficiency. By means of the first preset model, the server's dependence on the preset knowledge base can be reduced, thus lowering the requirement for knowledge input.

[0013] In some embodiments, the determining of the vehicle control instruction according to the intent label, the user intent, and the candidate function points includes:

[0014] Based on the second preset model and the preset knowledge base, the vehicle control instruction is determined according to the intent label, the user intent, and the candidate function points. The second preset model is configured with a second prompt.

[0015] In this way, based on the second preset model and the preset knowledge base, the server determines the vehicle control instruction according to the intent label, the user intent, and the candidate function points. The second preset model is configured with a second prompt. In this way, the second preset model and the preset knowledge base can help the server quickly and accurately generate the vehicle control instruction, thereby improving the efficiency of instruction generation.

[0016] In some embodiments, the intent label includes a first label, the first label is used to identify that the intent of the user perception information is clear, the second prompt includes a first sub-prompt, the preset knowledge base includes a first preset sub-knowledge base, and the determining of the vehicle control instruction based on the second preset model and the preset knowledge base according to the intent label, the user intent, and the candidate function points includes:

[0017] When the intent label is the first label, based on the first sub-prompt and the first preset sub-knowledge base, the vehicle control instruction is determined according to the user intent and the candidate function points.

[0018] In this way, when the intent label is the first label, the server determines the vehicle control instruction based on the first sub-prompt and the first preset sub-knowledge base according to the user intent and the candidate function points. In this way, for the scenario where the user intent is clear, reasoning is performed based on the first sub-prompt and the first preset sub-knowledge base to generate a control instruction that conforms to the actual situation, thereby improving the inference efficiency.

[0019] In some embodiments, the determining of the vehicle control instruction when the intent label is the first label, based on the first sub-prompt and the first preset sub-knowledge base, according to the user intent and the candidate function points includes:

[0020] Based on the first preset sub-knowledge base, determine the first target background knowledge according to the candidate function point, where the first target background knowledge includes the first information of the candidate function point, and the first information is used to describe the first adjustment operation of the candidate function point and the corresponding first applicable scenario, and the first adjustment operation includes the adjustment direction and / or adjustment gear of the candidate function point;

[0021] Based on the first sub-prompt word, determine the vehicle control instruction according to the first target background knowledge, the current perception information, the user intention and the candidate function point.

[0022] In this way, based on the first preset sub-knowledge base, the server determines the first target background knowledge according to the candidate function point, where the first target background knowledge includes the first information of the candidate function point, and the first information is used to describe the first adjustment operation of the candidate function point and the corresponding first applicable scenario, and the first adjustment operation includes the adjustment direction and / or adjustment gear of the candidate function point. Then, based on the first sub-prompt word, the server determines the vehicle control instruction according to the first target background knowledge, the current perception information, the user intention and the candidate function point. In this way, by obtaining the first target background knowledge of the candidate function point, the user's intention can be accurately understood, and a control instruction that meets the user's expectations can be generated.

[0023] In some embodiments, the intention label includes a second label, the second label is used to identify the ambiguity of the user perception information, the second prompt word includes a second sub-prompt word, the preset knowledge base includes a second preset sub-knowledge base, and based on the second preset model and the preset knowledge base, determine the vehicle control instruction according to the intention label, the user intention and the candidate function point, including:

[0024] When the intention label is the second label, based on the second sub-prompt word and the second preset sub-knowledge base, determine the vehicle control instruction according to the user intention and the candidate function point.

[0025] In this way, when the intention label is the second label, based on the second sub-prompt word and the second preset sub-knowledge base, the server determines the vehicle control instruction according to the user intention and the candidate function point. In this way, for the scenario where the user intention is ambiguous, reasoning based on the second sub-prompt word and the second preset sub-knowledge base can accurately understand the user's ambiguous intention and generate a control instruction that meets the user's expectations, thereby improving the reasoning efficiency.

[0026] In some embodiments, the step of, when the intention label is the second label, based on the second sub-prompt word and the second preset sub-knowledge base, determine the vehicle control instruction according to the user intention and the candidate function point, includes:

[0027] Based on the second preset sub-knowledge base, determine the second target background knowledge according to the candidate function point, where the second target background knowledge includes second information of the candidate function point, and the second information is used to describe a second adjustment operation of the candidate function point and a corresponding second applicable scenario, and the second adjustment operation includes an adjustment direction and an adjustment amplitude of the candidate function point;

[0028] Based on the second sub-prompt word, determine the vehicle control instruction according to the second target background knowledge, the current perception information, the user intention, and the candidate function point.

[0029] In this way, based on the second preset sub-knowledge base, the server determines the second target background knowledge according to the candidate function point, where the second target background knowledge includes second information of the candidate function point, and the second information is used to describe a second adjustment operation of the candidate function point and a corresponding second applicable scenario, and the second adjustment operation includes an adjustment direction and an adjustment amplitude of the candidate function point. Then, based on the second sub-prompt word, the server determines the vehicle control instruction according to the second target background knowledge, the current perception information, the user intention, and the candidate function point. In this way, by obtaining the second target background knowledge of the candidate function point, the fuzzy intention of the user can be accurately understood, and a control instruction more in line with the user's expectations can be generated.

[0030] In some embodiments, the preset knowledge base is constructed according to the following steps:

[0031] Obtain vehicle function points;

[0032] According to the vehicle function point, determine a third adjustment operation of the vehicle function point and a corresponding third applicable scenario, where the third adjustment operation includes an adjustment direction and / or an adjustment gear of the vehicle function point;

[0033] According to the vehicle function point, determine a fourth adjustment operation of the vehicle function point and a corresponding fourth applicable scenario, where the fourth adjustment operation includes an adjustment direction and an adjustment amplitude of the vehicle function point;

[0034] Construct the preset knowledge base according to the third adjustment operation and the corresponding third applicable scenario, and the fourth adjustment operation and the corresponding fourth applicable scenario.

[0035] Thus, first, obtain the vehicle function points. Then, based on the vehicle function points, determine the third adjustment operation and the corresponding third applicable scenario of the vehicle function points. The third adjustment operation includes the adjustment direction and / or adjustment gear of the vehicle function points. Next, based on the vehicle function points, determine the fourth adjustment operation and the corresponding fourth applicable scenario of the vehicle function points. The fourth adjustment operation includes the adjustment direction and adjustment amplitude of the vehicle function points. Finally, construct a preset knowledge base according to the third adjustment operation and the corresponding third applicable scenario, and the fourth adjustment operation and the corresponding fourth applicable scenario. In this way, the preset knowledge base provides detailed vehicle function knowledge, including the adjustment method, adjustment amplitude, and applicable scenario of each vehicle function point, thereby helping to understand the user's intention and generate control instructions.

[0036] An embodiment of the present application provides a server, which includes a processor and a memory. A computer program is stored on the memory. When the computer program is executed by the processor, the above vehicle control method is implemented.

[0037] An embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the vehicle control method as described above are implemented.

[0038] Additional aspects and advantages of the embodiments of the present application will be given in part in the following description, will become apparent in part from the following description, or will be understood through the practice of the embodiments of the present application. Description of the Drawings

[0039] The above and / or additional aspects and advantages of the present application will become apparent and easy to understand from the description of the embodiments in conjunction with the following drawings, where:

[0040] Figure 1 is one of the flow diagrams of the vehicle control method of some embodiments of the present application;

[0041] Figure 2 is another flow diagram of the vehicle control method of some embodiments of the present application;

[0042] Figure 3 is yet another flow diagram of the vehicle control method of some embodiments of the present application;

[0043] Figure 4 is still another flow diagram of the vehicle control method of some embodiments of the present application;

[0044] Figure 5 is yet still another flow diagram of the vehicle control method of some embodiments of the present application;

[0045] Figure 6It is the sixth schematic flowchart of the vehicle control method according to some embodiments of the present application;

[0046] Figure 7 It is the seventh schematic flowchart of the vehicle control method according to some embodiments of the present application;

[0047] Figure 8 It is the eighth schematic flowchart of the vehicle control method according to some embodiments of the present application. Detailed Embodiments

[0048] The following details the embodiments of the present application. The examples of the embodiments are shown in the accompanying drawings, where the same or similar reference numerals represent the same or similar elements or elements with the same or similar functions throughout. 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 should not be construed as a limitation to the embodiments of the present application.

[0049] In an intelligent cockpit system, for clear and definite user instructions, the system can effectively reason through advanced algorithms and logics to accurately meet the user's needs. For example, when the user issues specific instructions such as "Navigate to the nearest gas station" or "Increase the temperature inside the vehicle", the intelligent cockpit can quickly understand and execute the corresponding operations.

[0050] However, when the user uses vague user instructions, the reasoning ability of the intelligent cockpit is insufficient. For example, the user may say "I feel a bit cold" or "This music doesn't sound right". These expressions do not directly give specific operation instructions but require the system to understand the user's true intentions and feelings before generating corresponding vehicle control instructions. In this case, the intelligent cockpit system often has difficulty accurately inferring the user's needs, which may lead to incorrect operations or inaction, thus affecting the user experience.

[0051] Based on the above problems, please refer to Figure 1 , some embodiments of the present application provide a vehicle control method, which includes:

[0052] 01: Obtain the current perception information;

[0053] 02: Determine the intention label, user intention, and candidate function points according to the current perception information;

[0054] 03: Determine the vehicle control instruction according to the intention label, user intention, and candidate function points;

[0055] 04: Send the vehicle control instruction to the vehicle to complete vehicle control.

[0056] The embodiments of the present application also provide a server, including a memory and a processor. The vehicle control method of the embodiments of the present application can be implemented by the server of the embodiments of the present application. Specifically, a computer program is stored in the memory, and the processor is used to obtain the current perception information. And according to the current perception information, determine the intention label, user intention and candidate function points. The processor is also used to determine the vehicle control instruction according to the intention label, user intention and candidate function points. And send the vehicle control instruction to the vehicle to complete the vehicle control.

[0057] The embodiments of the present application also provide a vehicle control device. The vehicle control method of the embodiments of the present application can be implemented by the vehicle control device of the embodiments of the present application. Specifically, the vehicle control device includes a receiving module and a determining module. The receiving module is used to obtain the current perception information. The determining module is used to determine the intention label, user intention and candidate function points according to the current perception information. The determining module is also used to determine the vehicle control instruction according to the intention label, user intention and candidate function points. The sending module is used to send the vehicle control instruction to the vehicle to complete the vehicle control.

[0058] Specifically, the current perception information refers to the data collected by various sensors and systems of the vehicle, which are used to understand and respond to the user's needs and the environment around the vehicle, including user perception information and vehicle perception information.

[0059] User perception information refers to the data and information about vehicle users (such as drivers and passengers) collected by the vehicle through its sensors and systems, including user voice requests and user behaviors. User voice requests refer to requests made by users through voice commands, such as "Navigate to the nearest gas station" or "Play music". User behavior refers to the operation and behavior patterns of users, such as adjusting the seat, switching the driving mode, using the touch screen, etc. In some embodiments, the user perception information also includes user physiological information, such as heart rate, facial expression, etc., which can help the system evaluate the user's mood and comfort level.

[0060] Vehicle perception information refers to the data and information about the vehicle and the surrounding environment collected by the vehicle through its sensors, cameras and other detection devices, including the state perception information of in-vehicle device functions and the environmental state information of the current vehicle location. The state perception information involves the functional states of various devices inside the vehicle, such as whether information systems such as the navigation system, entertainment system, and communication system are running, and their current settings and states. Or whether comfort systems such as air conditioning, seat heating / ventilation, and window control are activated, and their current temperature, wind speed, etc. settings. It can also be whether driving assistance systems such as adaptive cruise control, lane keeping assistance, and automatic parking systems are enabled, and their current parameters and states. The environmental state information involves traffic conditions, weather conditions, and road condition information, etc.

[0061] An intent label refers to a label for classifying user instructions, used to distinguish whether the user's intent is a vague intent or a clear intent, which can help the intelligent cockpit system better understand the user's intent and select appropriate control strategies to meet the user's needs.

[0062] User intent refers to the expectations of the vehicle user obtained by the server for vehicle functions, performance, or services in a specific scenario.

[0063] Candidate function points refer to potential function options provided by the intelligent cockpit system based on the user's intent. By selecting appropriate candidate function points, the user's needs can be met and the user experience can be improved.

[0064] The server receives perception information, that is, the user perception information and vehicle perception information collected by the server through various sensors of the vehicle. For example, user voices such as "I'm a bit hot" and "Open the window", or user behaviors such as adjusting the seat and frequently touching the air conditioner panel. It also includes vehicle perception information such as air conditioner temperature, window switch state, seat angle, vehicle speed, and gear position.

[0065] Next, based on the currently collected perception information, the server determines the intent label, user intent, and candidate function points.

[0066] Then, the server generates specific vehicle control instructions according to the determined intent label, user intent, and candidate function points. For example, if the user intent is "I'm a bit hot", the server may generate vehicle control instructions such as "Lower the air conditioner temperature by 3 degrees" and "Reduce the air conditioner fan speed by 2 gears".

[0067] Finally, the server sends the generated vehicle control instructions to the vehicle to control the relevant devices to execute the instructions, such as adjusting the air conditioner temperature and fan speed, so as to meet the user's needs.

[0068] In summary, the server obtains the current perception information, which includes user perception information and vehicle perception information, and the user perception information includes user voice and / or user behavior. Next, the server determines the intent label, user intent, and candidate function points according to the current perception information. Then, the server determines the vehicle control instructions according to the intent label, user intent, and candidate function points. Finally, the server issues the vehicle control instructions to the vehicle to complete vehicle control. In this way, by combining multi-modal current perception information, distinguishing the intent label of the user's intent, and understanding and reasoning the user's intent, it can actively provide services that meet the user's expectations, thereby improving the user's driving experience.

[0069] Please refer to Figure 2 , in some embodiments, step 02 (determining the intent label, user intent, and candidate function points according to the current perception information) includes:

[0070] 021: Based on the first preset model, determine the intent label, user intent, and candidate function points according to the current perception information.

[0071] In some embodiments, the determination module is further configured to determine the intent label, user intent, and candidate function points based on the first preset model according to the current perception information.

[0072] In some embodiments, the processor is further configured to determine the intent label, user intent, and candidate function points based on the first preset model according to the current perception information.

[0073] Specifically, the first preset model refers to a pre-trained model for judging the intent label, user intent, and candidate function points according to the current perception information. The first prompt word is the first step in the user demand reasoning process, which can help the system judge the clarity of the user intent and provide a direction for further reasoning, that is, by judging the clarity of the user intent, select the appropriate reasoning path and knowledge base. In some embodiments, the basis for judging the clarity of the user intent is: analyze whether the user voice command or behavior contains a clear device name and operation command, or combine the current vehicle state information to judge whether the user command clearly points to a certain device or operation.

[0074] In some embodiments, the first prompt word may be:

[0075] You are a helpful assistant. You will receive a user's voice command, or a user's behavior in the vehicle, and the vehicle state. You need to make the following judgments:

[0076] Your first task is to make a judgment on the user's voice or behavior. Does the user have a clear intent?

[0077] Clear intent: means the running device is clear (such as the air conditioner), and the specific action is also clear (such as: lower the temperature by 2 degrees)

[0078] Vague intent: means the user has a certain need but doesn't know exactly what to do (such as: I'm a bit hot)

[0079] Please give your answer as clear or vague, where clear means the user has a clear intent, and vague means the user has a vague intent.

[0080] Your second task is to state the user's intent. Please use brief words (no more than 20 characters) to describe the user intent.

[0081] Your third task is to state the device(s) that may be needed to execute the intent.

[0082] Please provide your feedback in the following way:

[0083] Judgment: (Clear or vague, choose one)

[0084] Intention: (User intention, no more than 20 characters)

[0085] Candidate function points: (Vehicle function points to achieve user intention, can be multiple)

[0086] The input content is: rep

[0087] The server uses the first preset model and the "first prompt word" to analyze the current perception information, and finally determines the intention label, user intention and candidate function points.

[0088] The following uses an example to illustrate the vehicle control method provided by the embodiment of the present application. When the currently obtained perception information is "User instruction: Turn down the air conditioner by two gears; Vehicle state: The user is sitting in the driver's seat, P gear, the driver's window is open, the passenger window is closed, all vehicle doors are closed, the driver's seat backrest is vertical, the air conditioner is running, the air conditioner temperature is 18 degrees, the air conditioner wind speed is 6 gears (loud sound), the driver's reading light is on, and the passenger's reading light is off", the result obtained by the first preset model is "Judgment: Clear; User intention: The user hopes to reduce the air volume of the air conditioner to reduce noise. Candidate function points: Air conditioner air volume".

[0089] If the current perception information is "User instruction: I'm a bit cold; Vehicle state: The user is sitting in the driver's seat, P gear, the driver's window is open, the passenger window is closed, all vehicle doors are closed, the driver's seat backrest is vertical, the air conditioner is running, the air conditioner temperature is 18 degrees, the air conditioner wind speed is 6 gears (loud sound), the driver's reading light is on, and the passenger's reading light is off", the result obtained by the first preset model is "Judgment: Vague; User intention: The user feels cold and hopes to adjust the temperature inside the vehicle. Candidate function points: Air conditioner temperature, air conditioner air volume".

[0090] In this way, the first preset model can help the server quickly and accurately judge the intention label, determine the user intention, and select appropriate candidate function points, thereby improving the reasoning efficiency. Through the first preset model, the server's dependence on the preset knowledge base can be reduced, thereby reducing the need for knowledge input.

[0091] Please refer to Figure 3 , in some embodiments, step 03 (determine the vehicle control instruction according to the intention label, user intention and candidate function points) includes:

[0092] 031: Based on the second preset model and the preset knowledge base, determine the vehicle control instruction according to the intention label, user intention and candidate function points. The second preset model is configured with a second prompt word.

[0093] In some embodiments, the determination module is further configured to determine a vehicle control instruction based on a second preset model and a preset knowledge base according to the intent label, the user intent, and the candidate function points.

[0094] In some embodiments, the processor is further configured to determine a vehicle control instruction based on a second preset model and a preset knowledge base according to the intent label, the user intent, and the candidate function points.

[0095] Specifically, the second preset model refers to a pre-trained model for generating a vehicle control instruction according to the intent label, the user intent, and the candidate function points. It is configured with a second prompt word for guiding the model to perform reasoning. The preset knowledge base contains specific information about various functions of the vehicle, such as the wind speed at each gear position and the impact of each temperature setting on the interior temperature of the vehicle. The second prompt word can guide the system on how to adopt different processing methods according to different user intents.

[0096] The server uses the second preset model and the preset knowledge base, combines the intent label, the user intent, and the candidate function points, and generates a specific vehicle control instruction.

[0097] In this way, based on the second preset model and the preset knowledge base, the server determines the vehicle control instruction according to the intent label, the user intent, and the candidate function points, and the second preset model is configured with a second prompt word. In this way, the second preset model and the preset knowledge base can help the server quickly and accurately generate a vehicle control instruction, thereby improving the efficiency of instruction generation.

[0098] Please refer to Figure 4 , in some embodiments, the intent label includes a first label, the first label is used to identify that the intent of the user perception information is clear, the second prompt word includes a first sub-prompt word, the preset knowledge base includes a first preset sub-knowledge base, and step 031 (determining a vehicle control instruction based on a second preset model and a preset knowledge base according to the intent label, the user intent, and the candidate function points) includes:

[0099] 0311: When the intent label is the first label, based on the first sub-prompt word and the first preset sub-knowledge base, determine the vehicle control instruction according to the user intent and the candidate function points.

[0100] In some embodiments, the determination module is further configured to, when the intent label is the first label, determine the vehicle control instruction based on the first sub-prompt word and the first preset sub-knowledge base according to the user intent and the candidate function points.

[0101] In some embodiments, the processor is further configured to, when the intent label is the first label, determine the vehicle control instruction based on the first sub-prompt word and the first preset sub-knowledge base according to the user intent and the candidate function points.

[0102] Specifically, the first tag is used to clearly identify the user's intention of perceiving information, that is, the function and operation to be performed are clearly specified in the user's instruction. For example, "Reduce the air volume of the air conditioner by two gears."

[0103] The first sub-prompter is used to guide the system in reasoning and decision-making, including the following information: First, the scenario setting, which clarifies that the scenario occurs in a car and the interaction between the user and the system occurs inside the intelligent cockpit. Second, role-playing, setting the system as a specific role, which needs to judge whether the user needs service based on the user's behavior and vehicle status information and perform corresponding operations. Third, obtaining relevant background knowledge, providing relevant knowledge for adjusting candidate function points, including information such as wind speed, noise, and energy consumption at different gears, as well as the usage scenarios and effects of different gears. Fourth, obtaining detailed current perception information. Fifth, determining the well-defined user intention. Sixth, determining the well-defined candidate function points.

[0104] In some embodiments, the first sub-prompter may be:

[0105] Role: Personal assistant

[0106] You will act as a personal assistant in a car scenario. You need to combine the user's behavior and the vehicle's perception point information to judge whether the user needs your service.

[0107] # Background knowledge

[0108] 1. Air volume of the air conditioner

[0109] Turning up the air volume of the air conditioner: Increases the air output of the air conditioning system. Accelerates the air circulation speed inside the vehicle and improves the temperature adjustment efficiency. Provides a more rapid comfort experience when quickly changing the vehicle interior temperature (such as in hot summer or cold winter). May generate greater noise, affecting the riding comfort. Increases the load on the air conditioning system, may slightly increase energy consumption, and affects the vehicle's battery life.

[0110] Turning down the air volume of the air conditioner: Reduces the air output of the air conditioning system. Slows down the air flow speed inside the vehicle, suitable for passengers sensitive to the wind feeling. Reduces the operating noise of the air conditioning system, enhancing the riding quietness. Reduces the load on the air conditioning system, helps save energy, and extends the vehicle's battery life. Provides a more energy-efficient usage method when the temperature adjustment requirement is not urgent.

[0111] *Gear 1: The lowest gear of the air conditioner air volume, the slowest wind speed, suitable for fine-tuning the vehicle interior temperature or reducing noise.

[0112] *Gears 2 to 5: Medium-low to medium air volume settings, balancing the temperature adjustment speed and noise.

[0113] *Gears 6 to 8: Medium-high to high air volume settings, accelerating the temperature adjustment speed but accompanied by greater noise.

[0114] *9th to 10th gear: The highest air volume setting, providing the fastest temperature adjustment speed, but with the loudest noise.

[0115] ## Detailed perception point information

[0116] Vehicle status:

[0117] The user is sitting in the driver's seat,

[0118] The vehicle is in P gear,

[0119] The driver's side window is open, and the passenger's side window is closed,

[0120] All vehicle doors are closed,

[0121] The driver's seat backrest is vertical,

[0122] The air conditioner is running, the air conditioner temperature is 18 degrees, and the air conditioner air volume is at 6th gear (relatively loud),

[0123] The driver's reading light is on, and the passenger's reading light is off

[0124] ## User intention

[0125] The user hopes to reduce the air volume of the air conditioner to reduce the noise.

[0126] ## Candidate functions

[0127] {"Function name": "Air conditioner air volume", "Function value": {"Action": ["Adjust the air conditioner air volume to x gear"]}}

[0128] Please select the appropriate execution function:

[0129] When the intention label is the first label, the server uses the first sub-prompt word and the first preset sub-knowledge base, combines the user intention and the candidate function points, and generates a vehicle control instruction.

[0130] In this way, for scenarios with clear user intentions, reasoning is performed based on the first sub-prompt word and the first preset sub-knowledge base to generate control instructions that conform to the actual situation, thereby improving the reasoning efficiency.

[0131] Please refer to Figure 5 , in some embodiments, step 0311 (in the case where the intention label is the first label, based on the first sub-prompt word and the first preset sub-knowledge base, determine the vehicle control instruction according to the user intention and the candidate function points), includes:

[0132] 03111: Based on the first preset sub-knowledge base, determine the first target background knowledge according to the candidate function points;

[0133] 03112: Based on the first sub - prompt, determine the vehicle control command according to the first target background knowledge, current perception information, user intention, and candidate function points.

[0134] In some embodiments, the determination module is further configured to determine the first target background knowledge based on the first preset sub - knowledge base according to the candidate function points, and determine the vehicle control command based on the first sub - prompt, the first target background knowledge, current perception information, user intention, and candidate function points.

[0135] In some embodiments, the processor is further configured to determine the first target background knowledge based on the first preset sub - knowledge base according to the candidate function points, and determine the vehicle control command based on the first sub - prompt, the first target background knowledge, current perception information, user intention, and candidate function points.

[0136] Specifically, the first preset sub - knowledge base includes knowledge required for clear intentions, such as the adjustment direction, adjustment gear, and corresponding applicable scenarios of each function point.

[0137] The first target background knowledge includes the first information of the candidate function point, such as the adjustment direction, adjustment gear, and applicable scenario.

[0138] Based on the first preset sub - knowledge base and the candidate function points, determine the first target background knowledge. Then, use the first sub - prompt to guide the model to generate the vehicle control command according to the first target background knowledge, current perception information, user intention, and candidate function points.

[0139] Continuing the above example, if the current perception information is "User instruction: Reduce the air volume of the air conditioner by two gears; Vehicle state: The user is sitting in the driver's seat, in P gear, the driver's window is open, the passenger's window is closed, all vehicle doors are closed, the driver's seat backrest is vertical, the air conditioner is running, the air conditioner temperature is 18 degrees, the air volume of the air conditioner is 6 gears (loud sound), the driver's reading light is on, and the passenger's reading light is off", the first sub - prompt is "Role: Personal assistant"

[0140] You will act as a personal assistant in a car scenario. You need to combine the user's behavior and the vehicle's perception point information to determine whether the user needs your service.

[0141] # Background knowledge

[0142] 1. Air volume of the air conditioner

[0143] Turning up the air volume of the air conditioner: Increases the air output of the air - conditioning system. Speeds up the air circulation speed in the vehicle compartment and improves the temperature adjustment efficiency. Provides a more rapid comfort experience when quickly changing the interior temperature (such as in hot summer or cold winter). May generate greater noise, affecting the riding comfort. Increases the load on the air - conditioning system, may slightly increase energy consumption, and affects the vehicle's endurance.

[0144] Air conditioner air volume turned down: Reduce the air volume output of the air conditioning system. Slow down the air flow speed inside the vehicle compartment, suitable for passengers sensitive to the wind feeling. Reduce the operating noise of the air conditioning system and enhance the riding quietness. Reduce the load of the air conditioning system, help save energy, and extend the vehicle's cruising range. Provide a more energy-saving usage mode when the temperature adjustment requirement is not urgent.

[0145] *Gear 1: The lowest gear of the air conditioner air volume, the slowest wind speed, suitable for fine-tuning the temperature inside the vehicle or reducing noise.

[0146] *Gears 2 to 5: Medium-low to medium air volume settings, balancing the temperature adjustment speed and noise.

[0147] *Gears 6 to 8: Medium-high to high air volume settings, accelerating the temperature adjustment speed but accompanied by relatively large noise.

[0148] *Gears 9 to 10: The highest air volume settings, providing the fastest temperature adjustment speed but with the largest noise.

[0149] ## Detailed perception point information

[0150] Vehicle status:

[0151] The user is sitting in the driver's seat,

[0152] The vehicle is in P gear,

[0153] The driver's side window is open, the passenger's side window is closed,

[0154] All vehicle doors are closed,

[0155] The driver's seat backrest is vertical,

[0156] The air conditioner is running, the air conditioner temperature is 18 degrees, the air conditioner air volume is at gear 6 (relatively loud sound),

[0157] The driver's reading light is on, the passenger's reading light is off

[0158] ## User intention

[0159] The user hopes to reduce the air volume of the air conditioner to reduce noise.

[0160] ## Candidate functions

[0161] {"Function name": "Air conditioner air volume", "Function value": {"Action": ["Adjust the air conditioner air volume to gear x"]}}

[0162] Please select the appropriate execution function: ”, determine that the vehicle control command is “Adjust the air conditioner air volume to gear 4”.

[0163] Thus, based on the first preset sub-knowledge base, the server determines the first target background knowledge according to the candidate function points. The first target background knowledge includes the first information of the candidate function points, and the first information is used to describe the first adjustment operation of the candidate function points and the corresponding first applicable scenarios. The first adjustment operation includes the adjustment direction and / or adjustment gear of the candidate function points. Then, based on the first sub-prompt, the server determines the vehicle control instruction according to the first target background knowledge, the current perception information, the user intention, and the candidate function points. In this way, by obtaining the first target background knowledge of the candidate function points, the user intention can be accurately understood, and the control instruction that meets the user's expectations can be generated.

[0164] Please refer to Figure 6 , in some embodiments, the intention label includes a second label, the second label is used to identify the ambiguity of the user perception information, the second prompt includes a second sub-prompt, the preset knowledge base includes a second preset sub-knowledge base, and step 031 (based on the second preset model and the preset knowledge base, determine the vehicle control instruction according to the intention label, the user intention, and the candidate function points) includes:

[0165] 0312: When the intention label is the second label, based on the second sub-prompt and the second preset sub-knowledge base, determine the vehicle control instruction according to the user intention and the candidate function points.

[0166] In some embodiments, the determination module is further configured to, when the intention label is the second label, determine the vehicle control instruction according to the user intention and the candidate function points based on the second sub-prompt and the second preset sub-knowledge base.

[0167] In some embodiments, the processor is further configured to, when the intention label is the second label, determine the vehicle control instruction according to the user intention and the candidate function points based on the second sub-prompt and the second preset sub-knowledge base.

[0168] Specifically, the second label is used to identify the ambiguity of the user perception information, that is, the user instruction expresses a demand, but does not clearly specify the function and operation to be performed. The system needs to perform further reasoning to determine the target device and the target execution operation, such as "I'm a little hot".

[0169] The second sub-prompter is used to guide the system in reasoning and decision-making, corresponding to the second label, and includes the knowledge required for ambiguous intentions: First, scene setting, which clarifies that the scene occurs in a car and the interaction between the user and the system occurs inside the intelligent cockpit. Second, role-playing, which sets the system as a specific role and needs to judge whether the user needs services based on the user's behavior and vehicle status information and perform corresponding operations. Third, obtaining the corresponding background knowledge, which provides the relevant knowledge for adjusting the candidate function points, including information such as wind speed, noise, and energy consumption at different gears, as well as the usage scenarios and effects of different gears. Fourth, obtaining detailed current perception information. Fifth, determining the well-defined user intention. Sixth, determining the well-defined candidate function points.

[0170] When the intention label is the second label, the server uses the second sub-prompter and the second preset sub-knowledge base, combines the user intention and the candidate function points, and generates a vehicle control instruction.

[0171] In this way, when the intention label is the second label, based on the second sub-prompter and the second preset sub-knowledge base, the server determines the vehicle control instruction according to the user intention and the candidate function points. In this way, for the scenario where the user intention is ambiguous, based on the second sub-prompter and the second preset sub-knowledge base for reasoning, the server accurately understands the user's ambiguous intention and generates a control instruction that meets the user's expectations, thereby improving the reasoning efficiency.

[0172] Please refer to Figure 7 , in some embodiments, step 0312 (when the intention label is the second label, based on the second sub-prompter and the second preset sub-knowledge base, determine the vehicle control instruction according to the user intention and the candidate function points) includes:

[0173] 03121: Based on the second preset sub-knowledge base, determine the second target background knowledge according to the candidate function points;

[0174] 03122: Based on the second sub-prompter, determine the vehicle control instruction according to the second target background knowledge, the current perception information, the user intention, and the candidate function points.

[0175] In some embodiments, the determination module is further configured to determine the second target background knowledge based on the second preset sub-knowledge base according to the candidate function points. And based on the second sub-prompter, determine the vehicle control instruction according to the second target background knowledge, the current perception information, the user intention, and the candidate function points.

[0176] In some embodiments, the processor is further configured to determine the second target background knowledge based on the second preset sub-knowledge base according to the candidate function points. And based on the second sub-prompter, determine the vehicle control instruction according to the second target background knowledge, the current perception information, the user intention, and the candidate function points.

[0177] Specifically, the second preset sub-knowledge base includes knowledge required for fuzzy intentions, such as the adjustment direction, adjustment amplitude of each function point, and the corresponding applicable scenarios.

[0178] The second target background knowledge includes the second information of the candidate function points, such as the adjustment direction, adjustment amplitude, and applicable scenarios.

[0179] Based on the second preset sub-knowledge base and the candidate function points, determine the second target background knowledge. Then, use the second sub-prompt to guide the model to generate vehicle control instructions according to the second target background knowledge, the current perception information, the user intention, and the candidate function points.

[0180] In some embodiments, if the current perception information is "User instruction: I'm a bit cold; Vehicle state: The user is sitting in the driver's seat, in P gear, the driver's window is open, the passenger's window is closed, all vehicle doors are closed, the driver's seat backrest is vertical, the air conditioner is running, the air conditioner temperature is 18 degrees, the air conditioner air volume is at level 6 (relatively loud), the driver's reading light is on, and the passenger's reading light is off", and the second sub-prompt is "Role: Personal assistant

[0181] You will act as a personal assistant in a car scenario. You need to combine the user's behavior and the vehicle's perception point information to determine whether the user needs your service.

[0182] # Background knowledge

[0183] 1. Air conditioner temperature

[0184] Increase the air conditioner temperature by 8 degrees: Applicable when in extremely cold weather, quickly raising the interior temperature of the vehicle to a comfortable level when the passengers in the vehicle generally feel cold and need to significantly increase the interior temperature.

[0185] Increase the air conditioner temperature by 3 degrees: Applicable in slightly cool weather, slightly raising the interior temperature for a more comfortable riding experience when the passengers in the vehicle feel slightly cold but do not require a large adjustment of the temperature.

[0186] Decrease the air conditioner temperature by 3 degrees: Applicable in warm weather, slightly lowering the interior temperature for a more comfortable riding experience when the passengers in the vehicle feel slightly hot but do not require a large adjustment of the temperature.

[0187] Decrease the air conditioner temperature by 8 degrees: Applicable when in extremely hot weather, quickly lowering the interior temperature of the vehicle to a comfortable level when the passengers in the vehicle generally feel overheated and need to significantly lower the interior temperature.

[0188] 2. Air conditioner air volume

[0189] Increase the air conditioner air volume by 5 levels: Applicable when the difference between the interior temperature and the desired temperature is large, and it is necessary to quickly adjust the interior temperature of the vehicle to a comfortable level. There is smoke or odor in the vehicle, and it is necessary to quickly remove it to improve the air quality.

[0190] Increase the air volume of the air conditioner by 2 gears: applicable when the temperature inside the vehicle is close to comfort but still has a slight deviation. Slightly enhancing the ventilation effect is required to achieve the best comfort by fine-tuning the air volume, but excessive noise is not desired.

[0191] Decrease the air volume of the air conditioner by 2 gears: applicable when the desired temperature has been reached inside the vehicle, and passengers are sensitive to the air volume and hope to reduce the air volume to improve the riding comfort while reducing the noise generated by the operation of the air conditioner.

[0192] Decrease the air volume of the air conditioner by 5 gears: applicable during night driving, calls, and rest. Reducing the air volume can reduce noise and improve sleep quality. The temperature inside the vehicle has stabilized, and there is no need for a large amount of hot and cold air to flow, reducing energy consumption.

[0193] ## Detailed perception point information

[0194] Vehicle status:

[0195] The user is sitting in the driver's seat,

[0196] The vehicle is in P gear,

[0197] The driver's window is open, and the passenger's window is closed,

[0198] All vehicle doors are closed,

[0199] The driver's seat backrest is vertical,

[0200] The air conditioner is running, the air conditioner temperature is 18 degrees, and the air volume of the air conditioner is at gear 6 (relatively loud),

[0201] The driver's reading light is on, and the passenger's reading light is off

[0202] ## User intention

[0203] The user feels cold and hopes to adjust the temperature inside the vehicle.

[0204] ## Candidate functions

[0205] {"Function name": "Air conditioner temperature", "Function value": {"Action": ["Increase the air conditioner temperature by 8 degrees", "Increase the air conditioner temperature by 3 degrees", "Decrease the air conditioner temperature by 3 degrees", "Decrease the air conditioner temperature by 8 degrees"]}}, {"Function name": "Air conditioner air volume", "Function value": {"Action": ["Increase the air volume of the air conditioner by 5 gears", "Increase the air volume of the air conditioner by 2 gears", "Decrease the air volume of the air conditioner by 2 gears", "Decrease the air volume of the air conditioner by 5 gears"]}}

[0206] Please select the appropriate execution function:”. The determined vehicle control instructions are “Increase the air conditioner temperature by 3 degrees; Decrease the air conditioner fan speed by 2 gears”. In this way, the “direction + step size” inference scheme can avoid generating control instructions that do not conform to the actual situation. For example, when the air conditioner fan speed is already at gear 5, generating an instruction to “set the air conditioner fan speed to gear 5” again, thereby improving the efficiency of the system operation.

[0207] In this way, based on the second preset sub-knowledge base, the server determines the second target background knowledge according to the candidate function points. The second target background knowledge includes the second information of the candidate function points, and the second information is used to describe the second adjustment operation and the corresponding second applicable scenario of the candidate function points. The second adjustment operation includes the adjustment direction and adjustment amplitude of the candidate function points. Then, based on the second sub-prompts, the server determines the vehicle control instructions according to the second target background knowledge, the current perception information, the user intention, and the candidate function points. In this way, by obtaining the second target background knowledge of the candidate function points, the server can accurately understand the user's fuzzy intention and generate control instructions that better meet the user's expectations.

[0208] Please refer to Figure 8 , in some embodiments, the preset knowledge base is constructed according to the following steps:

[0209] 051: Obtain vehicle function points;

[0210] 052: According to the vehicle function points, determine the third adjustment operation and the corresponding third applicable scenario of the vehicle function points;

[0211] 053: According to the vehicle function points, determine the fourth adjustment operation and the corresponding fourth applicable scenario of the vehicle function points;

[0212] 054: According to the third adjustment operation and the corresponding third applicable scenario, and the fourth adjustment operation and the corresponding fourth applicable scenario, construct the preset knowledge base.

[0213] In some embodiments, the vehicle control device further includes a knowledge base construction module. The knowledge base construction module is used to obtain vehicle function points. And according to the vehicle function points, determine the third adjustment operation and the corresponding third applicable scenario of the vehicle function points. The knowledge base construction module is also used to according to the vehicle function points, determine the fourth adjustment operation and the corresponding fourth applicable scenario of the vehicle function points. And according to the third adjustment operation and the corresponding third applicable scenario, and the fourth adjustment operation and the corresponding fourth applicable scenario, construct the preset knowledge base.

[0214] In some embodiments, the processor is further configured to obtain vehicle function points, and determine a third adjustment operation and a corresponding third applicable scenario for the vehicle function points according to the vehicle function points. The processor is further configured to determine a fourth adjustment operation and a corresponding fourth applicable scenario for the vehicle function points according to the vehicle function points. And construct a preset knowledge base according to the third adjustment operation and the corresponding third applicable scenario, and the fourth adjustment operation and the corresponding fourth applicable scenario.

[0215] Specifically, obtain all adjustable function points in the vehicle, such as air conditioner temperature, air conditioner air volume, seat back angle, etc.

[0216] According to each function point, determine the adjustable operations that can be performed. For example, the air conditioner temperature can be adjusted in terms of direction (increase or decrease) and gear (1 - 30 degrees), and record the corresponding applicable scenarios, such as "increasing the air conditioner temperature is applicable to cold weather".

[0217] According to each function point, determine the adjustable range. For example, the adjustable range of the air conditioner temperature is 1 degree, and record the corresponding applicable scenarios, such as "for every 1 - degree increase in the air conditioner temperature, the vehicle interior temperature rises approximately 2 degrees".

[0218] Construct a preset knowledge base according to the third adjustment operation and the applicable scenario, and the fourth adjustment operation and the applicable scenario, which includes the adjustment direction, adjustment gear, adjustment range of each function point, and the corresponding applicable scenarios.

[0219] In this way, first, obtain the vehicle function points. Then, according to the vehicle function points, determine the third adjustment operation and the corresponding third applicable scenario for the vehicle function points. The third adjustment operation includes the adjustment direction and / or adjustment gear of the vehicle function points. Then, according to the vehicle function points, determine the fourth adjustment operation and the corresponding fourth applicable scenario for the vehicle function points. The fourth adjustment operation includes the adjustment direction and adjustment range of the vehicle function points. Finally, construct a preset knowledge base according to the third adjustment operation and the corresponding third applicable scenario, and the fourth adjustment operation and the corresponding fourth applicable scenario. In this way, the preset knowledge base provides detailed vehicle function knowledge, including the adjustment method, adjustment range, and applicable scenarios of each vehicle function point, thereby helping to understand the user's intention and generate control instructions.

[0220] The present application also provides a computer - readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the vehicle control method as described above are implemented.

[0221] It can be understood that a computer program includes computer program code. The computer program code can be in the form of source code, object code, executable file or some intermediate form, etc. The computer-readable storage medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), and software distribution medium, etc.

[0222] In the description of this specification, the descriptions referring to terms such as "specifically", "further", "specially", "understandably", etc. mean that the specific features, structures, materials or characteristics described in connection with the embodiments or examples are included in at least one embodiment or example of the present application. In this specification, the schematic expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in a suitable manner in any one or more embodiments or examples. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0223] Any process or method description shown in the flowchart or described in other ways herein can be understood as representing a module, segment or part of the executable request code including one or more steps for implementing a specific logical function or process, and the scope of the preferred embodiments of the present application includes additional implementations, where the functions can be executed in a substantially simultaneous manner or in the reverse order according to the functions involved, rather than in the order shown or discussed, which should be understood by those skilled in the technical field to which the embodiments of the present application belong.

[0224] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present application. Those of ordinary skill in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present application.

Claims

1. A vehicle control method, characterized in that: The method comprises: Acquiring current perception information, where the current perception information includes user perception information and vehicle perception information, where the user perception information includes user voice and / or user behavior; Determine the intent label, user intent, and candidate function points based on the current perception information; Determining a vehicle control instruction according to the intention label, the user intention, and the candidate function point; The vehicle control instruction is sent to the vehicle to complete the vehicle control.

2. The method according to claim 1, characterized in that Determining the intent label, user intent, and candidate function points based on the current perception information includes: Based on a first preset model, the intent label, the user intent and the candidate function point are determined according to the current perception information, and the first preset model is configured with a first prompt word.

3. The method according to claim 1, characterized in that The determining of the vehicle control instruction according to the intention label, the user intention and the candidate function point includes: Based on a second preset model and a preset knowledge base, a vehicle control instruction is determined according to the intention label, the user intention and the candidate function point, and the second preset model is configured with a second prompt word.

4. The method according to claim 3, characterized in that The intention label includes a first label, the first label is used to identify that the intention of the user perception information is clear, the second prompt word includes a first sub-prompt word, the preset knowledge base includes a first preset sub-knowledge base, and the vehicle control instruction is determined based on the second preset model and the preset knowledge base according to the intention label, the user intention and the candidate function point, including: In the case where the intention label is the first label, the vehicle control instruction is determined based on the first sub-prompt word and the first preset sub-knowledge base according to the user intention and the candidate function points.

5. The method according to claim 4, characterized in that In the case where the intention label is the first label, based on the first sub-prompt word and the first preset sub-knowledge base, according to the user intention and the candidate function point, determining the vehicle control instruction includes: Based on the first preset sub-knowledge base, first target background knowledge is determined according to the candidate function point, wherein the first target background knowledge includes first information of the candidate function point, the first information is used to describe a first adjustment operation of the candidate function point and a corresponding first applicable scenario, and the first adjustment operation includes an adjustment direction and / or an adjustment gear of the candidate function point; Based on the first sub-prompt word, the vehicle control instruction is determined according to the first target background knowledge, the current perception information, the user intention and the candidate function point.

6. The method according to claim 3, characterized in that The intention label includes a second label, the second label is used to identify the intention ambiguity of the user perception information, the second prompt word includes a second sub-prompt word, the preset knowledge base includes a second preset sub-knowledge base, and the determining of the vehicle control instruction based on the second preset model and the preset knowledge base according to the intention label, the user intention and the candidate function point includes: In the case where the intention label is the second label, the vehicle control instruction is determined based on the second sub-prompt word and the second preset sub-knowledge base according to the user intention and the candidate function points.

7. The method according to claim 6, characterized in that In the case where the intention label is the second label, based on the second sub-prompt word and the second preset sub-knowledge base, according to the user intention and the candidate function point, determining the vehicle control instruction includes: Based on the second preset sub-knowledge base, second target background knowledge is determined according to the candidate function point, wherein the second target background knowledge includes second information of the candidate function point, the second information is used to describe a second adjustment operation of the candidate function point and a corresponding second applicable scenario, and the second adjustment operation includes an adjustment direction and an adjustment amplitude of the candidate function point; Based on the second sub-prompt word, the vehicle control instruction is determined according to the second target background knowledge, the current perception information, the user intention and the candidate function point.

8. The method according to claim 3, characterized in that The preset knowledge base is constructed according to the following steps: Get vehicle function points; Determining, according to the vehicle function point, a third adjustment operation of the vehicle function point and a corresponding third applicable scenario, the third adjustment operation including an adjustment direction and / or an adjustment gear position of the vehicle function point; Determining, according to the vehicle function point, a fourth adjustment operation of the vehicle function point and a corresponding fourth applicable scenario, the fourth adjustment operation including an adjustment direction and an adjustment amplitude of the vehicle function point; The preset knowledge base is constructed according to the third adjustment operation and the corresponding third application scenario, and the fourth adjustment operation and the corresponding fourth application scenario.

9. A server, characterized in that: The server includes a processor and a memory, wherein a computer program is stored in the memory. 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 having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images