Vehicle operation method and device, electronic equipment and storage medium

Abstract

The application discloses a vehicle operation method and device, electronic equipment and storage medium. The method comprises the following steps: obtaining current behavior data of a driver of a vehicle and current vehicle operation data of the vehicle; inputting the current behavior data and the current vehicle operation data into an intention recognition model to obtain a first user intention and a second user intention output by the intention recognition model; wherein the intention recognition model is a model capable of determining the first user intention and the second user intention, the first user intention is used to represent driving demand of the driver, and the second user intention is used to represent driving demand of the vehicle; determining a to-be-executed task of the vehicle according to the first user intention and the second user intention, and controlling the vehicle to operate according to the to-be-executed task to realize the first user intention and the second user intention. That is, the scheme of the application can obtain multiple intentions generated by a user, improve the accuracy of user intention recognition, meet user demand, and improve the use experience of the user.

Description

Technical Field

[0001] This application relates to the field of intelligent vehicle technology, and in particular to a vehicle operation method, device, electronic device and storage medium. Background Technology

[0002] With the development of artificial intelligence and the rise of vehicle intelligence, people have higher and higher requirements for vehicles. In addition to requirements for vehicle safety and comfort, they also have higher requirements for vehicle intelligence.

[0003] Currently, in the in-vehicle systems of existing intelligent vehicles, the user's driving intentions are usually identified through preset commands to obtain the user's needs and plan the driving route.

[0004] However, recognizing a user's driving intention through preset commands can only yield one user intention, which cannot adapt to situations where the user has multiple intentions or the driving environment is complex. This results in poor accuracy in recognizing user intentions, which in turn makes it impossible to meet the user's needs when the vehicle is operated according to the determined user intentions, thus affecting the user's experience. Summary of the Invention

[0005] This application provides a vehicle operation method, device, electronic device, and storage medium to solve the problem that in the prior art, the user's driving intention can only be identified by a preset command, which cannot adapt to situations where the user has multiple intentions or the driving environment is complex. This results in poor accuracy of user intention recognition, which in turn makes it impossible to meet the user's needs when the vehicle is operated according to the determined user intention, thus affecting the user's experience.

[0006] In a first aspect, this application provides a vehicle operation method, the method comprising:

[0007] Acquire the driver's current behavior data and the vehicle's current operating data;

[0008] The current behavior data and the current vehicle operation data are input into the intent recognition model to obtain the first user intent and the second user intent output by the intent recognition model; wherein, the intent recognition model is a pre-trained model for determining the first user intent and the second user intent based on the current behavior data and the current vehicle operation data, the first user intent is used to characterize the driver's driving needs, and the second user intent is used to characterize the vehicle's driving needs.

[0009] The vehicle's task to be performed is determined based on the first user intent and the second user intent, and the vehicle is controlled to operate according to the task to be performed in order to realize the first user intent and the second user intent.

[0010] Secondly, this application provides a vehicle operating device, the device comprising:

[0011] The acquisition module is used to acquire the current behavior data of the driver of the vehicle and the current vehicle operation data;

[0012] An intent determination module is used to input the current behavior data and the current vehicle operation data into an intent recognition model to obtain a first user intent and a second user intent output by the intent recognition model; wherein, the intent recognition model is a pre-trained model used to determine the first user intent and the second user intent based on the current behavior data and the current vehicle operation data, the first user intent is used to characterize the driver's driving needs, and the second user intent is used to characterize the vehicle's driving needs.

[0013] The operation module is used to determine the task to be performed by the vehicle according to the first user intent and the second user intent, and control the vehicle to run according to the task to be performed, so as to realize the first user intent and the second user intent.

[0014] Thirdly, this application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement a vehicle operation method as described in any embodiment of this application.

[0015] Fourthly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the vehicle operation method as described in any embodiment of this application.

[0016] Fifthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the vehicle operation method as described in any embodiment of this application.

[0017] The solution of this application acquires the driver's current behavior data and the vehicle's current operation data; inputs the current behavior data and current vehicle operation data into an intent recognition model to obtain a first user intent and a second user intent output by the intent recognition model; wherein, the intent recognition model is a pre-trained model used to determine the first user intent and the second user intent based on the current behavior data and current vehicle operation data, the first user intent represents the driver's driving needs, and the second user intent represents the vehicle's driving needs; based on the first user intent and the second user intent, the vehicle's task to be executed is determined, and the vehicle is controlled to operate according to the task to be executed, so as to realize the first user intent and the second user intent. That is, the solution of this application, by inputting the current behavior data and current vehicle operation data into the intent recognition model, obtains the first user intent representing driving needs and the second user intent representing driving needs, thereby obtaining multiple intents generated by the user, applicable to complex driving environments, improving the accuracy of user intent recognition, and thus enabling the vehicle to run according to the determined user intent to meet user needs, thereby improving the user experience. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating the vehicle operation method provided in this application;

[0020] Figure 2 This is a schematic diagram of the training process of the intent recognition model for the vehicle operation method provided in this application;

[0021] Figure 3 This is a structural schematic diagram of the vehicle operating device provided in this application;

[0022] Figure 4 This is a schematic diagram of the structure of the electronic device provided in this application. Detailed Implementation

[0023] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0024] Figure 1 This is a flowchart illustrating a vehicle operation method provided in this application. This method can be executed by a vehicle operation device, which can be implemented using software and / or hardware. In a specific embodiment, the device can be applied to an electronic device, which can be an in-vehicle terminal. The following embodiments will illustrate this using the application of the device in an electronic device as an example. Figure 1 The method may specifically include the following steps:

[0025] Step 101: Obtain the driver's current behavior data and the vehicle's current operating data.

[0026] Specifically, the driver's current behavior data refers to data that indicates the driver's driving state, such as the driver's voice data, facial expression data, and body temperature data. The vehicle's current operating data refers to data that indicates the vehicle's driving state, such as vehicle speed data, vehicle status data, and traffic flow data for the road segment the vehicle is traveling on. For example, vehicle status data may include whether the vehicle's tires are leaking air, whether there are any safety hazards, and whether the vehicle's air conditioning is on. The acquisition of the driver's current behavior data and the vehicle's current operating data can be achieved by the onboard terminal executing this embodiment acquiring the data from vehicle status detection equipment and driver status detection equipment installed inside the vehicle. The vehicle status detection equipment may include vehicle status sensors capable of detecting vehicle status, and the driver status detection equipment may include a camera, a microphone, and a driver status detection sensor.

[0027] Optionally, current behavioral data includes driver voice data and image data. Current vehicle operation data includes vehicle status data and driving environment data.

[0028] Specifically, the current behavioral data includes the driver's voice data and image data. Voice data includes the driver's voice and the voices of other passengers. This embodiment can also perform semantic recognition on the driver's voice and the voices of other passengers to determine the driver's state data. Image data includes the driver's facial data; for example, when the driver is fatigued, the driver's facial data can indicate a higher blinking frequency, thus determining the driver's state data based on the image data. Current vehicle operation data includes vehicle status data and driving environment data. Vehicle status data includes data such as whether the vehicle's tires are leaking air, whether there are any safety hazards, whether the vehicle needs refueling, and whether the vehicle's air conditioning is on, indicating whether the vehicle is in a normal driving state. Driving environment data includes data such as the ambient temperature and the traffic flow of the road segment, determining whether the user needs rest.

[0029] Step 102: Input the current behavior data and the current vehicle operation data into the intent recognition model to obtain the first user intent and the second user intent output by the intent recognition model.

[0030] The intent recognition model is a pre-trained model used to determine the first user intent and the second user intent based on the current behavior data and the current vehicle operation data. The first user intent is used to represent the driver's driving needs, and the second user intent is used to represent the vehicle's driving needs.

[0031] Specifically, current behavioral data and current vehicle operation data are input into a pre-trained intent recognition model to obtain the first user intent and the second user intent output by the intent recognition model. The first user intent represents the driver's driving needs, which may include destination requirements and the driver's own state adjustment needs. For example, if the voice information in the car indicates that the driver is hungry and needs to eat lunch, then inputting the current behavioral data into the intent recognition model will yield the first user intent as "the user needs to eat lunch." Similarly, if the image information in the car indicates that the driver is sweating a lot and needs to turn on the air conditioning to cool down, then inputting the current behavioral data into the intent recognition model will yield the first user intent as "the user needs to turn on the air conditioning to cool down." The second user intent represents the vehicle's driving needs. For example, if the vehicle's status detection sensor sends the remaining fuel level to the onboard terminal, and the onboard terminal inputs the remaining fuel level into the intent recognition model, then the second user intent will yield "the vehicle needs to refuel," i.e., the vehicle needs to refuel. If the vehicle's energy type is another type, such as a refueling car or a hydrogen fuel cell vehicle, then the refueling method will be the corresponding refueling method for that energy type.

[0032] The intent recognition model can be located either inside the in-vehicle terminal or on a cloud server that can interact with the in-vehicle terminal; this solution does not impose any limitations on this. For example, the intent recognition model can be deployed on a cloud server, and the in-vehicle terminal connects to the cloud server through a request port to obtain the model's output results. The intent recognition model can also be deployed internally within the in-vehicle terminal, meaning the in-vehicle terminal itself can output the model's results. Alternatively, the intent recognition model can be deployed simultaneously on both the cloud server and the in-vehicle terminal. The training data used for the intent recognition model deployed on the in-vehicle terminal is less than that used for the intent recognition model deployed on the cloud server. Therefore, after inputting data, the output results of the intent recognition model on the cloud server can be used to correct the output results of the intent recognition model on the in-vehicle terminal, thereby enabling the in-vehicle terminal's intent recognition model to undergo reinforcement learning and improve the accuracy of intent recognition. After obtaining the first user intent and the second user intent, the intent recognition model can be further refined based on the user feedback regarding the accuracy of intent recognition, further improving the accuracy of intent recognition.

[0033] Step 103: Determine the task to be performed by the vehicle based on the first user intent and the second user intent, and control the vehicle to run according to the task to be performed, so as to realize the first user intent and the second user intent.

[0034] Specifically, after obtaining the first user intent and the second user intent, the vehicle's task to be executed is determined based on these intents. The task to be executed is a vehicle operation task that, upon execution, satisfies both the first and second user intents. For example, if the first user intent is that the user needs to take a nap, and the second user intent is that the vehicle needs to stop so the user can nap, then the vehicle's task to be executed is to determine a parking location and drive to that location, controlling the vehicle to stop so the user can nap. After determining the task to be executed, the vehicle is controlled to operate according to the task, that is, the vehicle is controlled to complete the task, thereby fulfilling the user's first and second user intents.

[0035] The process of determining the vehicle's task to be executed based on the first user intent and the second user intent can also be obtained from a trained deep learning model. For example, the user's first historical user intent, the user's second historical user intent, and the vehicle's actual task to be executed can be input into the model to be trained to obtain a task determination model that can determine the vehicle's task to be executed based on the first user intent and the second user intent. The model training here can be part of the intent recognition model in this embodiment, that is, the intent recognition model directly outputs the task to be executed, or it can be another separate model used to determine the vehicle's task to be executed based on the first user intent and the second user intent.

[0036] For example, the first user intent is that the user is feeling down and needs relief, and the second user intent is that the vehicle needs to move slowly. Based on the first and second user intents, the tasks to be performed by the vehicle are determined to be playing upbeat songs for the user and controlling the vehicle to move slowly. The vehicle is then controlled to perform the tasks, that is, to play upbeat songs for the user and move slowly, thereby fulfilling the user's first and second user intents.

[0037] Optionally, the first user intent includes a first target location, the second user intent includes a second target location, and the task to be performed includes a driving route. Determining the vehicle's task to be performed based on the first user intent and the second user intent can be achieved through steps 1031 to 1032.

[0038] Step 1031: Obtain the current position of the vehicle, and determine the distance between the current position of the vehicle and the first target position, the distance between the current position of the vehicle and the second target position, and the distance between the first target position and the second target position based on the current position of the vehicle, the first target position and the second target position.

[0039] Specifically, the first user intent includes the first target location to be reached to achieve the first user intent, and the second user intent includes the second target location to be reached to achieve the second user intent. The vehicle's current location is obtained, and based on the vehicle's current location, the first target location, and the second target location, the distances between the vehicle's current location and the first target location, the distances between the vehicle's current location and the second target location, and the distances between the first target location and the second target location are determined to avoid situations where the path to be traveled is not the shortest distance.

[0040] For example, the first target location is a fast food restaurant, the second target location is a gas station, and the distance between the vehicle's current location and the fast food restaurant is determined to be 2km, the distance between the vehicle's current location and the gas station is 3km, and the distance between the fast food restaurant and the gas station is 1km.

[0041] Step 1032: Determine the vehicle's travel path based on the distance between the vehicle's current position and the first target position, the distance between the vehicle's current position and the second target position, and the distance between the first target position and the second target position, so that the travel path includes the vehicle's current position, the first target position, and the second target position, and the travel distance of the travel path is the shortest distance between the vehicle's current position, the first target position, and the second target position.

[0042] Specifically, based on the distances between the vehicle's current position and the first target position, the distances between the vehicle's current position and the second target position, and the distances between the first and second target positions, the shortest distance between the vehicle's current position, the first target position, and the second target position is determined. This completes the determination of the driving path, ensuring that the driving path includes the vehicle's current position, the first target position, and the second target position, and that the driving path is the shortest distance between these three positions. Determining the driving path allows the vehicle to satisfy both the first and second user intentions when traveling along it. After determining the shortest distance, the driving path can be adjusted based on the traffic conditions. For example, if traffic congestion occurs, the driving path can be changed to avoid increasing the driver's driving time due to road congestion.

[0043] For example, the first target location is a fast food restaurant, and the second target location is a gas station. Based on the distance between the vehicle's current location and the fast food restaurant being 2km, the distance between the vehicle's current location and the gas station being 3km, and the distance between the fast food restaurant and the gas station being 1km, the driving route is determined. The driving route is from the current location to the fast food restaurant, and then to the gas station.

[0044] For example, the first user intent is that the user wants to eat lunch, and the second user intent is that the vehicle needs to refuel. The first target location is the location where the user can eat lunch, and the second target location is the location where the vehicle can refuel. Therefore, based on the obtained current vehicle location, the first user intent, and the second user intent, the driving path is determined so that the driving path includes both the location where the user can eat lunch and the location where the vehicle can refuel. The driving path including the location information for lunch and refueling can also be obtained by a trained deep learning model. For example, historical vehicle location information, map information, historical user meal information, and historical vehicle refueling information can be input into the model to be trained, resulting in a model that can determine the driving path including the location information for lunch and refueling based on the vehicle's location information and map information. This model training can be part of the intent recognition model in this embodiment, i.e., the intent recognition model directly outputs the driving path, or it can be a separate model used to output the driving path based on the vehicle's current location and map information.

[0045] The solution of this application acquires the driver's current behavior data and the vehicle's current operation data; inputs the current behavior data and current vehicle operation data into an intent recognition model to obtain a first user intent and a second user intent output by the intent recognition model; wherein, the intent recognition model is a pre-trained model used to determine the first user intent and the second user intent based on the current behavior data and current vehicle operation data, the first user intent represents the driver's driving needs, and the second user intent represents the vehicle's driving needs; based on the first user intent and the second user intent, the vehicle's task to be executed is determined, and the vehicle is controlled to operate according to the task to be executed, so as to realize the first user intent and the second user intent. That is, the solution of this application, by inputting the current behavior data and current vehicle operation data into the intent recognition model, obtains the first user intent representing driving needs and the second user intent representing driving needs, thereby obtaining multiple intents generated by the user, applicable to complex driving environments, improving the accuracy of user intent recognition, and thus enabling the vehicle to run according to the determined user intent to meet user needs, thereby improving the user experience.

[0046] Figure 2 This is a schematic diagram of the training process of the intent recognition model for the vehicle operation method provided in this application. This embodiment... Figure 1 Based on the illustrated embodiments and various optional implementation schemes, the training process of the intent recognition model is described in detail. For example... Figure 2 As shown, the method may include the following steps:

[0047] Step 201: Obtain the driver's first historical behavior data and the vehicle's first historical vehicle operation data.

[0048] Specifically, the first historical behavior data refers to the driver's behavior data within the most recent time period, such as the driver's driving habits, driving actions, facial expressions, and the correlation between driver behavior and time. The first historical vehicle operation data refers to the vehicle's operation data within the most recent time period, such as refueling locations. Both the first historical behavior data and the first historical vehicle operation data undergo data preprocessing, such as data cleaning and normalization, to make the data usable.

[0049] Step 202: Determine training behavior data based on the correlation between the first historical behavior data and the historical first user intent and the historical second user intent.

[0050] Specifically, among the various behavioral data corresponding to the first historical behavioral data, the correlation between each behavioral data and the historical first user intent and the historical second user intent is determined. This correlation can be represented by a correlation coefficient. If the correlation between behavioral data and the historical first user intent is less than a preset correlation coefficient, then the behavioral data is determined to be unrelated to the first user intent. If the correlation between behavioral data and the historical first user intent is greater than or equal to the preset correlation coefficient, then the behavioral data is determined to be related to the first user intent, and this behavioral data is then included in the training behavioral data. If the correlation between behavioral data and the historical second user intent is less than a preset correlation coefficient, then the behavioral data is determined to be unrelated to the second user intent. If the correlation between behavioral data and the historical second user intent is greater than or equal to the preset correlation coefficient, then the behavioral data is determined to be related to the second user intent, and this behavioral data is then included in the training behavioral data. For example, if the behavioral data is voice data generated when a driver sings, and the correlation between this data and both the historical first and historical second user intents is less than the preset correlation coefficient, then this data is determined to be unrelated to both the first and historical second user intents.

[0051] Alternatively, step 202 can be implemented via step 2021.

[0052] Step 2021: The data in the first historical behavior data that both the correlation with the first historical user intent and the correlation with the second historical user intent meet the first correlation requirement are identified as training behavior data.

[0053] For example, the first relevance requirement is that the relevance between the behavioral data and the user's intent is greater than or equal to a preset relevance. If the relevance between the user's facial data in the first historical behavioral data and both the historical first user intent and the historical second user intent is greater than the preset relevance, then the relevance between the user's facial data and both the historical first user intent and the historical second user intent satisfies the first relevance requirement, and therefore the user's facial data is determined as one type of training behavioral data.

[0054] Step 203: Determine the training vehicle operation data based on the correlation between the first historical vehicle operation data and the historical second user intent.

[0055] Specifically, among the various vehicle operation data corresponding to the first historical vehicle operation data, the correlation between each type of vehicle operation data and the historical second user intent is determined. This correlation can be characterized by a correlation coefficient. If the correlation between vehicle operation data and the historical second user intent is less than a preset correlation coefficient, then the vehicle operation data is determined to be unrelated to the second user intent. If the correlation between vehicle operation data and the historical second user intent is greater than or equal to the preset correlation coefficient, then the vehicle operation data is determined to be related to the second user intent, and this vehicle operation data is then selected as one type of training vehicle operation data. For example, if the vehicle operation data is the window opening size, and this data has a correlation with the historical second user intent that is less than the preset correlation coefficient, then this data is determined to be unrelated to the second user intent.

[0056] Alternatively, step 203 can be implemented via step 2031.

[0057] Step 2031: The data whose correlation between the first historical vehicle operation data and the historical second user intent meets the second correlation requirement is determined as training vehicle operation data.

[0058] For example, the second relevance requirement is that the relevance between vehicle operation data and historical second user intent is greater than or equal to a preset relevance. If the relevance between the vehicle fuel level data in the first historical vehicle operation data and the historical second user intent is greater than the preset relevance, then the vehicle fuel level data meets the second relevance requirement, and therefore the vehicle fuel level data is determined as one type of training vehicle operation data.

[0059] Step 204: Train the model to be trained based on the training behavior data and the training vehicle operation data to obtain the intent recognition model.

[0060] Specifically, the training model is trained using training behavior data and training vehicle operation data to obtain an intent recognition model capable of determining the first user intent and the second user intent based on the current behavior data and current vehicle operation data. The training model can be a deep learning model combining a long short-term memory network and a convolutional neural network.

[0061] Optionally, step 204 can be implemented through steps 2041 to 2042.

[0062] Step 2041: Input the training behavior data and training vehicle operation data into the model to be trained to obtain the first training user intent and the second training user intent output by the model to be trained.

[0063] Specifically, training behavior data and training vehicle operation data are input into the model to be trained to obtain the training results output by the model, namely the first training user intent and the second training user intent.

[0064] Step 2042: Train the model to be trained based on the first training user intent, the second training user intent, the historical first user intent, and the historical second user intent until the model to be trained meets the preset training termination condition, and obtain the intent recognition model.

[0065] Specifically, after obtaining the first and second training user intentions, the model output is revised based on the historical first and second user intentions until the model meets the preset training termination conditions. This allows the model to learn user behavior and preferences, ultimately resulting in an intent recognition model. The preset training termination conditions include reaching a preset number of iterations and the model's convergence exceeding a preset convergence level. After obtaining the model, cross-validation can be used to evaluate its performance. The model output is then validated based on the historical first and second user intentions, and model parameters, such as the learning rate, number of layers, and number of neurons, are adjusted based on the validation results. Regularization techniques can also be used during training to prevent overfitting.

[0066] Optionally, after step 2042, steps 41 to 44 may also be performed.

[0067] Step 41: Calculate the model evaluation parameters of the intent recognition model.

[0068] Specifically, model evaluation parameters include precision, recall, and F1 score. The test set includes test behavior data, test vehicle operation data, the first user intent, and the second user intent. The model evaluation parameters for the intent recognition model are calculated based on the test set, specifically the model's precision, recall, and F1 score.

[0069] Step 42: Determine the accuracy of the intent recognition model based on the model evaluation parameters.

[0070] Specifically, the accuracy of the intent recognition model can be determined based on the model evaluation parameters, as can the generalization ability of the intent recognition model.

[0071] Step 43: If the accuracy of the intent recognition model does not meet the preset accuracy requirements, then obtain the driver's second historical behavior data and the vehicle's second historical vehicle operation data.

[0072] Among them, the second historical behavior data is different from the first historical behavior data, and the second historical vehicle operation data is different from the first historical vehicle operation data.

[0073] Specifically, the preset accuracy requirement is that the intent recognition model must have high accuracy to effectively recognize user intent. If the accuracy of the intent recognition model does not meet the preset accuracy requirement, it indicates that the intent recognition model needs further training. At this point, the driver's second historical behavior data and the vehicle's second historical vehicle operation data are acquired. The second historical behavior data is the driver's behavior data within the first time period closest to its end. The second historical vehicle operation data is the vehicle's operation data within the first time period closest to its end.

[0074] Step 44: Use the second historical behavior data as the new first historical behavior data, the second historical vehicle operation data as the new first historical vehicle operation data, and the intent recognition model as the new model to be trained. Return to step 202 until the accuracy of the intent recognition model meets the preset accuracy requirements.

[0075] Specifically, the second historical behavior data is used as the new first historical behavior data, the second historical vehicle operation data is used as the new first historical vehicle operation data, and the intent recognition model is used as the new model to be trained. The intent recognition model is further trained until the accuracy of the intent recognition model meets the preset accuracy requirements, thereby improving the accuracy of the model in recognizing user intent.

[0076] The proposed solution uses historical behavioral data and historical vehicle driving data that are highly correlated with user intent to train the training model, thereby making user intent recognition more accurate. Furthermore, the trained model undergoes performance evaluation, and if the evaluation fails to meet accuracy requirements, the model is further trained, further improving the accuracy of user intent recognition and thus enhancing the user experience.

[0077] Figure 3 This is a schematic diagram of a vehicle operating device provided in this application, which is suitable for executing the vehicle operating method provided in this application. Figure 3 As shown, the device may specifically include:

[0078] The acquisition module 301 is used to acquire the current behavior data of the driver of the vehicle and the current vehicle operation data.

[0079] The intent determination module 302 is used to input the current behavior data and the current vehicle operation data into the intent recognition model to obtain the first user intent and the second user intent output by the intent recognition model; wherein, the intent recognition model is a pre-trained model used to determine the first user intent and the second user intent based on the current behavior data and the current vehicle operation data, the first user intent is used to characterize the driver's driving needs, and the second user intent is used to characterize the vehicle's driving needs.

[0080] The operation module 303 is used to determine the task to be performed by the vehicle according to the first user intention and the second user intention, and control the vehicle to run according to the task to be performed, so as to realize the first user intention and the second user intention.

[0081] In one embodiment, the intent recognition model of the intent determination module 302 is trained by a model training module. The model training module is specifically used to: acquire the driver's first historical behavior data and the vehicle's first historical vehicle operation data; determine training behavior data based on the correlation between the first historical behavior data and historical first user intent and historical second user intent; determine training vehicle operation data based on the correlation between the first historical vehicle operation data and historical second user intent; and train the model to be trained based on the training behavior data and the training vehicle operation data to obtain the intent recognition model.

[0082] In one embodiment, the model training module, in determining training behavior data based on the correlation between the first historical behavior data and historical first user intent and historical second user intent, specifically includes: determining the data in the first historical behavior data whose correlation with both historical first user intent and historical second user intent meets a first correlation requirement as the training behavior data; and determining the training vehicle operation data based on the correlation between the first historical vehicle operation data and historical second user intent includes: determining the data in the first historical vehicle operation data whose correlation with historical second user intent meets a second correlation requirement as the training vehicle operation data.

[0083] In one embodiment, the model training module, in training the model to be trained based on the training behavior data and the training vehicle operation data to obtain the intent recognition model, is specifically configured to: input the training behavior data and the training vehicle operation data into the model to be trained to obtain a first training user intent and a second training user intent output by the model to be trained; train the model to be trained based on the first training user intent, the second training user intent, the historical first user intent, and the historical second user intent until the model to be trained meets a preset training termination condition to obtain the intent recognition model.

[0084] In one embodiment, the model training module is further configured to: after obtaining the intent recognition model, calculate the model evaluation parameters of the intent recognition model; determine the accuracy of the intent recognition model based on the model evaluation parameters; if the accuracy of the intent recognition model does not meet the preset accuracy requirement, obtain the second historical behavior data of the driver and the second historical vehicle operation data of the vehicle; wherein the second historical behavior data is different from the first historical behavior data, and the second historical vehicle operation data is different from the first historical vehicle operation data; use the second historical behavior data as the new first historical behavior data, the second historical vehicle operation data as the new first historical vehicle operation data, and the intent recognition model as the new model to be trained, and return to execute the step of "determining training behavior data based on the correlation between the first historical behavior data and the historical first user intent and the historical second user intent" until the accuracy of the intent recognition model meets the preset accuracy requirement.

[0085] In one embodiment, the current behavior data obtained by the acquisition module 301 includes the driver's voice data and image data; the current vehicle operation data includes the vehicle's status data and driving environment data.

[0086] In one embodiment, the first user intent of the running module 303 includes a first target location, the second user intent includes a second target location, and the task to be executed includes a driving path. Specifically, the running module 303, in determining the task to be executed for the vehicle based on the first user intent and the second user intent, is configured to: obtain the current location of the vehicle, and determine the distance between the current location of the vehicle and the first target location, the distance between the current location of the vehicle and the second target location, and the distance between the first target location and the second target location based on the current location of the vehicle, the first target location, and the second target location; determine the driving path for the vehicle based on the distance between the current location of the vehicle and the first target location, the distance between the current location of the vehicle and the second target location, and the distance between the first target location and the second target location, such that the driving path includes the current location of the vehicle, the first target location, and the second target location, and the driving distance of the driving path is the shortest distance between the current location of the vehicle, the first target location, and the second target location.

[0087] The apparatus of this application acquires the driver's current behavior data and the vehicle's current operating data; inputs the current behavior data and the current vehicle operating data into an intent recognition model to obtain a first user intent and a second user intent output by the intent recognition model; wherein, the intent recognition model is a pre-trained model used to determine the first user intent and the second user intent based on the current behavior data and the current vehicle operating data, the first user intent represents the driver's driving needs, and the second user intent represents the vehicle's driving needs; based on the first user intent and the second user intent, the apparatus determines the vehicle's task to be executed and controls the vehicle to operate according to the task to be executed, so as to realize the first user intent and the second user intent. In other words, the solution of this application, by inputting the current behavior data and the current vehicle operating data into the intent recognition model, obtains the first user intent representing driving needs and the second user intent representing driving needs, thereby obtaining multiple intents generated by the user, applicable to complex driving environments, improving the accuracy of user intent recognition, and thus enabling the vehicle to meet user needs when operating according to the determined user intent, thereby improving the user experience.

[0088] This application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the vehicle operation method provided in any of the above embodiments.

[0089] This application also provides a computer-readable medium having a computer program stored thereon, which, when executed by a processor, implements the vehicle operation method provided in any of the above embodiments.

[0090] The following is for reference. Figure 4 It shows a schematic diagram of the structure of an electronic device 400 suitable for implementing the present application. Figure 4 The electronic device shown is merely an example and should not impose any limitations on the functionality and scope of this application.

[0091] like Figure 4 As shown, the electronic device 400 includes a central processing unit (CPU) 401, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 402 or a program loaded from a storage section 408 into a random access memory (RAM) 403. The RAM 403 also stores various programs and data required for the operation of the electronic device 400. The CPU 401, ROM 402, and RAM 403 are interconnected via a bus 404. An input / output (I / O) interface 405 is also connected to the bus 404.

[0092] The following components are connected to I / O interface 405: an input section 406 including a keyboard, mouse, etc.; an output section 407 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and speakers, etc.; a storage section 408 including a hard disk, etc.; and a communication section 409 including a network interface card such as a LAN card, modem, etc. The communication section 409 performs communication processing via a network such as the Internet. A drive 410 is also connected to I / O interface 405 as needed. A removable medium 411, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on drive 410 as needed so that computer programs read from it can be installed into storage section 408 as needed.

[0093] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 409, and / or installed from removable medium 411. When the computer program is executed by central processing unit (CPU) 401, it performs the functions defined above in the system of this application.

[0094] It should be noted that the computer-readable medium shown in this application can be a computer-readable signal medium or a computer-readable storage medium, or any combination of the two. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this application, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In this application, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media can also be any computer-readable medium other than computer-readable storage media, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wireless, wire, optical fiber, RF, etc., or any suitable combination thereof.

[0095] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0096] The modules and / or units described in this application can be implemented in software or hardware. The described modules and / or units can also be housed in a processor; for example, a processor may be described as including an acquisition module, an intent determination module, and an execution module. The names of these modules do not necessarily limit the module itself.

[0097] In another aspect, this application also provides a computer-readable medium, which may be included in the device described in the above embodiments; or it may exist independently and not assembled into the device. The computer-readable medium carries one or more programs that, when executed by the device, cause the device to perform the following operations:

[0098] The system acquires the driver's current behavior data and the vehicle's current operation data; it inputs these data into an intent recognition model to obtain a first user intent and a second user intent output by the model. The intent recognition model is a pre-trained model used to determine the first and second user intents based on the current behavior and vehicle operation data. The first user intent represents the driver's driving needs, and the second user intent represents the vehicle's driving needs. Based on the first and second user intents, the system determines the vehicle's task to be executed and controls the vehicle to operate according to the task, thereby realizing the first and second user intents.

[0099] According to the technical solution of this application, the current behavior data of the driver and the current vehicle operation data are obtained; the current behavior data and the current vehicle operation data are input into an intent recognition model to obtain a first user intent and a second user intent output by the intent recognition model; wherein, the intent recognition model is a pre-trained model used to determine the first user intent and the second user intent based on the current behavior data and the current vehicle operation data, the first user intent is used to represent the driver's driving needs, and the second user intent is used to represent the vehicle's driving needs; the vehicle's task to be executed is determined based on the first user intent and the second user intent, and the vehicle is controlled to run according to the task to be executed, so as to realize the first user intent and the second user intent. That is, the solution of this application, by inputting the current behavior data and the current vehicle operation data into the intent recognition model, obtains the first user intent representing driving needs and the second user intent representing driving needs, thereby obtaining multiple intents generated by the user, applicable to complex driving environments, improving the accuracy of user intent recognition, and thus enabling the vehicle to run according to the determined user intent to meet user needs, thereby improving the user experience.

[0100] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the vehicle operation method provided in any embodiment of this application.

[0101] In the implementation of the computer program product, computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof. Programming languages ​​include object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0102] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this application can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this application can be achieved, and this is not limited herein.

[0103] The specific embodiments described above do not constitute a limitation on the scope of protection of this application. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can occur depending on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for operating a vehicle, characterized in that, The method includes: Acquire the driver's current behavior data and the vehicle's current operating data; The current behavior data and the current vehicle operation data are input into the intent recognition model to obtain the first user intent and the second user intent output by the intent recognition model; wherein, the intent recognition model is a pre-trained model for determining the first user intent and the second user intent based on the current behavior data and the current vehicle operation data, the first user intent is used to characterize the driver's driving needs, and the second user intent is used to characterize the vehicle's driving needs. The intent recognition model is trained using the following method: acquiring the driver's first historical behavior data and the vehicle's first historical vehicle operation data; identifying data from the first historical behavior data whose correlation with both the historical first user intent and the historical second user intent meets a first correlation requirement as training behavior data; identifying data from the first historical vehicle operation data whose correlation with the historical second user intent meets a second correlation requirement as training vehicle operation data; and training the model to be trained based on the training behavior data and the training vehicle operation data to obtain the intent recognition model. The vehicle's task to be performed is determined based on the first user intent and the second user intent, and the vehicle is controlled to operate according to the task to be performed in order to realize the first user intent and the second user intent.

2. The method according to claim 1, characterized in that, The step of training the model to be trained based on the training behavior data and the training vehicle operation data to obtain the intent recognition model includes: The training behavior data and the training vehicle operation data are input into the model to be trained to obtain the first training user intent and the second training user intent output by the model to be trained. The training model is trained based on the first training user intent, the second training user intent, the historical first user intent, and the historical second user intent until the training model meets the preset training termination condition, thereby obtaining the intent recognition model.

3. The method according to claim 2, characterized in that, After obtaining the intent recognition model, the method further includes: Calculate the model evaluation parameters of the intent recognition model; The accuracy of the intent recognition model is determined based on the model evaluation parameters. If the accuracy of the intent recognition model does not meet the preset accuracy requirement, then the second historical behavior data of the driver and the second historical vehicle operation data of the vehicle are obtained; wherein, the second historical behavior data is different from the first historical behavior data, and the second historical vehicle operation data is different from the first historical vehicle operation data; The second historical behavior data is used as the new first historical behavior data, the second historical vehicle operation data is used as the new first historical vehicle operation data, and the intent recognition model is used as the new model to be trained. The process of "determining training behavior data based on the correlation between the first historical behavior data and the historical first user intent and the historical second user intent" is repeated until the accuracy of the intent recognition model meets the preset accuracy requirements.

4. The method according to claim 1, characterized in that, The current behavior data includes the driver's voice data and image data; The current vehicle operation data includes the vehicle status data and driving environment data.

5. The method according to claim 1, characterized in that, The first user intent includes a first target location, the second user intent includes a second target location, the task to be performed includes a driving route, and determining the task to be performed by the vehicle based on the first user intent and the second user intent includes: The current position of the vehicle is obtained, and based on the current position of the vehicle, the first target position, and the second target position, the distance between the current position of the vehicle and the first target position, the distance between the current position of the vehicle and the second target position, and the distance between the first target position and the second target position are determined. Based on the distance between the vehicle's current position and the first target position, the distance between the vehicle's current position and the second target position, and the distance between the first target position and the second target position, a driving path for the vehicle is determined, such that the driving path includes the vehicle's current position, the first target position, and the second target position, and the driving distance of the driving path is the shortest distance between the vehicle's current position, the first target position, and the second target position.

6. A vehicle operating device, characterized in that, The device includes: The acquisition module is used to acquire the current behavior data of the driver of the vehicle and the current vehicle operation data; An intent determination module is used to input the current behavior data and the current vehicle operation data into an intent recognition model to obtain a first user intent and a second user intent output by the intent recognition model; wherein, the intent recognition model is a pre-trained model used to determine the first user intent and the second user intent based on the current behavior data and the current vehicle operation data, the first user intent is used to characterize the driver's driving needs, and the second user intent is used to characterize the vehicle's driving needs. The intent recognition model in the intent determination module is trained by the model training module. Specifically, the model training module is used to: acquire the driver's first historical behavior data and the vehicle's first historical vehicle operation data; determine the data in the first historical behavior data whose correlation with both the historical first user intent and the historical second user intent meets a first correlation requirement as training behavior data; determine the data in the first historical vehicle operation data whose correlation with the historical second user intent meets a second correlation requirement as training vehicle operation data; and train the model to be trained based on the training behavior data and the training vehicle operation data to obtain the intent recognition model. The operation module is used to determine the task to be performed by the vehicle according to the first user intent and the second user intent, and control the vehicle to run according to the task to be performed, so as to realize the first user intent and the second user intent.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the vehicle operation method as described in any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the vehicle operation method as described in any one of claims 1 to 5.

Images