Vehicle control method and device, vehicle and electronic equipment

Abstract

According to the vehicle control method and device, the vehicle and the electronic equipment provided by the invention, the vehicle environment information of the target vehicle and the feature information of the monitored object in the target vehicle are obtained, and the vehicle environment information and the feature information are subjected to feature extraction processing to obtain the standard feature data; inputting the standard feature data into a pre-trained scene recognition model, respectively calculating the similarity between the standard feature data and each piece of preset scene data, and determining target scene data corresponding to the standard feature data according to the similarity; and generating a corresponding control instruction according to the vehicle control parameters contained in the target scene data, and performing vehicle control on the target vehicle according to the control instruction. Compared with the prior art, the driving scene of the vehicle can be automatically recognized through the pre-trained scene recognition model, the control instruction is generated to control the target vehicle, and various parameters of the vehicle can be automatically adjusted according to different driving environments.

Description

Technical Field

[0001] The present disclosure relates to the field of vehicle processing technology, and in particular to a vehicle control method and device, a vehicle, and an electronic device. Background Art

[0002] With the development of the automotive industry and the improvement of people's living standards, more and more people are using cars, and vehicle comfort is gaining increasing attention. In-car air quality significantly affects passenger comfort, and vehicle air conditioning is becoming increasingly important as a system for enhancing passenger comfort. The automotive air conditioning system (AC) is the system that regulates the air inside the vehicle.

[0003] The rapid development of electronic technology has made vehicle control systems more intelligent. Modern vehicle air conditioning is already controlled by computers. A complete vehicle computer-controlled air conditioning system can not only adjust the temperature, humidity, cleanliness, air volume and direction of the air in the car to ensure that passengers are in a comfortable air environment under various external climates and conditions, but also perform fault detection.

[0004] However, the air conditioning control method in the existing technology requires manual adjustment of temperature, wind speed and wind direction. When the driver is in different driving environments (such as cities, suburbs, highways, etc.), the air conditioning parameters need to be frequently adjusted, which increases the driver's operating burden. Therefore, how to automatically adjust the air conditioning parameters according to different driving environments is a problem that needs to be solved urgently. Summary of the Invention

[0005] The present disclosure provides a vehicle control method and device, a vehicle, and an electronic device, the main purpose of which is to automatically adjust air conditioning parameters according to different driving environments.

[0006] According to a first aspect of the present disclosure, a vehicle control method is provided, comprising:

[0007] Acquiring vehicle environment information of a target vehicle and feature information of a monitored object in the target vehicle, and performing feature extraction processing on the vehicle environment information and the feature information to obtain standard feature data;

[0008] Inputting the standard feature data into a pre-trained scenario recognition model, calculating the similarity between the standard feature data and each preset scenario data, and determining the target scenario data corresponding to the standard feature data based on the similarity; wherein the preset scenario data is pre-configured driving scenario data including vehicle control parameters, and different preset scenario data include different vehicle control parameters;

[0009] A corresponding control instruction is generated according to the vehicle control parameters included in the target scenario data, and the target vehicle is controlled according to the control instruction.

[0010] Optionally, the performing feature extraction processing on the vehicle environment information and the feature information to obtain standard feature data includes:

[0011] Encoding the vehicle environment information and the characteristic information respectively using a preset encoding method to obtain encoded information data;

[0012] Standardizing the coded information data using a preset standardized format to obtain the standard information data;

[0013] The standard information data is subjected to feature extraction processing by a preset feature selection algorithm to obtain the standard feature data.

[0014] Optionally, after controlling the target vehicle according to the control instruction, the method further includes:

[0015] Acquiring status information of the monitored object and a vehicle control record of the monitored object controlling the target vehicle;

[0016] Performing data extraction processing based on the state information and the vehicle control record to obtain update scenario data;

[0017] Optimizing the model parameters in the pre-trained scenario recognition model using a preset optimization algorithm according to the target scenario data and the update scenario data to obtain updated model parameters;

[0018] The updated model parameters are configured in the pre-trained scenario recognition model.

[0019] Optionally, before obtaining the vehicle environment information of the target vehicle and the characteristic information of the monitored object in the target vehicle, the method further includes:

[0020] Acquire a preset amount of vehicle usage data; wherein the vehicle usage data includes training vehicle environment information, training feature information, and training label scenario data;

[0021] Performing feature extraction processing on the training vehicle environment information and the training feature information to obtain training standard feature data, and dividing the training standard feature data into a training feature data set and a test feature data set by a preset division method;

[0022] Inputting the training feature data set into the scenario recognition model for model training to obtain trained model parameters, and optimizing the scenario recognition model using a preset loss function based on the trained model parameters and the training label scenario data to obtain a trained scenario recognition model;

[0023] performing model evaluation processing on the trained scenario recognition model using the test feature data set to obtain a model evaluation parameter, and continuing model training on the scenario recognition model if it is determined that the model evaluation parameter is less than or equal to a preset evaluation threshold;

[0024] When it is determined that the model evaluation parameter is greater than the preset evaluation threshold, the pre-trained scenario recognition model is obtained.

[0025] Optionally, the step of inputting the training feature dataset into a scenario recognition model for model training to obtain trained model parameters includes:

[0026] Obtaining the number of features in the training feature dataset and calculating the feature variance of each feature;

[0027] Perform parameter calculation processing based on the number of features and the feature variance to obtain model parameters;

[0028] The model parameters are verified using a preset verification algorithm to obtain the trained model parameters.

[0029] Optionally, determining the target scenario data corresponding to the standard feature data according to the similarity includes:

[0030] Calculating a probability value of the standard feature data belonging to each of the preset scenario data by using the similarity and the preset weights in the pre-trained scenario recognition model;

[0031] In each of the preset scenario data, the preset scenario data having a higher probability value of containing the standard feature data is selected as the target scenario data.

[0032] Optionally, after controlling the target vehicle according to the control instruction, the method further includes:

[0033] Obtaining execution information of the control instruction;

[0034] When it is determined according to the execution information that the control instruction fails to be executed, re-execute the control instruction and obtain the number of failures of the control instruction execution;

[0035] When the number of failures exceeds a preset number threshold, acquiring the vehicle environment information and the characteristic information;

[0036] When it is determined according to the execution information that the control instruction is successfully executed, the vehicle environment information and the feature information are continued to be acquired.

[0037] According to a second aspect of the present disclosure, there is provided a vehicle control device, comprising:

[0038] A first acquisition unit is used to acquire vehicle environment information of a target vehicle and feature information of a monitored object in the target vehicle;

[0039] a first extraction unit, configured to perform feature extraction processing on the vehicle environment information and the feature information to obtain standard feature data;

[0040] a calculation unit, configured to input the standard feature data into a pre-trained scenario recognition model and respectively calculate the similarity between the standard feature data and each preset scenario data;

[0041] A first determining unit is configured to determine target scenario data corresponding to the standard feature data based on the similarity; wherein the preset scenario data is pre-configured driving scenario data including vehicle control parameters, and different preset scenario data include different vehicle control parameters;

[0042] a generating unit, configured to generate corresponding control instructions according to the vehicle control parameters included in the target scenario data;

[0043] A control unit is used to control the target vehicle according to the control instruction.

[0044] Optionally, the first extraction unit includes:

[0045] An encoding module, configured to encode the vehicle environment information and the characteristic information respectively using a preset encoding method to obtain encoded information data;

[0046] a processing module, configured to perform standardization processing on the coded information data using a preset standardized format to obtain the standard information data;

[0047] The extraction module is used to perform feature extraction processing on the standard information data through a preset feature selection algorithm to obtain the standard feature data.

[0048] Optionally, the first acquiring unit is further configured to acquire status information of the monitored object and a vehicle control record of the monitored object controlling the target vehicle;

[0049] The first extraction unit is further configured to perform data extraction processing based on the state information and the vehicle control record to obtain update scenario data;

[0050] The device further comprises:

[0051] an optimization unit, configured to optimize the model parameters of the pre-trained scenario recognition model using a preset optimization algorithm according to the target scenario data and the update scenario data to obtain updated model parameters;

[0052] A configuration unit is used to configure the updated model parameters in the pre-trained scenario recognition model.

[0053] Optionally, the device further includes:

[0054] A second acquisition unit is configured to acquire a preset amount of vehicle usage data; wherein the vehicle usage data includes training vehicle environment information, training feature information, and training label scenario data;

[0055] a second extraction unit, configured to perform feature extraction processing on the training vehicle environment information and the training feature information to obtain standard feature data for training;

[0056] A division unit, configured to divide the training standard feature data into a training feature data set and a test feature data set by a preset division method;

[0057] a training unit, configured to input the training feature data set into a scenario recognition model for model training to obtain trained model parameters, and optimize the scenario recognition model using a preset loss function based on the trained model parameters and the training labeled scenario data to obtain a trained scenario recognition model;

[0058] an evaluation unit, configured to perform a model evaluation process on the trained scenario recognition model using the test feature data set to obtain model evaluation parameters;

[0059] The training unit is further configured to, when determining that the model evaluation parameter is less than or equal to a preset evaluation threshold, continue to perform model training on the scenario recognition model;

[0060] The second determining unit is configured to obtain the pre-trained scenario recognition model when it is determined that the model evaluation parameter is greater than the preset evaluation threshold.

[0061] Optionally, the training unit includes:

[0062] An acquisition module, configured to acquire the number of features in the training feature dataset;

[0063] Calculation module, used to calculate the feature variance of each feature;

[0064] The calculation module is further configured to perform parameter calculation processing according to the number of features and the feature variance to obtain model parameters;

[0065] The verification module is used to verify the model parameters through a preset verification algorithm to obtain the trained model parameters.

[0066] Optionally, the first determining unit is further configured to:

[0067] Calculating a probability value of the standard feature data belonging to each of the preset scenario data by using the similarity and the preset weights in the pre-trained scenario recognition model;

[0068] In each of the preset scenario data, the preset scenario data having a higher probability value of containing the standard feature data is selected as the target scenario data.

[0069] Optionally, the first acquiring unit is further configured to acquire execution information of the control instruction;

[0070] The control unit is further configured to, when it is determined according to the execution information that the execution of the control instruction has failed, re-execute the control instruction and obtain the number of failures of the execution of the control instruction;

[0071] The first acquisition unit is further configured to acquire the vehicle environment information and the feature information when the number of failures exceeds a preset number threshold;

[0072] The first acquisition unit is further configured to, when it is determined according to the execution information that the control instruction is successfully executed, continue to acquire the vehicle environment information and the feature information.

[0073] According to a third aspect of the present disclosure, a vehicle is provided, wherein the vehicle includes the vehicle control device as described in the second aspect of the present disclosure.

[0074] According to a fourth aspect of the present disclosure, there is provided an electronic device, including:

[0075] at least one processor; and

[0076] a memory communicatively connected to the at least one processor; wherein,

[0077] The memory stores instructions that can be executed by the at least one processor. The instructions are executed by the at least one processor to enable the at least one processor to perform the method described in the first aspect.

[0078] According to a fifth aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable the computer to execute the method described in the first aspect.

[0079] According to a sixth aspect of the present disclosure, a computer program product is provided, comprising a computer program, wherein when the computer program is executed by a processor, the computer program implements the method as described in the first aspect above.

[0080] The vehicle control method and device, vehicle and electronic device provided by the present disclosure obtain vehicle environment information of a target vehicle and feature information of a monitored object in the target vehicle, and perform feature extraction processing on the vehicle environment information and the feature information to obtain standard feature data; the standard feature data is input into a pre-trained scenario recognition model, and the similarity between the standard feature data and each preset scenario data is calculated respectively, and the target scenario data corresponding to the standard feature data is determined based on the similarity; wherein the preset scenario data is pre-configured driving scenario data containing vehicle control parameters, and different preset scenario data contain different vehicle control parameters; corresponding control instructions are generated based on the vehicle control parameters contained in the target scenario data, and the target vehicle is controlled based on the control instructions. Compared with the related art, the embodiment of the present disclosure can automatically identify the driving scenario of the vehicle through the pre-trained scenario recognition model, and generate control instructions to control the target vehicle based on the vehicle control parameters in the driving scenario, and can automatically adjust various parameters of the vehicle according to different driving environments.

[0081] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0082] The accompanying drawings are provided to facilitate a better understanding of the present invention and do not constitute a limitation of the present disclosure.

[0083] Figure 1 A flow chart of a vehicle control method provided by an embodiment of the present disclosure;

[0084] Figure 2 A schematic diagram of a feature extraction process provided by an embodiment of the present disclosure;

[0085] Figure 3 A flowchart of a training method for a scenario recognition model provided by an embodiment of the present disclosure;

[0086] Figure 4A flowchart of a training process provided by an embodiment of the present disclosure;

[0087] Figure 5 A schematic diagram of the principle of a vehicle control method provided by an embodiment of the present disclosure;

[0088] Figure 6 A schematic structural diagram of a vehicle control device provided by an embodiment of the present disclosure;

[0089] Figure 7 A schematic structural diagram of another vehicle control device provided by an embodiment of the present disclosure;

[0090] Figure 8 A schematic block diagram of an electronic device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0091] The following description of exemplary embodiments of the present disclosure is made in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding. These details should be considered as merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0092] The following describes a vehicle control method and apparatus, a vehicle, and an electronic device according to embodiments of the present disclosure with reference to the accompanying drawings.

[0093] Figure 1 A flowchart of a vehicle control method provided by an embodiment of the present disclosure.

[0094] like Figure 1 As shown, the method comprises the following steps:

[0095] Step 101 : Acquire vehicle environment information of a target vehicle and feature information of a monitored object in the target vehicle, and perform feature extraction processing on the vehicle environment information and the feature information to obtain standard feature data.

[0096] The vehicle control described in the embodiments of the present disclosure includes but is not limited to: control of vehicle air conditioning, control of vehicle seats, etc. Specifically, the embodiments of the present disclosure do not limit the specific components of the vehicle control.

[0097] To facilitate understanding of the embodiments of the present disclosure, the following description will be given by taking the control of a vehicle air-conditioning system as an example.

[0098] In the embodiment of the present disclosure, the vehicle environmental information includes but is not limited to: data such as temperature, humidity, air quality inside and outside the vehicle, and data such as vehicle speed, time, and vehicle location. The characteristic information includes but is not limited to: characteristic data such as gender, age, clothing, physical condition of the driver and passengers in the vehicle, and data such as the interaction history between the passengers in the vehicle and the air conditioner. The vehicle environmental information is mainly obtained through sensors configured in the vehicle (for example: temperature sensors, humidity sensors, speed sensors, etc.). The method of obtaining the characteristic information includes but is not limited to: linking smart bracelets, watches, steering wheels and other human body monitoring devices containing biometric technology and in-vehicle camera equipment.

[0099] After obtaining the vehicle environment information and feature information, the vehicle environment information and feature information will be sent to the on-board computing unit for data preprocessing. The preprocessing methods include but are not limited to: data cleaning, formatting, etc. Preprocessing can remove duplicate data and erroneous data in the vehicle environment information and feature information.

[0100] After data preprocessing, the vehicle environment information and feature information undergo further data processing, namely feature extraction. The steps of feature extraction include, but are not limited to: converting all data in the vehicle environment information and feature information into feature vectors that can be used for machine learning; normalizing and standardizing all feature vectors to obtain standard data; and performing feature selection on the standard data to obtain standard feature data. The standard feature data contains features from multiple data conversions.

[0101] Step 102: Input the standard feature data into a pre-trained scenario recognition model, calculate the similarity between the standard feature data and each preset scenario data, and determine the target scenario data corresponding to the standard feature data based on the similarity; wherein the preset scenario data is pre-configured driving scenario data containing vehicle control parameters, and different preset scenario data contain different vehicle control parameters.

[0102] In the embodiment of the present disclosure, the preset scenario data is pre-customized and configured based on typical driving scenarios, such as: daily commuting, suburban travel, high-speed driving, "morning start mode", "rainy day mode", "summer afternoon mode", sweating after exercise, different levels of fatigue driving, pregnant women in the car, children in the car, patients in the car, etc. Different scenario modes are defined. Each preset scenario data includes vehicle control parameters. For example: in the summer afternoon mode, the vehicle control parameters are air-conditioning temperature at 22°C, etc. It should be noted that when determining the target scenario data, more than one scenario data can be determined, that is, multiple target scenario data can appear.

[0103] The types of pre-trained scenario recognition models include but are not limited to: Support Vector Machine (SVM) model, etc. When performing similarity calculation, the calculation can be performed by, but not limited to: Radial Basis Function (RBF) of the SVM model as the kernel function. When determining the target scenario data based on the similarity, it is necessary to make a decision based on the similarity and the learned weights, and determine the probability that the standard feature data belongs to the corresponding preset scenario data, and use the preset scenario data with a higher probability as the target scenario data corresponding to the standard feature data. For example, the preset scenario data determined according to the model are the probabilities of the preset scenario data corresponding to the vehicle's air-conditioning control parameters being "low", "medium", and "high", and the following probabilities (hypothetical values) are given respectively:

[0104] Low: 0.1

[0105] Medium: 0.7

[0106] High: 0.2

[0107] Since the probability of "medium" is the highest, the target scenario data is the preset scenario data in which the vehicle's air conditioning control parameter is "medium".

[0108] Step 103 : generating corresponding control instructions according to the vehicle control parameters included in the target scenario data, and controlling the target vehicle according to the control instructions.

[0109] In the embodiment of the present disclosure, when generating control instructions, it is carried out through a preset scenario mode environmental control system based on a large model. The scenario mode environmental control system can output sequence instructions corresponding to the vehicle control parameters contained in the scenario mode, and intelligently control the vehicle's air-conditioning system according to the sequence instructions to achieve the purpose of environmental control.

[0110] Regarding generating corresponding control instructions based on the vehicle control parameters, including but not limited to: the vehicle control parameter is the air-conditioning temperature at 22°C, the control instruction generated is to adjust the vehicle's air-conditioning temperature to 22°C. It should be noted that since multiple target scenario data may appear when determining the target scenario data, multiple control instructions will appear. When multiple control instructions appear, all control instructions will be executed at the same time. For example: there are three target scenario data, including three vehicle control parameters: the air-conditioning mode is cooling, the air-conditioning temperature is 23°C, and the air-conditioning wind speed is medium wind speed. The control instructions generated at this time are: adjusting the vehicle's air-conditioning mode to cooling, adjusting the vehicle's air-conditioning temperature to 23°C, and adjusting the vehicle's air-conditioning wind speed to medium wind speed. When executing the instructions, the three instructions will be executed together.

[0111] The vehicle control method provided by the present disclosure obtains vehicle environment information of a target vehicle and feature information of a monitored object in the target vehicle, and performs feature extraction processing on the vehicle environment information and the feature information to obtain standard feature data; the standard feature data is input into a pre-trained scenario recognition model, and the similarity between the standard feature data and each preset scenario data is calculated respectively, and the target scenario data corresponding to the standard feature data is determined based on the similarity; wherein the preset scenario data is pre-configured driving scenario data containing vehicle control parameters, and different preset scenario data contain different vehicle control parameters; corresponding control instructions are generated based on the vehicle control parameters contained in the target scenario data, and the target vehicle is controlled based on the control instructions. Compared with the related art, the embodiment of the present disclosure can automatically identify the driving scenario of the vehicle through the pre-trained scenario recognition model, and generate control instructions to control the target vehicle based on the vehicle control parameters in the driving scenario, and can automatically adjust various parameters of the vehicle according to different driving environments.

[0112] In one possible implementation of the embodiment of the present disclosure, as a refinement of the above step 101, regarding the feature extraction processing of the vehicle environment information and the feature information, the embodiment of the present disclosure provides a flow chart of feature extraction, such as Figure 2 As shown, including:

[0113] Step 201 : Encode the vehicle environment information and the characteristic information respectively using a preset encoding method to obtain encoded information data.

[0114] In the embodiment of the present disclosure, the preset encoding method is a custom-selected encoding method, such as binary encoding, geographic encoding, etc. Specifically, the embodiment of the present disclosure does not limit the preset encoding method.

[0115] Encoding the vehicle environment information and the feature information separately means: converting the various data in the vehicle environment information and the feature information into feature vectors that can be used for machine learning, for example: time: converted into binary code of the hour of the day or the day of the week (8.00am can be converted into the time period of the day: 0.25); longitude and latitude data can be converted into geographic code or distance from a specific location (30km from the city center can be converted into binary features, urban area = 1, suburbs = 0); vehicle speed: directly as a numerical feature; outside temperature and inside temperature: directly as numerical features; number and distribution of passengers: number as a numerical feature, distribution can be encoded as the presence or absence of passengers in different areas in the car; historical passenger preferences: such as temperature preferences, wind volume preferences, etc. are encoded as features; current weather conditions: weather conditions (sunny, rainy, snowy, etc.) are encoded as category features.

[0116] Step 202: Standardize the coded information data using a preset standardized format to obtain the standard information data.

[0117] In the embodiment of the present disclosure, the preset normalization format is a custom-selected normalization format, such as Z-score normalization, Min-Max normalization, etc. Specifically, the embodiment of the present disclosure does not limit the preset normalization format.

[0118] Normalization of the coded information data refers to scaling the features of the data to a uniform scale. For example, a vehicle speed of 80 km / h can be normalized to: \frac{80-0}{200-0}=0.4, and an outside temperature of 5°C can be normalized to: \frac{5-(-20)}{40-(-20)}=0.357.

[0119] Step 203: Perform feature extraction processing on the standard information data using a preset feature selection algorithm to obtain the standard feature data.

[0120] In the embodiment of the present disclosure, the preset feature selection algorithm is a custom-selected algorithm, such as recursive feature elimination (RFE), model-based feature selection or univariate statistical test, etc. Specifically, the embodiment of the present disclosure does not limit the preset feature selection algorithm.

[0121] By performing feature extraction processing on the standard information data, only the most useful features for scene mode determination can be retained, thereby improving the accuracy and efficiency of scene mode determination.

[0122] In one implementable manner of the embodiment of the present disclosure, a pre-trained scenario recognition model will provide a more comfortable driving environment for the vehicle and can be continuously optimized as more data is collected. In order to optimize the pre-trained scenario recognition model, the following methods may also be used but are not limited to: obtaining the status information of the monitored object and the vehicle control record of the monitored object controlling the target vehicle; performing data extraction processing based on the status information and the vehicle control record to obtain updated scenario data; optimizing the model parameters in the pre-trained scenario recognition model through a preset optimization algorithm based on the target scenario data and the updated scenario data to obtain updated model parameters; and configuring the updated model parameters in the pre-trained scenario recognition model.

[0123] In the embodiment of the present disclosure, the preset optimization algorithm is a custom-set algorithm, such as the Stochastic Gradient Descent (SGD) algorithm, the Elastic Weight Consolidation (EWC) algorithm, etc. The status information includes but is not limited to the posture and facial expression feedback of the occupants of the vehicle, and the vehicle control records include but are not limited to driving habits, preferences for frequent adjustment of air-conditioning settings, vehicle adjustment content, etc., wherein the basis for updating the pre-trained scenario recognition model according to the status information and vehicle control records is: when the expressions of the occupants of the vehicle show negative emotions or the posture of the occupants of the vehicle show abnormal conditions, it indicates that the control of the vehicle is inappropriate, and the pre-trained scenario recognition model needs to be further updated according to the vehicle control records.

[0124] The updating scenario data is obtained through vehicle control records. For example, after the vehicle is controlled by a control instruction, the status information shows that the occupants of the vehicle begin to sweat, and the vehicle control record indicates that the occupants of the vehicle have lowered the air-conditioning temperature. Then, the updating scenario data is the parameters after the air-conditioning temperature is lowered. After the pre-trained scenario recognition model is optimized by the updating scenario data, the next time the vehicle is automatically adjusted, the pre-trained scenario recognition model will indicate in the output control instruction that the air-conditioning temperature should be lowered a little.

[0125] In one possible implementation of the embodiment of the present disclosure, the pre-trained scenario recognition model is obtained by training the scenario recognition model. Regarding the training of the scenario recognition model, the embodiment of the present disclosure provides a flow chart of a scenario recognition model training method, as shown in FIG. Figure 3 As shown, including:

[0126] Step 301 : Acquire a preset amount of vehicle usage data; wherein the vehicle usage data includes training vehicle environment information, training feature information, and training label scenario data.

[0127] In the embodiment of the present disclosure, the preset number is a number of custom settings, for example, 2500, 3000, etc. Specifically, the embodiment of the present disclosure does not limit the preset number.

[0128] A vehicle usage data set represents the process from when the vehicle is started to when it is shut down. At the same time, the vehicle usage data set includes all data collected during the driving process, namely, vehicle environment information for training and feature information for training, such as: vehicle speed (0-200 km / h); outside temperature (-20°C to 40°C); inside temperature (15°C to 30°C); time (24-hour format, for example, 8.00 represents 8:00 am); location (distance relative to the city center, in km); number of passengers (1 to 5); passenger interaction history with the air conditioner (frequency index from 0 (least) to 100 (most)), etc.

[0129] Among them, the training labeled scenario data is a data sample (scenario data) containing a predetermined output or result. For example, if the model needs to be trained to automatically adjust the vehicle air-conditioning temperature, the labeled data is the actual air-conditioning system temperature setting in each vehicle usage data.

[0130] Step 302 : performing feature extraction processing on the training vehicle environment information and the training feature information to obtain training standard feature data, and dividing the training standard feature data into a training feature data set and a test feature data set by a preset division method.

[0131] In the embodiment of the present disclosure, the training standard feature data can refer to the description of the standard feature data in the above step 101, so it will not be described in detail here.

[0132] Since the training standard feature data contains multiple data and the scenario recognition model needs to be optimized and evaluated, the training standard feature data needs to be divided into a training feature data set and a test feature data set. The preset division method is a custom division method, for example: using 70% of the training standard feature data as the training feature data set, and 30% of the training standard feature data as the test feature data set, etc. Specifically, the embodiments of the present disclosure do not limit the preset division method.

[0133] In step 303, the training feature data set is input into the scenario recognition model for model training to obtain trained model parameters, and the scenario recognition model is optimized using a preset loss function based on the trained model parameters and the training label scenario data to obtain a trained scenario recognition model.

[0134] In the embodiment of the present disclosure, when performing model training, data training is required through the kernel function in the model to obtain trained model parameters. Specifically, the process of model training includes but is not limited to: calculating hyperparameters through feature data in the training feature data set; based on the hyperparameters, using Gaussian radial basis function (RBF) as the kernel function to perform model training to obtain model parameters; verifying the model parameters to obtain a trained scenario recognition model.

[0135] The preset loss function is a custom-selected loss function, such as Mean Squared Error (MSE), Mean Absolute Error (MAE), Cross-Entropy Loss (Cross-EntropyLoss), Softmax Loss (Softmax Loss), etc. Specifically, the embodiments of the present disclosure do not limit the preset loss function.

[0136] Step 304: Perform model evaluation processing on the trained scenario recognition model using the test feature data set to obtain model evaluation parameters, and continue model training on the scenario recognition model when it is determined that the model evaluation parameters are less than or equal to a preset evaluation threshold.

[0137] In the embodiment of the present disclosure, when performing model evaluation processing on the trained model parameters, it can also be implemented in but not limited to the following manner: configuring the trained model parameters in the scenario recognition model to obtain the trained scenario recognition model, inputting the test feature data set into the trained scenario recognition model to perform scenario recognition, and obtaining the accuracy, precision and recall rate of scenario recognition. When any one of the accuracy, precision and recall rate is less than or equal to a preset evaluation threshold, it is determined that the trained model parameters need further training processing.

[0138] Step 305: When it is determined that the model evaluation parameter is greater than the preset evaluation threshold, the pre-trained scenario recognition model is obtained.

[0139] In an embodiment of the present disclosure, when the accuracy, precision, and recall are all greater than a preset evaluation threshold, the trained scenario recognition model is determined to be the pre-trained scenario recognition model.

[0140] In one possible implementation of the embodiment of the present disclosure, as a refinement of the above step 303, regarding the training process of the model, the embodiment of the present disclosure provides a flow chart of the training process, such as Figure 4 As shown, including:

[0141] Step 401: Obtain the number of features in the training feature dataset and calculate the feature variance of each feature.

[0142] In the embodiment of the present disclosure, the training feature data set includes multiple feature data. The process of calculating the feature variance includes but is not limited to: centering (Mean Normalization), squaring, summing and dividing by the number of samples, etc. Specifically, for the calculation method of the feature variance, please refer to the description in the relevant technology, so it will not be repeated here.

[0143] The feature variance indicates the fluctuation of each feature. The larger the variance, the more dispersed the feature distribution is and the greater the difference between samples. The smaller the variance, the more concentrated the feature distribution is and the smaller the difference between samples.

[0144] Step 402: Parameter calculation is performed based on the number of features and the feature variance to obtain model parameters.

[0145] In the embodiments of the present disclosure, the model parameters include but are not limited to: hyperparameters of the kernel function, etc. The calculation of the hyperparameters includes but is not limited to:

[0146] [\gamma = \frac{1}{n_{\text{features}}\cdot X.\text{var}()}], where (n_{\text{features}}) is the number of features in the dataset, (X.\text{var}()) is the variance of all sample features, and (\gamma) is a kernel hyperparameter that controls the width of the kernel. Another important hyperparameter is (C). It is a regularization parameter that controls the error term in the loss function, balancing the strictness of the model's fit to the training data while ensuring the model's generalization ability.

[0147] The kernel function includes but is not limited to a Gaussian radial basis function (RBF), which is expressed as:

[0148] [K(x,x')=\exp(-\gamma|x-x'|^2)], where (x) and (x') are two sample points.

[0149] Step 403: Verify the model parameters using a preset verification algorithm to obtain the trained model parameters.

[0150] In the embodiment of the present disclosure, the preset verification algorithm is a custom-set algorithm, such as cross-validation, grid search, etc. Specifically, the embodiment of the present disclosure does not limit the preset verification algorithm.

[0151] Regarding cross-validation: Cross-validation can estimate the performance of the model under different hyperparameter settings. If 5-fold cross-validation is used, it means that the training feature dataset will be divided into 5 parts, and then one part will be set aside as the test set while the remaining 4 parts are used for training. This process will be repeated 5 times, and a different part will be selected as the test set each time. Specifically, the code implementation of cross-validation includes but is not limited to:

[0152] Cross validation can be achieved in Python using cross_val_score in the sklearn.model_selection module:

[0153] from sklearn.model_selection import cross_val_score

[0154] from sklearn.svm import SVC

[0155] #Assume that X is the feature data and y is the corresponding label.

[0156] svm_model=SVC(kernel='rbf', C=1.0, gamma='scale')

[0157] scores=cross_val_score(svm_model,X,y,cv=5)

[0158] About grid search: It is a method that automatically traverses multiple combinations of hyperparameter values ​​and evaluates the performance of each combination. In sklearn, grid search is provided by the GridSearchCV class. At the same time, grid search includes cross-validation, so it can complete the tasks of parameter selection and model validation in one go. The code for implementing grid search includes but is not limited to:

[0159] from sklearn.model_selection import GridSearchCV

[0160] #Define parameter grid

[0161] param_grid = {

[0162] 'C':[0.1,1,10,100],

[0163] 'gamma':['scale',0.001,0.01,0.1]

[0164] }

[0165] #Create SVM model

[0166] svm_model=SVC(kernel='rbf')

[0167] #Instantiate the grid search object

[0168] grid_search=GridSearchCV(svm_model,param_grid,cv=5,n_jobs=-1)

[0169] # Perform grid search

[0170] grid_search.fit(X,y)

[0171] #Extract the best parameters and the best model

[0172] best_params=grid_search.best_params_

[0173] best_estimator=grid_search.best_estimator_

[0174] In one possible implementation of the embodiment of the present disclosure, as a refinement of the above-mentioned step 102, regarding determining the target scenario data corresponding to the standard feature data based on the similarity, it can also be implemented in but not limited to the following manner: calculating the probability value of the standard feature data belonging to each of the preset scenario data through the similarity and the preset weights in the pre-trained scenario recognition model; in each of the preset scenario data, selecting the preset scenario data with a higher probability value of containing the standard feature data as the target scenario data.

[0175] In an embodiment of the present disclosure, the preset weight is a weight learned during the training process of the scenario recognition model, wherein the target scenario data is determined based on the similarity and the probability value calculated by the preset weight, thereby improving the inference accuracy of the pre-trained scenario recognition model.

[0176] In one implementable method of the embodiment of the present disclosure, if vehicle control fails, more user feature data will continue to be collected, and more demand materials will be collected to prepare for more accurate identification of scenario patterns. At the same time, continuous training of the scenario pattern large model will also improve the accuracy and efficiency of scenario pattern recognition, and continuously optimize the user experience. Therefore, in order to determine whether vehicle control is successful, the following method may also be used but is not limited to: obtaining the execution information of the control instruction; when it is determined that the execution of the control instruction fails according to the execution information, re-execute the control instruction and obtain the number of failures of the control instruction; when the number of failures exceeds a preset number threshold, obtain the vehicle environment information and the feature information; when it is determined that the execution of the control instruction is successful according to the execution information, continue to obtain the vehicle environment information and the feature information.

[0177] In one possible implementation of the embodiment of the present disclosure, since the vehicle serves people, the vehicle user's control instructions for the vehicle take precedence over the automatic adjustment control instructions. Therefore, in order to ensure the priority of human control of the vehicle, the following method can also be adopted but is not limited to: receiving vehicle control instructions and controlling the target vehicle according to the vehicle control instructions.

[0178] At the same time, after receiving the vehicle control command, the system will change the scene mode to manual operation mode or the driver actively selects the scene mode, and adjust the air conditioning system parameters to match the current mode. At the same time, the system may obtain the user's confirmation or adjustment request for the current scene mode through the user interface.

[0179] In summary of all the above embodiments, in order to facilitate understanding of the implementation process of the vehicle control method described in the present disclosure, the present disclosure provides a schematic diagram of the principle of the vehicle control method, as shown in FIG. Figure 5 As shown, the scenario mode environment control system based on the large model is the vehicle control system, which can be continuously trained to improve the accuracy of environment control in different scenario modes and improve user experience. Figure 5 , the embodiment of the present disclosure provides a scenario description:

[0180] Examples of intelligent reasoning scenarios: The following are specific examples of using this model for intelligent reasoning:

[0181] (1) Scenario: On a weekend morning, a family plans to drive to the suburbs for a day trip.

[0182] (2) Input data:

[0183] Vehicle speed: 80 km / h (normalized to (\frac{80-0}{200-0}=0.4))

[0184] Outside temperature: 5°C (normalized to (\frac{5-(-20)}{40-(-20)}=0.357))

[0185] In-car temperature: 20°C (a fixed reference value that can be used to adjust the air conditioning output temperature)

[0186] Time: 8:00am (converted to time of day: 0:25)

[0187] Location: 30 km from the city center (converted to a binary feature, urban area = 1, suburban area = 0)

[0188] Number of passengers: 4 (raw integer value)

[0189] Passenger-air conditioner interaction history: frequency index 45 (normalized by Z-score: (\frac{45 - \text{mean}}{\text{standard deviation}}), assuming the resulting value is 0.5)

[0190] (3) Preprocessed input vector: [X_{\text{new}}=[0.4,0.357,0.25,0,4,0.5]]

[0191] (4) Model parameters:

[0192] Kernel function: RBF

[0193] (C): 1.0 (assuming optimality)

[0194] (\gamma):'scale' (as set above)

[0195] (5) Model reasoning:

[0196] The input vector (X_{\text{new}}) is fed into the trained SVM model.

[0197] The SVM model uses the RBF kernel function to calculate the similarity between (X_{\text{new}}) and the support vector.

[0198] The model makes decisions based on these similarities and the learned weights.

[0199] (6) Output: Imagine that the model outputs the probabilities corresponding to the three labels "low", "medium", and "high". In this scenario, the model gives the following probabilities (hypothetical values):

[0200] Low: 0.1

[0201] Medium: 0.7

[0202] High: 0.2

[0203] Since "medium" has the highest probability, the model recommends a medium air conditioning temperature setting.

[0204] (7) Intelligent reasoning logic: Based on the output results and real-time data analysis, the model reasoning is as follows:

[0205] Since it was morning, the speed indicator indicated that the vehicle was probably on its way to its destination, in long-distance travel mode.

[0206] The slightly cooler outside temperature and the number of passengers being four indicate that the interior of the vehicle may require slightly warmer conditions for comfort.

[0207] Combined with the personalized user-air conditioning interaction history, it shows that passengers do not adjust the air conditioning frequently.

[0208] (8) Action: The system will intelligently adjust the air conditioning setting to a suitable medium temperature to ensure the comfort of all passengers in the car and optimize energy efficiency.

[0209] This process demonstrates how to combine specific scenarios, real-time input data, and pre-trained models to perform intelligent reasoning and automatically adjust the in-car environment. Over time, as more data is collected, the model can be iteratively optimized, making intelligent reasoning even more accurate and personalized.

[0210] It should also be noted that the pre-trained scenario recognition model can collaborate with the cloud platform to continuously obtain new data and algorithm updates from the cloud to enrich the judgment of scenario patterns and optimize the algorithm execution efficiency.

[0211] In summary, the embodiments of the present disclosure can achieve the following effects:

[0212] 1. The embodiment of the present disclosure can automatically identify the vehicle driving scene through a pre-trained scenario recognition model, and generate control instructions to control the target vehicle based on the vehicle control parameters in the driving scene, and can automatically adjust various parameters of the vehicle according to different driving environments.

[0213] 2. The disclosed embodiment can perform precise temperature control through a pre-trained scenario recognition model. By using a large model to predict external environmental changes and internal cabin status, the intelligent air-conditioning system can accurately adjust the temperature and provide a more comfortable riding environment.

[0214] 3. The embodiment of the present disclosure can reduce energy consumption through a pre-trained scenario recognition model. By predicting and automatically identifying the optimal cooling and heating loads, the air-conditioning system can use energy more efficiently and reduce energy consumption.

[0215] 4. The embodiment of the present disclosure uses a pre-trained scenario recognition model to enable users to adjust the air-conditioning system without having to do so manually. The system can automatically adjust the settings based on changes in the external environment and the real-time needs of passengers in the car, thereby improving convenience.

[0216] 5. The embodiment of the present disclosure can learn and automatically adjust to the most suitable air-conditioning mode according to the preferences of different drivers or passengers through a pre-trained scenario recognition model.

[0217] 6. The disclosed embodiment can maintain a suitable temperature inside the vehicle through a pre-trained scenario recognition model, helping to keep the driver awake and focused, thereby potentially reducing the risk of accidents caused by fatigue or discomfort.

[0218] 7. The embodiment of the present disclosure can achieve the desired temperature and comfort level more quickly through real-time recognition and rapid adaptation to different scenario modes.

[0219] Corresponding to the above-mentioned vehicle control method, the present invention also provides a vehicle control device. Since the device embodiment of the present invention corresponds to the above-mentioned method embodiment, details not disclosed in the device embodiment can be referred to the above-mentioned method embodiment and will not be repeated in this invention.

[0220] Figure 6 A schematic diagram of the structure of a vehicle control device provided by an embodiment of the present disclosure is shown in FIG. Figure 6 As shown, including:

[0221] A first acquisition unit 601 is configured to acquire vehicle environment information of a target vehicle and feature information of a monitored object within the target vehicle;

[0222] A first extraction unit 602 is configured to perform feature extraction processing on the vehicle environment information and the feature information to obtain standard feature data;

[0223] A calculation unit 603 is used to input the standard feature data into a pre-trained scenario recognition model and calculate the similarity between the standard feature data and each preset scenario data;

[0224] A first determining unit 604 is configured to determine target scenario data corresponding to the standard feature data based on the similarity; wherein the preset scenario data is pre-configured driving scenario data including vehicle control parameters, and different preset scenario data include different vehicle control parameters;

[0225] A generating unit 605 is configured to generate corresponding control instructions according to the vehicle control parameters included in the target scenario data;

[0226] The control unit 606 is configured to control the target vehicle according to the control instruction.

[0227] The vehicle control device provided by the present disclosure obtains vehicle environment information of a target vehicle and feature information of a monitored object in the target vehicle, and performs feature extraction processing on the vehicle environment information and the feature information to obtain standard feature data; the standard feature data is input into a pre-trained scenario recognition model, and the similarity between the standard feature data and each preset scenario data is calculated respectively, and the target scenario data corresponding to the standard feature data is determined based on the similarity; wherein the preset scenario data is pre-configured driving scenario data containing vehicle control parameters, and different preset scenario data contain different vehicle control parameters; corresponding control instructions are generated based on the vehicle control parameters contained in the target scenario data, and the target vehicle is controlled based on the control instructions. Compared with the related art, the embodiment of the present disclosure can automatically identify the driving scenario of the vehicle through the pre-trained scenario recognition model, and generate control instructions to control the target vehicle based on the vehicle control parameters in the driving scenario, and can automatically adjust various parameters of the vehicle according to different driving environments.

[0228] Furthermore, in a possible implementation of the embodiment of the present disclosure, as Figure 7 As shown, the first extraction unit 602 includes:

[0229] The encoding module 6021 is used to encode the vehicle environment information and the characteristic information respectively using a preset encoding method to obtain encoded information data;

[0230] The processing module 6022 is configured to perform standardization processing on the coded information data using a preset standardized format to obtain the standard information data;

[0231] The extraction module 6023 is used to perform feature extraction processing on the standard information data using a preset feature selection algorithm to obtain the standard feature data.

[0232] Furthermore, in a possible implementation of the embodiment of the present disclosure, the first acquiring unit 601 is further configured to acquire the status information of the monitored object and the vehicle control record of the monitored object controlling the target vehicle;

[0233] The first extraction unit 602 is further configured to perform data extraction processing based on the state information and the vehicle control record to obtain update scenario data;

[0234] like Figure 7 As shown, the device also includes:

[0235] An optimization unit 607 is configured to optimize the model parameters of the pre-trained scenario recognition model using a preset optimization algorithm according to the target scenario data and the update scenario data to obtain updated model parameters;

[0236] The configuration unit 608 is configured to configure the updated model parameters in the pre-trained scenario recognition model.

[0237] Furthermore, in a possible implementation of the embodiment of the present disclosure, as Figure 7 As shown, the device also includes:

[0238] A second acquisition unit 609 is configured to acquire a preset amount of vehicle usage data, wherein the vehicle usage data includes training vehicle environment information, training feature information, and training label scenario data;

[0239] A second extraction unit 610 is configured to perform feature extraction processing on the training vehicle environment information and the training feature information to obtain standard feature data for training;

[0240] A division unit 611 is configured to divide the training standard feature data into a training feature data set and a testing feature data set by a preset division method;

[0241] A training unit 612 is configured to input the training feature dataset into a scenario recognition model for model training to obtain trained model parameters, and optimize the scenario recognition model using a preset loss function based on the trained model parameters and the training labeled scenario data to obtain a trained scenario recognition model;

[0242] An evaluation unit 613 is configured to perform a model evaluation process on the trained scene recognition model using the test feature dataset to obtain model evaluation parameters;

[0243] The training unit 612 is further configured to, when it is determined that the model evaluation parameter is less than or equal to a preset evaluation threshold, continue to perform model training on the scenario recognition model;

[0244] The second determining unit 614 is configured to obtain the pre-trained scenario recognition model when it is determined that the model evaluation parameter is greater than the preset evaluation threshold.

[0245] Furthermore, in a possible implementation of the embodiment of the present disclosure, as Figure 7 As shown, the training unit 612 includes:

[0246] An acquisition module 6121 is used to obtain the number of features in the training feature dataset;

[0247] A calculation module 6122 is used to calculate the feature variance of each feature;

[0248] The calculation module 6122 is further configured to perform parameter calculation processing based on the number of features and the feature variance to obtain model parameters;

[0249] The verification module 6123 is used to verify the model parameters using a preset verification algorithm to obtain the trained model parameters.

[0250] Furthermore, in a possible implementation of the embodiment of the present disclosure, the first determining unit 604 is further configured to:

[0251] Calculating a probability value of the standard feature data belonging to each of the preset scenario data by using the similarity and the preset weights in the pre-trained scenario recognition model;

[0252] In each of the preset scenario data, the preset scenario data having a higher probability value of containing the standard feature data is selected as the target scenario data.

[0253] Furthermore, in a possible implementation of the embodiment of the present disclosure, the first acquiring unit 601 is further configured to acquire execution information of the control instruction;

[0254] The control unit 606 is further configured to, when it is determined according to the execution information that the execution of the control instruction fails, re-execute the control instruction and obtain the number of failures of the execution of the control instruction;

[0255] The first acquisition unit 601 is further configured to acquire the vehicle environment information and the feature information when the number of failures exceeds a preset number threshold;

[0256] The first acquiring unit 601 is further configured to, when it is determined according to the execution information that the control instruction is successfully executed, continue to acquire the vehicle environment information and the feature information.

[0257] It should be noted that the above explanation of the method embodiment is also applicable to the device of the embodiment of the present disclosure, and the principles are the same, which is no longer limited in the embodiment of the present disclosure.

[0258] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0259] Figure 8A schematic block diagram of an example electronic device 800 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are provided as examples only and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0260] like Figure 8 As shown, the device 800 includes a computing unit 801, which can perform various appropriate actions and processes according to a computer program stored in a ROM (Read-Only Memory) 802 or a computer program loaded from a storage unit 808 into a RAM (Random Access Memory) 803. Various programs and data required for the operation of the device 800 can also be stored in the RAM 803. The computing unit 801, the ROM 802, and the RAM 803 are connected to each other via a bus 804. An I / O (Input / Output) interface 805 is also connected to the bus 804.

[0261] Various components in device 800 are connected to I / O interface 805, including an input unit 806, such as a keyboard, mouse, etc.; an output unit 807, such as various types of displays, speakers, etc.; a storage unit 808, such as a magnetic disk, optical disk, etc.; and a communication unit 809, such as a network card, modem, wireless communication transceiver, etc. The communication unit 809 allows device 800 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0262] The computing unit 801 can be any general-purpose and / or specialized processing component with processing and computing capabilities. Some examples of the computing unit 801 include, but are not limited to, a CPU (Central Processing Unit), a GPU (Graphic Processing Unit), various specialized AI (Artificial Intelligence) computing chips, various computing units that run machine learning model algorithms, a DSP (Digital Signal Processor), and any suitable processor, controller, microcontroller, etc. The computing unit 801 performs the various methods and processes described above, such as the vehicle control method. For example, in some embodiments, the vehicle control method can be implemented as a computer software program tangibly embodied in a machine-readable medium, such as the storage unit 808. In some embodiments, part or all of the computer program can be loaded and / or installed onto the device 800 via the ROM 802 and / or the communication unit 809. When the computer program is loaded into the RAM 803 and executed by the computing unit 801, one or more steps of the method described above can be performed. Alternatively, in other embodiments, the computing unit 801 may be configured to execute the aforementioned vehicle control method in any other appropriate manner (for example, by means of firmware).

[0263] Various embodiments of the systems and techniques described herein can be implemented in digital electronic circuit systems, integrated circuit systems, FPGAs (Field Programmable Gate Arrays), ASICs (Application-Specific Integrated Circuits), ASSPs (Application Specific Standard Products), SOCs (System on Chips), CPLDs (Complex Programmable Logic Devices), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system that includes at least one programmable processor, which can be a special-purpose or general-purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0264] The program code for implementing the method of the present disclosure can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device so that when the program code is executed by the processor or controller, the functions / operations specified in the flow chart and / or block diagram are implemented. The program code can be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0265] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or apparatus. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or apparatus, or any suitable combination of the foregoing. More specific examples of machine-readable storage media may include an electrical connection based on one or more wires, a portable computer disk, a hard disk, RAM, ROM, EPROM (Electrically Programmable Read-Only-Memory) or flash memory, optical fiber, CD-ROM (Compact Disc Read-Only Memory), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0266] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (Cathode-Ray Tube) or LCD (Liquid Crystal Display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0267] The systems and techniques described herein can be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: LAN (Local Area Network), WAN (Wide Area Network), the Internet, and blockchain networks.

[0268] A computer system may include a client and a server. The client and server are generally remote from each other and typically interact via a communication network. This client-server relationship is established by computer programs running on the respective computers, establishing a client-server relationship. The server may be a cloud server, also known as a cloud computing server or cloud host, a host product within the cloud computing service ecosystem that addresses the management difficulties and limited scalability of traditional physical hosts and VPS services ("Virtual Private Servers" or simply "VPS"). The server may also be a server in a distributed system or a server integrated with blockchain.

[0269] It's important to note that artificial intelligence (AI) is the study of how computers can simulate certain human thought processes and intelligent behaviors (such as learning, reasoning, thinking, and planning). This encompasses both hardware and software technologies. AI hardware technologies generally include sensors, specialized AI chips, cloud computing, distributed storage, and big data processing. AI software technologies primarily encompass computer vision, speech recognition, natural language processing, machine learning / deep learning, big data processing, and knowledge graphs.

[0270] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in this disclosure can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved. This is not limited herein.

[0271] The above specific embodiments do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure shall be included within the scope of protection of this disclosure.

Claims

1. A vehicle control method, characterized in that: include: Acquiring vehicle environment information of a target vehicle and feature information of a monitored object in the target vehicle, and performing feature extraction processing on the vehicle environment information and the feature information to obtain standard feature data; Inputting the standard feature data into a pre-trained scenario recognition model, calculating the similarity between the standard feature data and each preset scenario data, and determining the target scenario data corresponding to the standard feature data based on the similarity; wherein the preset scenario data is pre-configured driving scenario data including vehicle control parameters, and different preset scenario data include different vehicle control parameters; A corresponding control instruction is generated according to the vehicle control parameters included in the target scenario data, and the target vehicle is controlled according to the control instruction.

2. The method according to claim 1, characterized in that The performing feature extraction processing on the vehicle environment information and the feature information to obtain standard feature data includes: Encoding the vehicle environment information and the characteristic information respectively using a preset encoding method to obtain encoded information data; Standardizing the coded information data using a preset standardized format to obtain the standard information data; The standard information data is subjected to feature extraction processing by a preset feature selection algorithm to obtain the standard feature data.

3. The method according to claim 1, characterized in that After controlling the target vehicle according to the control instruction, the method further includes: Acquiring status information of the monitored object and a vehicle control record of the monitored object controlling the target vehicle; Performing data extraction processing based on the state information and the vehicle control record to obtain update scenario data; Optimizing the model parameters in the pre-trained scenario recognition model using a preset optimization algorithm according to the target scenario data and the update scenario data to obtain updated model parameters; The updated model parameters are configured in the pre-trained scenario recognition model.

4. The method according to claim 1, wherein Before obtaining vehicle environment information of a target vehicle and characteristic information of a monitored object in the target vehicle, the method further includes: Acquire a preset amount of vehicle usage data; wherein the vehicle usage data includes training vehicle environment information, training feature information, and training label scenario data; Performing feature extraction processing on the training vehicle environment information and the training feature information to obtain training standard feature data, and dividing the training standard feature data into a training feature data set and a test feature data set by a preset division method; Inputting the training feature data set into the scenario recognition model for model training to obtain trained model parameters, and optimizing the scenario recognition model using a preset loss function based on the trained model parameters and the training label scenario data to obtain a trained scenario recognition model; performing model evaluation processing on the trained scenario recognition model using the test feature data set to obtain a model evaluation parameter, and continuing model training on the scenario recognition model if it is determined that the model evaluation parameter is less than or equal to a preset evaluation threshold; When it is determined that the model evaluation parameter is greater than the preset evaluation threshold, the pre-trained scenario recognition model is obtained.

5. The method according to claim 4, characterized in that The inputting of the training feature data set into the scene recognition model for model training to obtain trained model parameters includes: Obtaining the number of features in the training feature dataset and calculating the feature variance of each feature; Perform parameter calculation processing based on the number of features and the feature variance to obtain model parameters; The model parameters are verified using a preset verification algorithm to obtain the trained model parameters.

6. The method according to claim 1, characterized in that Determining the target scenario data corresponding to the standard feature data according to the similarity includes: Calculating a probability value of the standard feature data belonging to each of the preset scenario data by using the similarity and the preset weights in the pre-trained scenario recognition model; In each of the preset scenario data, the preset scenario data having a higher probability value of containing the standard feature data is selected as the target scenario data.

7. The method according to claim 1, characterized in that After controlling the target vehicle according to the control instruction, the method further includes: Obtaining execution information of the control instruction; When it is determined according to the execution information that the control instruction fails to be executed, re-execute the control instruction and obtain the number of failures of the control instruction execution; When the number of failures exceeds a preset number threshold, acquiring the vehicle environment information and the characteristic information; When it is determined according to the execution information that the control instruction is successfully executed, the vehicle environment information and the feature information are continued to be acquired.

8. A vehicle control device, characterized in that: include: A first acquisition unit is used to acquire vehicle environment information of a target vehicle and feature information of a monitored object in the target vehicle; a second extraction unit, configured to perform feature extraction processing on the vehicle environment information and the feature information to obtain standard feature data; a calculation unit, configured to input the standard feature data into a pre-trained scenario recognition model and respectively calculate the similarity between the standard feature data and each preset scenario data; A first determining unit is configured to determine target scenario data corresponding to the standard feature data based on the similarity; wherein the preset scenario data is pre-configured driving scenario data including vehicle control parameters, and different preset scenario data include different vehicle control parameters; a generating unit, configured to generate corresponding control instructions according to the vehicle control parameters included in the target scenario data; A control unit is used to control the target vehicle according to the control instruction.

9. A vehicle, characterized in that: The vehicle includes the vehicle control device according to claim 8.

10. An electronic device, characterized in that: include: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 7.

Images