Information providing method and device, computer equipment and storage medium

Abstract

The invention relates to an information providing method and device, computer equipment and a storage medium. The method comprises the steps that information used for prediction and corresponding to a target moment is acquired, the information used for prediction comprises a plurality of first vehicle parameters of a target vehicle at the target moment and a road surface image related to the target moment, and the road surface image comprises a road surface object representing a target road surface; using a target neural network to predict a prediction result related to a target road surface according to the information for prediction, the prediction result including a road surface adhesion coefficient of the target road surface, a bumping degree of the target road surface, and a road surface gradient of the target road surface; and providing each item in the prediction result to a corresponding system of the target vehicle, so that the corresponding system performs corresponding adjustment according to the corresponding item in the prediction result.

Description

Technical Field

[0001] The present invention relates to the field of vehicles, and in particular to an information providing method, device, computer equipment and storage medium. Background Art

[0002] Predicting road-related information and providing the vehicle's system with road-related information so that the vehicle's corresponding system can be adjusted according to the road-related information is an important function of the assisted driving system.

[0003] In the related art, the road surface related information provided to the vehicle system is the road surface type. However, the road condition of a road surface type is divided into multiple road conditions. For example, the road condition of a snow road surface type is divided into deep snow road conditions, shallow snow road conditions, dry snow road conditions and other road conditions. The road surface type of the road surface on which the vehicle is traveling cannot reflect the road condition of the road surface on which the vehicle is traveling, and the accuracy of the road surface related information provided to the vehicle's corresponding system is low. The low accuracy of the related information of the road surface on which the vehicle is traveling provided to the vehicle's corresponding system will make it difficult for the vehicle's corresponding system to make more accurate corresponding adjustments. How to improve the accuracy of the related information of the road surface on which the vehicle is traveling provided to the vehicle's system has become a problem that needs to be solved. Summary of the invention

[0004] One of the purposes of the present invention is to provide an information providing method, apparatus, computer device and storage medium to solve the problem of how to improve the accuracy of relevant information provided to a vehicle system.

[0005] In order to achieve the above object, the technical solution adopted by the present invention is as follows:

[0006] A method for providing information, the method comprising:

[0007] Acquiring information for prediction corresponding to a target time, the information for prediction comprising: a plurality of first vehicle parameters of the target vehicle at the target time, and a road surface image related to the target time, the road surface image comprising a road surface object representing the target road surface;

[0008] Using a target neural network, based on the information for prediction, predicting a prediction result related to the target road surface, the prediction result including: a road surface adhesion coefficient of the target road surface, a bumpiness of the target road surface, and a road surface slope of the target road surface;

[0009] Each item in the prediction results is provided to a corresponding system of the target vehicle, so that the corresponding system can make corresponding adjustments according to the corresponding item in the prediction results.

[0010] Further, providing each item in the prediction result to a corresponding system of the target vehicle respectively, so that the corresponding system performs corresponding adjustments according to the corresponding item in the prediction result, includes:

[0011] providing the road surface adhesion coefficient of the target road surface to the traction control system of the target vehicle so that the traction control system of the target vehicle adjusts the output power of the engine of the target vehicle according to the road surface adhesion coefficient of the target road surface;

[0012] Providing the bumpiness of the target road surface to a suspension system of a target vehicle so that the suspension system of the target vehicle adjusts a damping rate according to the bumpiness of the target road surface;

[0013] The road surface gradient of the target road surface is provided to a torque distribution system of the target vehicle, so that the torque distribution system of the target vehicle adjusts the engine torque of the target vehicle according to the road surface gradient of the target road surface.

[0014] Further, using the target neural network, based on the information for prediction, predicting a prediction result related to the target road surface includes:

[0015] Generate a hidden state vector for each of the multiple first vehicle parameters, and extract a road surface image feature vector of the road surface image; generate a target hidden state vector based on the hidden state vector of each of the multiple first vehicle parameters and the hidden state vector of each of the multiple second vehicle parameters of the target vehicle at a previous moment before the target moment; generate features for prediction based on the target hidden state vector and the road surface image feature vector, and predict the prediction result based on the features for prediction.

[0016] Further, generating a target hidden state vector according to the hidden state vector of each first vehicle parameter in the plurality of first vehicle parameters and the hidden state vector of each second vehicle parameter in the plurality of second vehicle parameters of the target vehicle at a previous moment before the target moment includes:

[0017] Determining an attention parameter of each first vehicle parameter according to a hidden state vector of each first vehicle parameter among the plurality of first vehicle parameters;

[0018] Generate an attention hidden state vector of each first vehicle parameter according to the hidden state vector of each first vehicle parameter in the plurality of first vehicle parameters and the attention parameter of each first vehicle parameter;

[0019] A target hidden state vector is generated based on the attention hidden state vector of each first vehicle parameter and the hidden state vector of each second vehicle parameter.

[0020] Further, according to the hidden state vector of each first vehicle parameter among the plurality of first vehicle parameters, determining the attention parameter of each first vehicle parameter comprises:

[0021] Determining a first attention score for each of the plurality of first vehicle parameters according to a hidden state vector of the first vehicle parameter;

[0022] For each of the first vehicle parameters, determining a second attention score of the first vehicle parameter according to a hidden state vector of the first vehicle parameter and a hidden state vector of a second vehicle parameter corresponding to the first vehicle parameter, wherein a type of the first vehicle parameter is the same as a type of the second vehicle parameter corresponding to the first vehicle parameter;

[0023] An attention parameter of each first vehicle parameter is determined according to the first attention score of each first vehicle parameter and the second attention score of each first vehicle parameter.

[0024] Further, generating a target hidden state vector according to the attention hidden state vector of each first vehicle parameter and the hidden state vector of each second vehicle parameter includes:

[0025] Generate an attention hidden state vector of each second vehicle parameter according to the hidden state vector of each second vehicle parameter and the attention parameter of each second vehicle parameter;

[0026] A target hidden state vector is generated according to the attention hidden state vector of each first vehicle parameter and the attention hidden state vector of each second vehicle parameter.

[0027] Further, generating a target hidden state vector according to the attention hidden state vector of each first vehicle parameter and the attention hidden state vector of each second vehicle parameter includes:

[0028] generating a first vehicle parameter combination vector including an attention hidden state vector of each of the first vehicle parameters, and generating a second vehicle parameter combination vector including an attention hidden state vector of each of the second vehicle parameters;

[0029] The first vehicle parameter combination vector is added to the second vehicle parameter combination vector to obtain a target hidden state vector.

[0030] Furthermore, it is characterized in that: generating features for prediction according to the target hidden state vector and the road surface image feature vector includes:

[0031] Multiplying the target hidden state vector by the vehicle parameter feature weight to obtain a first weighted vector, and multiplying the road surface image feature vector by the road surface image feature weight to obtain a second weighted vector;

[0032] A feature for prediction is generated according to the first weighting vector and the second weighting vector.

[0033] Further, generating features for prediction according to the first weighted vector and the second weighted vector includes:

[0034] The first weighted vector and the second weighted vector are added together to obtain features for prediction.

[0035] Further, before obtaining the information for prediction corresponding to the target time, the method also includes:

[0036] The vehicle parameter feature weights and the road surface image feature weights are determined according to the accuracy of the vehicle parameter feature extraction network in the target neural network on the verification set of the target neural network and the accuracy of the road surface image feature extraction network in the target neural network on the verification set.

[0037] According to the above technical means, it is considered that the road conditions of the road surface of the same road surface type are divided into multiple types, and the different road conditions of the road surface of the same road surface type have different road surface adhesion coefficients, bumpiness, and road surface slopes. For example, the road conditions of the road surface type of snow road surface are divided into deep snow road conditions, shallow snow road conditions, dry snow road conditions, etc. Deep snow road conditions, shallow snow road conditions, and dry snow road conditions have different road surface adhesion coefficients, and deep snow road conditions, shallow snow road conditions, and dry snow road conditions have different bumpiness. The road surface adhesion coefficient of the target road surface, the bumpiness of the target road surface, and the road surface slope of the target road surface predicted by the present invention are related to the road conditions of the target road surface. Among them, the road surface adhesion coefficient of the target road surface can describe the friction between the tire and the target road surface, the bumpiness of the target road surface can reflect the road surface flatness or vibration degree of the target road surface, and the road surface slope of the target road surface can reflect the inclination angle of the target road surface. The accuracy of the relevant information provided to the system of the vehicle is high. The road adhesion coefficient of the target road surface, the bumpiness of the target road surface, and the road surface slope of the target road surface related to the road condition of the target road surface are respectively provided to the corresponding system of the target vehicle, so that the corresponding system of the target vehicle can make adjustments suitable for the road condition of the target road surface according to the information related to the road condition of the target road surface, and the adjustment accuracy is high. The road adhesion coefficient of the target road surface, the bumpiness of the target road surface, and the road surface slope of the target road surface can be provided to the chassis system of the target vehicle, so that the chassis system of the target vehicle can make adjustments suitable for the road condition of the target road surface according to the road adhesion coefficient of the target road surface, the bumpiness of the target road surface, and the road surface slope of the target road surface, thereby improving the dynamic stability, safety, and comfort of the target vehicle.

[0038] Different road conditions of the same road type have different road adhesion coefficients, bumpiness, and road slopes. The corresponding system is adjusted according to at least one of the road adhesion coefficient, bumpiness, and road slope. Therefore, the information providing method provided in the embodiment of the present invention adopts different adjustment methods for the corresponding system for multiple different road conditions of the same road type. Compared with the related art in which the corresponding systems for multiple different road conditions of the same road type are adjusted according to the same road type, the corresponding system in the embodiment of the present invention has a high degree of refinement in adjustment.

[0039] Since various road conditions of various types of roads each have corresponding road adhesion coefficients, bumpiness, and road slopes, at least one of the road adhesion coefficients, bumpiness, and road slopes can be provided to the corresponding system for various types of roads. The information providing method provided by the embodiment of the present invention has strong applicability. It can also be said that the information providing method provided by the embodiment of the present invention covers a wider range of road conditions. The information providing method provided by the embodiment of the present invention can avoid the following situations: inaccurate identification of road type, such as inaccurate identification of road type due to the complexity of the road type of the target road, inaccurate identification of road type due to the target road being a composite road condition of partially waterlogged and partially muddy, and the corresponding system is adjusted only according to the road type, and the adjustment accuracy is low.

[0040] An information providing device, comprising:

[0041] an acquisition unit, configured to acquire information for prediction corresponding to a target time, the information for prediction comprising: a plurality of first vehicle parameters of the target vehicle at the target time, and a road surface image related to the target time, the road surface image comprising a road surface object representing the target road surface;

[0042] A prediction unit, used to predict a prediction result related to the target road surface by using a target neural network according to the information for prediction, wherein the prediction result includes: a road surface adhesion coefficient of the target road surface, a bumpiness of the target road surface, and a road surface slope of the target road surface;

[0043] A providing unit is used to provide each item in the prediction result to a corresponding system of the target vehicle, so that the corresponding system can make corresponding adjustments according to the corresponding item in the prediction result.

[0044] Furthermore, a unit is provided, which is further used to provide each item in the prediction results to the corresponding system of the target vehicle respectively, so that the corresponding system makes corresponding adjustments according to the corresponding items in the prediction results, including: providing the road adhesion coefficient of the target road to the traction control system of the target vehicle so that the traction control system of the target vehicle adjusts the output power of the engine of the target vehicle according to the road adhesion coefficient of the target road; providing the bumpiness of the target road to the suspension system of the target vehicle so that the suspension system of the target vehicle adjusts the damping rate according to the bumpiness of the target road; providing the road slope of the target road to the torque distribution system of the target vehicle so that the torque distribution system of the target vehicle adjusts the engine torque of the target vehicle according to the road slope of the target road.

[0045] Furthermore, the prediction unit is further used to generate a hidden state vector of each first vehicle parameter among the multiple first vehicle parameters, and to extract a road surface image feature vector of the road surface image; to generate a target hidden state vector based on the hidden state vector of each first vehicle parameter among the multiple first vehicle parameters and the hidden state vector of each second vehicle parameter among the multiple second vehicle parameters of the target vehicle at the previous moment before the target moment; to generate features for prediction based on the target hidden state vector and the road surface image feature vector, and to predict the prediction result based on the features for prediction.

[0046] Furthermore, the prediction unit is further used to determine the attention parameter of each first vehicle parameter among the multiple first vehicle parameters based on the hidden state vector of each first vehicle parameter; generate the attention hidden state vector of each first vehicle parameter according to the hidden state vector of each first vehicle parameter among the multiple first vehicle parameters and the attention parameter of each first vehicle parameter; generate the target hidden state vector according to the attention hidden state vector of each first vehicle parameter and the hidden state vector of each second vehicle parameter.

[0047] Furthermore, the prediction unit is further used to determine a first attention score of each first vehicle parameter among the multiple first vehicle parameters based on a hidden state vector of each first vehicle parameter; for each first vehicle parameter, determine a second attention score of the first vehicle parameter based on the hidden state vector of the first vehicle parameter and a hidden state vector of a second vehicle parameter corresponding to the first vehicle parameter, wherein a type of the first vehicle parameter is the same as a type of the second vehicle parameter corresponding to the first vehicle parameter; determine an attention parameter of each first vehicle parameter based on the first attention score of each first vehicle parameter and the second attention score of each first vehicle parameter.

[0048] Furthermore, the prediction unit is further used to generate an attention hidden state vector of each second vehicle parameter based on the hidden state vector of each second vehicle parameter and the attention parameter of each second vehicle parameter; and to generate a target hidden state vector based on the attention hidden state vector of each first vehicle parameter and the attention hidden state vector of each second vehicle parameter.

[0049] Furthermore, the prediction unit is further used to generate a first vehicle parameter combination vector including the attention hidden state vector of each first vehicle parameter, and to generate a second vehicle parameter combination vector including the attention hidden state vector of each second vehicle parameter; and to add the first vehicle parameter combination vector to the second vehicle parameter combination vector to obtain a target hidden state vector.

[0050] Furthermore, the prediction unit is further used to multiply the target hidden state vector by the vehicle parameter feature weight to obtain a first weighted vector, and to multiply the road surface image feature vector by the road surface image feature weight to obtain a second weighted vector; and generate features for prediction based on the first weighted vector and the second weighted vector.

[0051] Furthermore, the prediction unit is further used to add the first weighted vector and the second weighted vector to obtain a feature for prediction.

[0052] Furthermore, the information providing device comprises:

[0053] The weight determination unit is used to determine the vehicle parameter feature weight and the road surface image feature weight according to the accuracy of the vehicle parameter feature extraction network in the target neural network on the verification set of the target neural network and the accuracy of the road surface image feature extraction network in the target neural network on the verification set before obtaining the information for prediction corresponding to the target time.

[0054] A computer device comprising:

[0055] A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the above method by executing the computer instructions.

[0056] A computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the above method.

[0057] A computer program product, characterized in that it includes computer instructions, wherein the computer instructions are used to enable a computer to execute the above method.

[0058] Beneficial effects of the present invention:

[0059] The target neural network is used to directly predict the target road surface adhesion coefficient, target road surface bumpiness, and target road surface slope required for the corresponding system of the target vehicle to make corresponding adjustments based on multiple first vehicle parameters of the target vehicle at the target time and the road surface image related to the target time. The target road surface adhesion coefficient, target road surface bumpiness, and target road surface slope required for the corresponding system of the target vehicle to make corresponding adjustments are directly provided to the corresponding system of the target vehicle, so that the corresponding system of the target vehicle can make corresponding adjustments based on the target road surface adhesion coefficient, target road surface bumpiness, and target road surface slope, so that the driving and riding experience of the personnel in the target vehicle when the target vehicle is traveling on the target road surface is more comfortable.

[0060] The information providing method provided by the embodiment of the present invention takes into account that the road conditions of the road surface of the same road surface type are divided into multiple types, and different road conditions of the road surface of the same road surface type have different road surface adhesion coefficients, bumpiness, and road surface slopes. For example, the road conditions of the road surface type of snow road surface are divided into deep snow road conditions, shallow snow road conditions, dry snow road conditions, etc. Deep snow road conditions, shallow snow road conditions, and dry snow road conditions have different road surface adhesion coefficients, and deep snow road conditions, shallow snow road conditions, and dry snow road conditions have different bumpiness. The road surface adhesion coefficient of the target road surface, the bumpiness of the target road surface, and the road surface slope of the target road surface predicted by the present invention are related to the road conditions of the target road surface. Among them, the road surface adhesion coefficient of the target road surface can describe the friction between the tire and the target road surface, the bumpiness of the target road surface can reflect the road surface flatness or vibration degree of the target road surface, and the road surface slope of the target road surface can reflect the inclination angle of the target road surface. The relevant information provided to the system of the vehicle has a high accuracy. The road adhesion coefficient of the target road surface, the bumpiness of the target road surface, and the road surface slope of the target road surface related to the road condition of the target road surface are respectively provided to the corresponding system of the target vehicle, so that the corresponding system of the target vehicle can make adjustments suitable for the road condition of the target road surface according to the information related to the road condition of the target road surface, and the adjustment accuracy is high. The road adhesion coefficient of the target road surface, the bumpiness of the target road surface, and the road surface slope of the target road surface can be provided to the chassis system of the target vehicle, so that the chassis system of the target vehicle can make adjustments suitable for the road condition of the target road surface according to the road adhesion coefficient of the target road surface, the bumpiness of the target road surface, and the road surface slope of the target road surface, thereby improving the dynamic stability, safety, and comfort of the target vehicle.

[0061] Different road conditions of the same road type have different road adhesion coefficients, bumpiness, and road slopes. The corresponding system is adjusted according to at least one of the road adhesion coefficient, bumpiness, and road slope. Therefore, the information providing method provided in the embodiment of the present invention adopts different adjustment methods for the corresponding system for multiple different road conditions of the same road type. Compared with the related art in which the corresponding systems for multiple different road conditions of the same road type are adjusted according to the same road type, the corresponding system in the embodiment of the present invention has a high degree of refinement in adjustment.

[0062] Since various road conditions of various types of roads each have corresponding road adhesion coefficients, bumpiness, and road slopes, at least one of the road adhesion coefficients, bumpiness, and road slopes can be provided to the corresponding system for various types of roads. The information providing method provided by the embodiment of the present invention has strong applicability. It can also be said that the information providing method provided by the embodiment of the present invention covers a wider range of road conditions. The information providing method provided by the embodiment of the present invention can avoid the following situations: inaccurate identification of road type, such as inaccurate identification of road type due to the complexity of the road type of the target road, inaccurate identification of road type due to the target road being a composite road condition of partially waterlogged and partially muddy, and the corresponding system is adjusted only according to the road type, and the adjustment accuracy is low.

[0063] In an embodiment of the present invention, the road adhesion coefficient of the target road surface, the bumpiness of the target road surface, and the road slope of the target road surface required for the corresponding system of the target vehicle to be adjusted are directly provided to the corresponding system of the target vehicle respectively. Thus, the corresponding system of the target vehicle does not need to convert the information provided to the corresponding system of the target vehicle into the corresponding parameters required for the corresponding adjustment before the corresponding adjustment is performed. The corresponding system of the target vehicle directly performs corresponding adjustments according to the corresponding items of the road adhesion coefficient of the target road surface, the bumpiness of the target road surface, and the road slope of the target road surface. Thus, the efficiency of the corresponding system of the target vehicle in performing corresponding adjustments is improved, and the overall control efficiency of the target vehicle is improved, so that the target vehicle quickly reaches a better state. For example, the bumpiness of the target road surface is directly provided to the suspension system of the target vehicle, and the suspension system of the target vehicle can be directly adaptively adjusted according to the bumpiness of the target road surface to provide a smoother driving experience. The road adhesion coefficient of the target road surface is directly provided to the power distribution system of the target vehicle, and the power distribution system of the target vehicle can directly adjust the driving force according to the road adhesion coefficient of the target road surface to ensure that the target vehicle maintains good traction and stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 A schematic diagram of a flow chart of an information providing method provided in an embodiment of the present invention;

[0065] Figure 2 A flowchart of an example of combining a hidden state vector of vehicle parameters with a road surface image feature vector of a road surface image related to a target moment to predict a prediction result related to a target road surface;

[0066] Figure 3 is a schematic diagram of an example structure of a target neural network;

[0067] Figure 4 A schematic diagram of a flow chart of another information providing method provided by an embodiment of the present invention;

[0068] Figure 5 A flowchart for an example of generating features for prediction;

[0069] Figure 6 A schematic diagram of the hardware structure of a computer device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0070] The following will describe the embodiments of the present invention with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are only for illustrating the present invention, not for limiting the scope of protection of the present invention.

[0071] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention, and thus the drawings only show components related to the present invention rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0072] refer to Figure 1 , which shows a flow chart of the information providing method provided by an embodiment of the present invention. It should be noted that, during the driving of the target vehicle, step S101 to step S102 may be repeatedly performed.

[0073] In step S101, information for prediction corresponding to a target time is obtained.

[0074] The information for prediction corresponding to the target time includes: a plurality of first vehicle parameters of the target vehicle at the target time, and a road surface image related to the target time.

[0075] The road surface image related to the target moment includes: a road surface object representing the target road surface.

[0076] The target road surface may be: the road surface on which the target vehicle is traveling at the target time.

[0077] In a possible implementation, the multiple first vehicle parameters of the target vehicle at the target time include but are not limited to: the wheel inertia torque of the target vehicle at the target time, the wheel longitudinal force of the target vehicle at the target time, the suspension height of the target vehicle at the target time, the body vertical acceleration of the target vehicle at the target time, and the yaw rate of the target vehicle at the target time. In other words, the wheel inertia torque of the target vehicle at the target time, the wheel longitudinal force of the target vehicle at the target time, the suspension height of the target vehicle at the target time, the body vertical acceleration of the target vehicle at the target time, and the yaw rate of the target vehicle at the target time are respectively used as the first vehicle parameters of the target vehicle at the target time.

[0078] In the embodiment of the present invention, the first vehicle parameter of the target vehicle at the target time is collected at the target time. In other words, the collection time of the first vehicle parameter of the target vehicle at the target time is the target time.

[0079] As an example, multiple first vehicle parameters of the target vehicle at the target time include: the inertia torque of the wheels of the target vehicle at the target time, the longitudinal force of the wheels of the target vehicle at the target time, the suspension height of the target vehicle at the target time, the body vertical acceleration of the target vehicle at the target time, and the yaw angular velocity of the target vehicle at the target time.

[0080] In the embodiment of the present invention, the target moment and the moment before the target moment are two adjacent moments used for prediction among all moments used for prediction.

[0081] In a possible implementation, the moment used for prediction is: a moment whose time interval with the moment when the target vehicle is started is a preset time interval or a multiple of the preset time interval.

[0082] In a possible implementation, the road surface image related to the target time may be: a road surface image having a collection time closest to the target time.

[0083] In a possible implementation, in order to make the acquisition time of the road image correspond more accurately to the acquisition time of the vehicle parameter of the target vehicle, for the vehicle parameter i of the target vehicle, the acquisition time of the vehicle parameter i of the target vehicle can be obtained by correcting the original acquisition time of the vehicle parameter i output by the sensor used to collect the vehicle parameter i. The vehicle parameter i of the target vehicle can be any collected parameter of the target vehicle. image is the acquisition time of the road image with the closest acquisition time to the original acquisition time of vehicle parameter i, t sensoris the original acquisition time of vehicle parameter i output by the sensor for acquiring vehicle parameter i. v is the speed of the target vehicle at the original acquisition time of the vehicle parameter i, The collection time of vehicle parameter i of the target vehicle The acquisition time t of the vehicle parameter i of the target vehicle adjusted The acquisition time t of the road image with the closest acquisition time to the original acquisition time of vehicle parameter i image Closer.

[0084] In step S102, a target neural network is used to predict a prediction result related to a target road surface based on the prediction information corresponding to the target moment.

[0085] The prediction results related to the target road surface include, but are not limited to: the road surface adhesion coefficient of the target road surface, the bumpiness of the target road surface, and the road surface slope of the target road surface.

[0086] In step S102, the information for prediction corresponding to the target time is input into the target neural network, and the target neural network outputs the road adhesion coefficient of the target road surface, the bumpiness of the target road surface, and the road slope of the target road surface.

[0087] The target neural network can be any neural network that can perform regression tasks.

[0088] As an example, the target neural network is a convolutional neural network or a deep neural network.

[0089] It should be noted that the target neural network is trained. Before step S101, the target neural network is trained using the training data set of the target neural network.

[0090] During the training of the target neural network, the parameters of the target neural network are updated. In order to obtain the training data set of the target neural network, the vehicle used for training is driven on each of the various road surfaces such as snow, mud, rock, sand, asphalt, cement, and wading.

[0091] For each of the multiple road surfaces, while the vehicle for training is traveling on the road surface, the vehicle for training collects data of the type to which the data in the information for prediction corresponding to the target moment belongs. Based on the data collected while the vehicle for training is traveling on the road surface, the training data for the road surface is determined. The training data for each of the multiple road surfaces constitutes a training data set for the target neural network.

[0092] As an example, the vehicle used for training collects vehicle-side parameters such as the inertia torque of the wheels of the vehicle used for training, the longitudinal force of the wheels of the vehicle used for training, the suspension height of the vehicle used for training, the body vertical acceleration of the vehicle used for training, the yaw angular velocity of the vehicle used for training, and collects road surface images including road surface objects representing the road surface on which the vehicle used for training is traveling.

[0093] The training data in the training data set of the target neural network includes: a plurality of first vehicle parameters of the target vehicle at a time for training, and a road surface image related to the time for training.

[0094] The training data in the training data set of the target neural network are labeled to obtain the labeling result of the training data of the target neural network. The labeling result of the training data of the target neural network is: the result expected to be output by the target neural network when the training data is input into the target neural network.

[0095] The labeling results of the training data include: the labeled road adhesion coefficient, the labeled road bumpiness, and the labeled road slope.

[0096] During the training of the target neural network, the training data in the training data set of the target neural network is input into the target neural network, and the target neural network outputs the prediction result corresponding to the training data. The loss between the prediction result corresponding to the training data and the labeled result of the training data is calculated, and the loss is used to update the parameters of the target neural network.

[0097] In a possible implementation, step S102 includes: step S1021-step S1023. Step S1021-step S1023 combines the hidden state vector of the vehicle parameters with the road surface image feature vector of the road surface image related to the target moment to predict the prediction result related to the target road surface.

[0098] Through steps S1021 to S1023, the road adhesion coefficient of the target road surface, the bumpiness of the target road surface, and the road slope of the target road surface are predicted by using various types of information, namely, the vehicle parameters of the target vehicle and the road surface image related to the target moment, so as to improve the richness of the relevant information used to predict the road surface on which the vehicle is traveling, and the predicted road adhesion coefficient of the target road surface, the bumpiness of the target road surface, and the road slope of the target road surface have high accuracy. Specifically, the prediction results related to the road surface on which the vehicle is traveling are predicted by using relatively rich features, namely, the features of the vehicle parameters of the target vehicle, namely, the hidden state vectors of the vehicle parameters of the target vehicle, and the road surface image feature vectors of the road surface image related to the target moment, and the predicted results related to the target road surface have high accuracy, thereby improving the accuracy of the prediction results related to the target road surface and improving the accuracy of the adjustment of the corresponding system of the target vehicle. At the same time, the characteristics of the vehicle parameters of the target vehicle at two moments with a high degree of temporal correlation, namely, the hidden state vector of the first vehicle parameter of the target vehicle at the target moment and the hidden state vector of the second vehicle parameter of the target vehicle at the moment before the target moment, are used to predict the prediction result related to the road surface on which the vehicle is traveling. The characteristics of the vehicle parameters of the target vehicle used to predict the prediction result related to the road surface on which the vehicle is traveling are relatively rich, further improving the accuracy of predicting the prediction result related to the road surface on which the vehicle is traveling. Further improve the accuracy of the corresponding system of the target vehicle in making corresponding adjustments according to the corresponding items in the prediction result related to the target road surface.

[0099] refer to Figure 2 , which shows a flowchart of an example of combining a hidden state vector of a vehicle parameter with a road surface image feature vector of a road surface image related to a target moment to predict a prediction result related to a target road surface.

[0100] In this example, information for prediction corresponding to the target time is input into the target neural network. The information for prediction corresponding to the target time includes: a plurality of first vehicle parameters of the target vehicle at the target time, and a road surface image related to the target time.

[0101] In this example, the target neural network generates a hidden state vector of each first vehicle parameter of the target vehicle at the target time. The target neural network extracts a road surface image feature vector of the road surface image associated with the target time.

[0102] In this example, the target neural network generates a target hidden state vector based on the hidden state vector of multiple first vehicle parameters of the target vehicle at the target time and the hidden state vector of each second vehicle parameter of multiple second vehicle parameters of the target vehicle at the previous moment of the target time.

[0103] It should be noted that the hidden state vector of the plurality of second vehicle parameters of the target vehicle at the moment before the target moment is generated when predicting the prediction result related to the target road surface at the moment before the target moment.

[0104] In this example, the target neural network generates features for prediction based on the target hidden state vector and the road surface image feature vector of the road surface image related to the target time. The target neural network predicts a prediction result related to the target road surface based on the features for prediction.

[0105] The prediction results related to the target road surface include: the road adhesion coefficient of the target road surface, the road bumpiness of the target road surface, and the road slope of the target road surface.

[0106] refer to Figure 3 , which shows a schematic diagram of an example structure of a target neural network.

[0107] The target neural network includes: a vehicle parameter feature extraction network, a road surface image feature extraction network, a feature generation unit for prediction, and a prediction unit.

[0108] The type of vehicle parameter feature extraction network is Long Short-Term Memory (LSTM) neural network. The type of road surface image feature extraction network is Convolutional Neural Network.

[0109] The vehicle parameter feature extraction network receives the corresponding parameters at the target time and outputs the hidden state vector of the corresponding parameters at the corresponding time.

[0110] The road surface image feature extraction network receives the corresponding road surface image and outputs a road surface image feature vector of the corresponding road surface image.

[0111] The feature generation unit for prediction generates features for prediction according to the hidden state vector output by the vehicle parameter feature extraction network and the road surface image feature vector output by the road surface image feature extraction network.

[0112] The prediction unit may include: a fully connected layer for predicting the road adhesion coefficient, bumpiness, and road slope. The features used for prediction are input into the fully connected layer for predicting the road adhesion coefficient, bumpiness, and road slope. The fully connected layer for predicting the road adhesion coefficient, bumpiness, and road slope linearly changes the feature vector features used for prediction through a weight matrix and a bias vector. The fully connected layer for predicting the road adhesion coefficient, bumpiness, and road slope has three output nodes, and the three output nodes are respectively an output node for outputting the road adhesion coefficient, an output node for outputting the road bumpiness, and an output node for outputting the road slope.

[0113] In step S1021, a hidden state vector of each first vehicle parameter of a plurality of first vehicle parameters of the target vehicle at a target time is generated, and a road surface image feature vector of a road surface image related to the target time is extracted.

[0114] In step S1021, a hidden state vector of each first vehicle parameter of a target vehicle at a target time is generated by a vehicle parameter feature extraction network in a target neural network.

[0115] It should be noted that the hidden state vectors of the multiple second vehicle parameters of the target vehicle at the previous moment of the target moment are generated when predicting the prediction results related to the road surface represented by the road surface object in the road surface image related to the previous moment. Among them, predicting the prediction results related to the road surface represented by the road surface object in the road surface image related to the previous moment is before step S301. The process of generating the hidden state vectors of the multiple second vehicle parameters of the target vehicle at the previous moment of the target moment is the same as the process of generating the hidden state vectors of each first vehicle parameter among the multiple first vehicle parameters of the target vehicle at the target moment.

[0116] The second vehicle parameter of the target vehicle at the moment before the target moment is collected at the moment before the target moment, that is, the collection moment of the second vehicle parameter of the target vehicle at the moment before the target moment is: the moment before the target moment.

[0117] In the embodiment of the present invention, the number of first vehicle parameters included in the plurality of first vehicle parameters is equal to the number of second vehicle parameters included in the plurality of second vehicle parameters.

[0118] The plurality of first vehicle parameters correspond one-to-one to the plurality of second vehicle parameters.

[0119] For a first vehicle parameter among the plurality of first vehicle parameters, a type of the first vehicle parameter is the same as a type of a second vehicle parameter corresponding to the first vehicle parameter.

[0120] As an example, a plurality of first vehicle parameters of the target vehicle at the target time include: the inertia torque of the wheels of the target vehicle at the target time, the longitudinal force of the wheels of the target vehicle at the target time, the suspension height of the target vehicle at the target time, the vertical acceleration of the body of the target vehicle at the target time, and the yaw rate of the target vehicle at the target time. A plurality of second vehicle parameters of the target vehicle at the last moment before the target time include: the inertia torque of the wheels of the target vehicle at the last moment before the target time, the longitudinal force of the wheels of the target vehicle at the last moment before the target time, the suspension height of the target vehicle at the last moment before the target time, the vertical acceleration of the body of the target vehicle at the last moment before the target time, and the yaw rate of the target vehicle at the last moment before the target time. The inertia torque of the target vehicle's wheels at the target moment corresponds to the inertia torque of the target vehicle's wheels at the previous moment before the target moment, the wheel longitudinal force of the target vehicle at the target moment corresponds to the wheel longitudinal force of the target vehicle at the previous moment before the target moment, the suspension height of the target vehicle at the target moment corresponds to the suspension height of the target vehicle at the previous moment before the target moment, the body vertical acceleration of the target vehicle at the target moment corresponds to the body vertical acceleration of the target vehicle at the previous moment before the target moment, and the yaw angular velocity of the target vehicle at the target moment corresponds to the yaw angular velocity of the target vehicle at the previous moment before the target moment.

[0121] In step S1021, for each first vehicle parameter of a plurality of first vehicle parameters of the target vehicle at a target time, a vehicle parameter feature extraction network receives the first vehicle parameter, and the vehicle parameter feature extraction network outputs a hidden state vector of the first vehicle parameter.

[0122] The type of vehicle parameter feature extraction network is: long short-term memory neural network.

[0123] In one possible implementation, for a first vehicle parameter of the target vehicle at a target time, the input layer of the target neural network encodes the first vehicle parameter to obtain an embedding vector of the first vehicle parameter; the embedding vector of the first vehicle parameter is used as the input of a hidden unit in a vehicle parameter feature extraction network for generating a hidden state vector of a parameter of the type to which the first vehicle parameter belongs; and the hidden unit for generating a hidden state vector of a parameter of the type to which the first vehicle parameter belongs outputs the hidden state vector of the first vehicle parameter.

[0124] The hidden state vector for generating the first vehicle parameter can be expressed as:

[0125] h k =[σ(S o ·s k +S o ·h k-1 +bo )]⊙tanh(c k )

[0126] c k =[σ(S f ·s k +S f ·h k-1 +b f )]⊙c k-1 +[σ(S i ·s k +S i ·h k-1 +b i )]⊙[tanh(S c ·s k +S c ·h k-1 +b c )]

[0127] Among them, h k The hidden state vector representing the first vehicle parameter, s k represents the embedding vector of the first vehicle parameter, h k-1 represents the hidden state vector of the second vehicle parameter corresponding to the first vehicle parameter, c k represents the cell state at the corresponding moment, c k-1 represents the unit state at the previous moment of the target moment, S o represents the weight matrix of the output gate, b o Represents the bias of the output gate, S f represents the weight matrix of the forget gate, b f represents the bias of the forget gate, S i represents the weight matrix of the input gate, b i represents the bias of the input gate, S c The weight matrix representing the unit state, b c represents the bias of the unit state, σ represents the activation function, and ⊙ represents the effective product.

[0128] In step S1021, a road surface image related to a target time may be received by a road surface image feature extraction network, and the road surface image feature extraction network outputs a road surface image feature vector of the road surface image.

[0129] In step S1022, a target hidden state vector is generated based on the hidden state vector of each first vehicle parameter of the target vehicle at the target time and the hidden state vector of each second vehicle parameter of the target vehicle at the previous moment before the target time.

[0130] In a possible implementation of step S1022, a vehicle parameter combination vector including a hidden state vector of each first vehicle parameter among a plurality of first vehicle parameters of the target vehicle at a target time is generated, wherein the hidden state vector of each first vehicle parameter among the plurality of first vehicle parameters of the target vehicle at the target time is respectively used as a component of the vehicle parameter combination vector including the hidden state vector of each first vehicle parameter among the plurality of first vehicle parameters of the target vehicle at the target time; a vehicle parameter combination vector including a hidden state vector of each second vehicle parameter of the target vehicle at a previous moment before the target time is generated, wherein the hidden state vector of each second vehicle parameter of the target vehicle at a previous moment before the target time is respectively used as a component of the vehicle parameter combination vector including the hidden state vector of each second vehicle parameter of the target vehicle at a previous moment before the target time; and the vehicle parameter combination vector including the hidden state vector of each first vehicle parameter among the plurality of first vehicle parameters of the target vehicle at the target time is added to the vehicle parameter combination vector including the hidden state vector of each second vehicle parameter of the target vehicle at a previous moment before the target time to obtain a target hidden state vector.

[0131] In step S1023, features for prediction are generated based on the target hidden state vector and the road surface image feature vector of the road surface image related to the target moment, and a prediction result related to the target road surface is predicted based on the features for prediction.

[0132] In a possible implementation of step S1023, the target hidden state vector is added to the road surface image feature vector to obtain the feature for prediction. When the hidden state vector of the vehicle parameters is combined with the road surface image feature vector of the road surface image related to the target moment to predict the prediction result related to the target road surface through steps S1021-S1023, the target neural network can be trained in the following manner: the target neural network is trained using the training data set of the target neural network. During the training of the target neural network, the parameters of the target neural network are updated. In order to obtain the training data set of the target neural network, the vehicle used for training is driven on each of the various road surfaces such as snow, mud, rock, sand, asphalt, cement, wading, etc. For each of the various road surfaces, while the vehicle used for training is driving on the road surface, the vehicle used for training collects data of the type of data corresponding to the information used for prediction at the target moment. According to the data collected by the vehicle used for training during the driving on the road surface, the training data of the road surface is determined. The training data of each of the various road surfaces constitute the training data set of the target neural network. As an example, the vehicle used for training collects vehicle-side parameters such as the inertia torque of the wheels of the vehicle used for training, the longitudinal force of the wheels of the vehicle used for training, the suspension height of the vehicle used for training, the vertical acceleration of the body of the vehicle used for training, and the yaw rate of the vehicle used for training, and collects road surface images including road surface objects representing the road surface on which the vehicle used for training is traveling. The training data in the training data set of the target neural network includes: multiple first vehicle parameters of the target vehicle at the time of training, multiple second vehicle parameters of the target vehicle at the previous time of the target time of training, and road surface images related to the time of training. The training data in the training data set of the target neural network are labeled to obtain the labeling results of the training data of the target neural network. The labeling results of the training data of the target neural network are: the results expected to be output by the target neural network when the training data is input into the target neural network. The labeling results of the training data include: the labeled road adhesion coefficient, the labeled road bumpiness, and the labeled road slope. During the training of the target neural network, the training data in the training data set of the target neural network are input into the target neural network, and the target neural network outputs the prediction results corresponding to the training data. The loss between the predicted results corresponding to the training data and the labeled results of the training data is calculated, and the loss is used to update the parameters of the target neural network.

[0133] In step S103, each item in the prediction result is provided to a corresponding system of the target vehicle, so that the corresponding system of the target vehicle makes corresponding adjustments according to the corresponding item in the prediction result.

[0134] The prediction results related to the target road surface include: the road surface adhesion coefficient of the target road surface, the bumpiness of the target road surface, and the road surface slope of the target road surface. The road surface adhesion coefficient of the target road surface, the bumpiness of the target road surface, and the road surface slope of the target road surface can each be provided to a corresponding system of the target vehicle.

[0135] As an example, the bumpiness of the target road surface is directly provided to the suspension system of the target vehicle, and the suspension system of the target vehicle can be adaptively adjusted directly according to the bumpiness of the target road surface to provide a smoother driving experience. The road adhesion coefficient of the target road surface is directly provided to the power distribution system of the target vehicle, and the power distribution system of the target vehicle can directly adjust the driving force according to the road adhesion coefficient of the target road surface to ensure that the target vehicle maintains good traction and stability. The road slope of the target road surface is directly provided to the power output system of the target vehicle.

[0136] In addition, the prediction results related to the target road surface may also include: in addition to the road adhesion coefficient of the target road surface, the bumpiness of the target road surface, and the road slope of the target road surface, other information required for the corresponding system of the target vehicle to make corresponding adjustments. For example, the prediction results related to the target road surface may also include: the wetness of the target road surface. The wetness of the target road surface may be provided to the vehicle control system of the target vehicle, and the vehicle control system of the target vehicle may automatically adjust the vehicle speed directly according to the wetness of the road surface to reduce the risk of skidding.

[0137] In step S103, the road adhesion coefficient of the target road, the bumpiness of the target road, and the road slope of the target road can be provided to the chassis system of the target vehicle, so that the chassis system of the target vehicle can optimize the suspension of the target vehicle, the braking of the target vehicle, and the traction control of the target vehicle according to the road adhesion coefficient of the target road, the bumpiness of the target road, and the road slope of the target road.

[0138] In a possible implementation, each item in the prediction result is provided to the corresponding system of the target vehicle, so that the corresponding system of the target vehicle makes corresponding adjustments according to the corresponding items in the prediction result, including: providing the road adhesion coefficient of the target road surface to the traction control system of the target vehicle so that the traction control system of the target vehicle adjusts the output power of the engine of the target vehicle according to the road adhesion coefficient of the target road surface; providing the bumpiness of the target road surface to the suspension system of the target vehicle so that the suspension system of the target vehicle adjusts the damping rate according to the bumpiness of the target road surface; providing the road surface slope of the target road surface to the torque distribution system of the target vehicle so that the torque distribution system of the target vehicle adjusts the engine torque of the target vehicle according to the road surface slope of the target road surface. Wherein, the adhesion coefficient represents the friction between the tire and the road surface, and is an important indicator for measuring the vehicle's grip. When the road adhesion coefficient of the target road surface is low (such as the target road surface is slippery, the target road surface is an icy road surface), the traction control (TCS) system of the chassis system of the target vehicle can limit the output power of the engine of the target vehicle, prevent the drive wheel from idling, and ensure that the target vehicle can accelerate smoothly. When the target road surface has a high adhesion coefficient (such as a dry asphalt road), the traction control system will allow higher power output and improve acceleration performance. A road surface with a low adhesion coefficient usually means poor grip. When the target road surface has a low adhesion coefficient, the suspension system of the chassis system of the target vehicle can choose a softer mode to absorb vibration and improve the continuity of tire contact. When the target road surface has a high adhesion coefficient, the suspension system of the target vehicle can choose a harder mode to increase handling and improve driving stability. Bumpiness reflects the unevenness of the road surface and will directly affect the stability and ride comfort of the vehicle. The chassis system can adjust the suspension damping, stiffness and other parameters to enhance the response to vibration and improve the driving experience. When the target road surface has a high bumpiness (such as potholes on the target road surface and gravel on the target road surface), the suspension system of the target vehicle can dynamically adjust the damping rate according to the bumpiness information, so that the suspension system of the target vehicle absorbs the impact of the road surface, reduces the body shaking and vibration of the target vehicle, and improves the comfort in the car. When the target road surface has a low bumpiness, such as a flat road surface, the target vehicle's suspension system can increase damping to provide better handling. The road slope of the target road surface provides information about the angle of the road surface inclination, which has an important impact on the vehicle's traction and braking force distribution. When the road slope of the target road surface indicates a large slope, such as when the target vehicle is driving on an uphill section with a large slope, the target vehicle's chassis system's torque distribution system will increase the target vehicle's engine torque and distribute more traction to the drive wheels to prevent the target vehicle from slipping during the climbing process and ensure stable acceleration performance. When the road slope of the target road surface indicates a downhill slope, the target vehicle's engine torque will be actively reduced to prevent the target vehicle's downhill speed from getting out of control.When the road gradient of the target road indicates that the target vehicle is going down a steep slope, the Hill Descent Control (HDC) system of the chassis system of the target vehicle can be activated. The Hill Descent Control system will gradually reduce the vehicle speed by controlling the brakes to prevent the target vehicle from sliding down too fast due to excessive slope, thereby improving the downhill stability and safety of the target vehicle.

[0139] The target road surface adhesion coefficient, target road surface bumpiness, and target road surface slope reflecting the road condition of the target road surface are provided to the chassis system of the target vehicle, so that the chassis system of the target vehicle can make corresponding adjustments suitable for the road condition of the target road surface according to the information reflecting the road condition of the target road surface, thereby improving the dynamic stability, safety, and comfort of the target vehicle.

[0140] refer to Figure 4 , which shows a flow chart of another information providing method provided by an embodiment of the present invention.

[0141] In step S401, information for prediction corresponding to the target time is obtained.

[0142] In step S402, using the target neural network, based on the information for prediction corresponding to the target moment, the prediction result related to the target road surface is predicted, including: generating a target hidden state vector based on the attention hidden state vector of each first vehicle parameter of the target vehicle at the target moment and the hidden state vector of each second vehicle parameter of the target vehicle at the previous moment of the target moment.

[0143] Step S4022 includes: step S40221-step S40223.

[0144] In step S40221, an attention parameter of each first vehicle parameter of the target vehicle at the target time is determined based on the hidden state vector of each first vehicle parameter of the target vehicle at the target time.

[0145] In step S40222, based on the hidden state vector of each first vehicle parameter among the multiple first vehicle parameters of the target vehicle at the target time and the attention parameter of each first vehicle parameter of the target vehicle at the target time, an attention hidden state vector of each first vehicle parameter of the target vehicle at the target time is generated.

[0146] In step S40223, a target hidden state vector is generated based on the attention hidden state vector of each first vehicle parameter of the target vehicle at the target moment and the hidden state vector of each second vehicle parameter of the target vehicle at the previous moment of the target moment.

[0147] Step S4022 takes into account the importance of the hidden state vector of the first vehicle parameter of the target vehicle at the target time for predicting the prediction result related to the target road surface. The size of the attention parameter of the first vehicle parameter of the target vehicle at the target time can reflect the importance of the hidden state vector of the first vehicle parameter of the target vehicle at the target time. The larger the attention parameter of the first vehicle parameter of the target vehicle at the target time, the higher the importance of the hidden state vector of the first vehicle parameter of the target vehicle at the target time. The smaller the attention parameter of the first vehicle parameter of the target vehicle at the target time, the lower the importance of the hidden state vector of the first vehicle parameter of the target vehicle at the target time. Thus, the accuracy of the target hidden state vector can be improved, and the accuracy of the prediction result related to the road surface on which the vehicle is traveling can be improved.

[0148] In a possible implementation of step S40221, the target neural network includes: a first linear layer. In step S40221, for each first vehicle parameter of the target vehicle at the target time, the hidden state vector of the first vehicle parameter is input to the input first linear layer, the first linear layer applies the first score function to the hidden state vector of the first vehicle parameter, and the first linear layer outputs the first linear layer output of the first vehicle parameter; the activation function corresponding to the first linear layer is applied to the first linear layer output of the first vehicle parameter to obtain the first attention score of the first vehicle parameter. The first attention score of each first vehicle parameter of the target vehicle at the target time is added to obtain the sum of the first attention scores of each first vehicle parameter of the target vehicle at the target time. For each first vehicle parameter of the target vehicle at the target time, the first attention score of the first vehicle parameter is divided by the sum of the first attention scores of each first vehicle parameter of the target vehicle at the target time to obtain the normalized first attention score of the first vehicle parameter at the target time, and the normalized first attention score of the first vehicle parameter is determined as the attention parameter of the first vehicle parameter.

[0149] The parameters of the first linear layer include: a weight matrix of the first linear layer and a bias of the first linear layer.

[0150] The first score function includes: an item representing a weight matrix of the first linear layer, an item representing a bias of the first linear layer, and an item representing a hidden state vector as an independent variable.

[0151] The activation function corresponding to the first linear layer can be a tanh function.

[0152] Using the first linear layer and the activation function corresponding to the first linear layer, a first attention score of a first vehicle parameter can be expressed as:

[0153] γ(h i)=tanh(h i .w a +b a )

[0154] Among them, γ(h i ) represents the first attention score of the first vehicle parameter, h i represents the hidden state vector of the first vehicle parameters, w a represents the weight matrix of the first linear layer, b a represents the bias of the first linear layer.

[0155] In another possible implementation of step S40221, step S40221 includes: step S40221a-step S40221c.

[0156] In step S40221a, a first attention score of each first vehicle parameter of the target vehicle at the target time is determined based on the hidden state vector of each first vehicle parameter of the target vehicle at the target time.

[0157] In step S40221b, for each first vehicle parameter of the target vehicle at the target moment, a second attention score of the first vehicle parameter is determined based on the hidden state vector of the first vehicle parameter and the hidden state vector of the second vehicle parameter corresponding to the first vehicle parameter, wherein the type of the first vehicle parameter is the same as the type of the second vehicle parameter corresponding to the first vehicle parameter.

[0158] In step S40221c, the attention parameter of each first vehicle parameter of the target vehicle at the target moment is determined based on the first attention score of each first vehicle parameter of the target vehicle at the target moment and the second attention score of each first vehicle parameter of the target vehicle at the target moment.

[0159] Step S40221 considers the importance of the hidden state vector of the first vehicle parameter of the target vehicle at the target time when considering the hidden state vector of the second vehicle parameter corresponding to the first vehicle parameter, thereby improving the accuracy of the attention parameter of each first vehicle parameter of the target vehicle at the target time.

[0160] In a possible implementation of step S40221a, in step S40221a, for each first vehicle parameter of the target vehicle at the target time, the hidden state vector of the first vehicle parameter is input into the input first linear layer, the first linear layer applies the first score function to the hidden state vector of the first vehicle parameter, and the first linear layer outputs the first linear layer output of the first vehicle parameter; the activation function corresponding to the first linear layer is applied to the first linear layer output of the first vehicle parameter to obtain a first attention score of the first vehicle parameter.

[0161] In a possible implementation of step S40221b, the target neural network includes: a second linear layer. In step S40221b, for each first vehicle parameter of the target vehicle at the target time, the hidden state vector of the second vehicle parameter corresponding to the first vehicle parameter is pooled to obtain the pooling result of the hidden state vector of the second vehicle parameter corresponding to the first vehicle parameter; the hidden state vector of the first vehicle parameter and the pooling result of the hidden state vector of the second vehicle parameter corresponding to the first vehicle parameter are input to the second linear layer, the second linear layer applies the second score function to the hidden state vector of the first vehicle parameter and the pooling result of the hidden state vector of the second vehicle parameter corresponding to the first vehicle parameter, and the second linear layer outputs the second linear layer output of the first vehicle parameter; the activation function corresponding to the second linear layer is applied to the second linear layer output of the first vehicle parameter to obtain the second attention score of the first vehicle parameter.

[0162] The parameters of the second linear layer include: a first weight matrix, a second weight matrix, and a bias of the second linear layer.

[0163] The second score function includes: an item representing the first weight matrix, an item representing the second weight matrix, a bias item representing the second linear layer, and an item representing the hidden state vector as an independent variable.

[0164] The activation function corresponding to the second linear layer can be a tanh function.

[0165] Using the second linear layer and the activation function corresponding to the second linear layer, a second attention score of the first vehicle parameter can be expressed as:

[0166]

[0167] in, represents a second attention score of the first vehicle parameter, represents the hidden state vector of the first vehicle parameters, represents the hidden state vector of the second vehicle parameter corresponding to the first vehicle parameter, w1 represents the first weight matrix, w2 represents the second weight matrix, b represents the bias of the second linear layer, and pool represents pooling.

[0168] In a possible implementation of step S40221c, for each first vehicle parameter of the target vehicle at the target time, an exponential function with base e is applied to the second attention score of the first vehicle parameter, that is, the second attention score of the first vehicle parameter is taken as the value of the independent variable of the exponential function with base e, and the second function value corresponding to the first vehicle parameter is obtained. For each first vehicle parameter of the target vehicle at the target time, an exponential function with base e is applied to the first attention score of the first vehicle parameter, that is, the first attention score of the first vehicle parameter is taken as the value of the independent variable of the exponential function with base e, and the first function value corresponding to the first vehicle parameter is obtained. The first function values ​​corresponding to each first vehicle parameter of the target vehicle at the target time are added together to obtain the sum of the first function values ​​corresponding to each first vehicle parameter of the target vehicle at the target time. For each first vehicle parameter of the target vehicle at the target time, the second function value corresponding to the first vehicle parameter is divided by the sum of the first function values ​​corresponding to each first vehicle parameter of the target vehicle at the target time to obtain the attention parameter of the first vehicle parameter.

[0169] The attention parameter of the first vehicle parameter can be expressed as:

[0170]

[0171] Among them, α t represents the attention parameter of the first vehicle parameter, exp represents the exponential function with e as the base, represents the second attention score of the first vehicle parameter, γ(h k ) represents the first attention score of the kth first vehicle parameter of the target vehicle at the target time.

[0172] In step S40222, for each first vehicle parameter of the target vehicle at the target time, the attention parameter of the first vehicle parameter is multiplied by the hidden state vector of the first vehicle parameter to obtain the attention hidden state vector of the first vehicle parameter. The i-th component in the attention hidden state vector of the first vehicle parameter is: the product of the i-th component in the hidden state vector of the first vehicle parameter and the attention parameter of the first vehicle parameter.

[0173] In a possible implementation of S40223, in step S40223, a vehicle parameter combination vector including an attention hidden state vector of each first vehicle parameter of the target vehicle at the target moment is generated, wherein the attention hidden state vector of each first vehicle parameter of the target vehicle at the target moment is respectively used as a component of the vehicle parameter combination vector including the attention hidden state vector of each first vehicle parameter of the target vehicle at the target moment; a vehicle parameter combination vector including a hidden state vector of each second vehicle parameter of the target vehicle at a moment before the target moment is generated, wherein the hidden state vector of each second vehicle parameter of the target vehicle at a moment before the target moment is respectively used as a component of the vehicle parameter combination vector including the hidden state vector of each second vehicle parameter of the target vehicle at a moment before the target moment; the vehicle parameter combination vector including the attention hidden state vector of each first vehicle parameter of the target vehicle at the target moment is added to the vehicle parameter combination vector including the hidden state vector of each second vehicle parameter of the target vehicle at a moment before the target moment to obtain a target hidden state vector.

[0174] In another possible implementation of S40223, S40223 includes: S40223a.

[0175] In step S40223a, an attention hidden state vector of each second vehicle parameter of the target vehicle at the previous moment before the target moment is generated based on the hidden state vector of each second vehicle parameter of the target vehicle at the previous moment before the target moment and the attention parameter of each second vehicle parameter of the target vehicle at the previous moment before the target moment; a target hidden state vector is generated based on the attention hidden state vector of each first vehicle parameter of the target vehicle at the target moment and the attention hidden state vector of each second vehicle parameter of the target vehicle at the previous moment before the target moment.

[0176] The attention parameter of each second vehicle parameter of the target vehicle at a previous moment before the target moment is generated according to the hidden state vector of each second vehicle parameter of the target vehicle at a previous moment before the target moment.

[0177] It should be noted that the hidden state vector of each second vehicle parameter of the target vehicle at the previous moment before the target moment and the attention parameter of each second vehicle parameter of the target vehicle at the previous moment before the target moment are generated when predicting the prediction result related to the road surface represented by the road surface object in the road surface image related to the previous moment. Predicting the prediction result related to the road surface represented by the road surface object in the road surface image related to the previous moment is before step S401.

[0178] Step S40223a takes into account the importance of the hidden state vector of the second vehicle parameter of the target vehicle at the target moment for predicting the prediction result related to the target road surface. The size of the attention parameter of the second vehicle parameter of the target vehicle at the previous moment of the target moment can reflect the importance of the hidden state vector of the second vehicle parameter of the target vehicle at the target moment. The larger the attention parameter of the second vehicle parameter of the target vehicle at the previous moment of the target moment, the higher the importance of the hidden state vector of the second vehicle parameter of the target vehicle at the previous moment of the target moment. The smaller the attention parameter of the second vehicle parameter of the target vehicle at the previous moment of the target moment, the lower the importance of the hidden state vector of the second vehicle parameter of the target vehicle at the previous moment of the target moment. Thus, the accuracy of the target hidden state vector can be improved, and the accuracy of the prediction result related to the road surface on which the vehicle is traveling can be improved.

[0179] The process of determining the attention parameter of each second vehicle parameter of the target vehicle at the previous moment before the target moment according to the hidden state vector of each second vehicle parameter of the target vehicle at the previous moment before the target moment is the same as the process of determining the attention parameter of each first vehicle parameter of the target vehicle at the target moment according to the hidden state vector of each first vehicle parameter among the multiple first vehicle parameters of the target vehicle at the target moment in step S40221.

[0180] According to the hidden state vector of each second vehicle parameter of the target vehicle at the previous moment before the target moment, the implementation of determining the attention parameter of each second vehicle parameter of the target vehicle at the previous moment before the target moment can refer to any possible implementation method of step S40221.

[0181] The target moment is recorded as moment t, the moment before the target moment is recorded as moment t-1, and the moment before moment t-1 is recorded as moment t-2.

[0182] By replacing the first vehicle parameters of the target vehicle at time t in a possible implementation method of the above step S40221 with the second vehicle parameters of the target vehicle at time t-1, an implementation method of determining the attention parameters of each second vehicle parameter of the target vehicle at the previous moment before the target moment can be obtained based on the hidden state vector of each second vehicle parameter of the target vehicle at the previous moment before the target moment.

[0183] By replacing the first vehicle parameters of the target vehicle at time t in another possible implementation of step S40221 above with the second vehicle parameters of the target vehicle at time t-1, and replacing the vehicle parameters of the target vehicle at time t-2 with the second vehicle parameters of the target vehicle at time t-1 in another possible implementation of step S40221 above, another implementation method of determining the attention parameters of each second vehicle parameter of the target vehicle at the previous moment before the target moment can be obtained based on the hidden state vector of each second vehicle parameter of the target vehicle at the previous moment before the target moment.

[0184] The attention hidden state vector of each second vehicle parameter of the target vehicle at the previous moment of the target moment is generated based on the attention parameter of each second vehicle parameter of the target vehicle at the previous moment of the target moment and the hidden state vector of each second vehicle parameter of the target vehicle at the previous moment of the target moment.

[0185] For each second vehicle parameter of the target vehicle at a previous moment before the target moment, the attention hidden state vector of the second vehicle parameter is obtained by multiplying the attention parameter of the second vehicle parameter by the hidden state vector of the second vehicle parameter.

[0186] In a possible implementation of step 40223a, a first vehicle parameter combination vector is generated, including an attention hidden state vector of each first vehicle parameter of the target vehicle at the target moment, wherein the attention hidden state vector of each first vehicle parameter of the target vehicle at the target moment is respectively used as a component of the first vehicle parameter combination vector; a second vehicle parameter combination vector is generated, including an attention hidden state vector of each second vehicle parameter of the target vehicle at a moment before the target moment, wherein the attention hidden state vector of each second vehicle parameter of the target vehicle at a moment before the target moment is respectively used as a component of the second vehicle parameter combination vector; the first vehicle parameter combination vector is added to the second vehicle parameter combination vector to obtain a target hidden state vector.

[0187] refer to Figure 5 , which shows a flowchart of an example of obtaining a target hidden state vector.

[0188] In this example, time t is the target time, and time t-1 is the previous time of the target time.

[0189] In this example, the hidden state vector of each second vehicle parameter of the target vehicle at time t-1 is obtained by inputting each second vehicle parameter of the target vehicle at time t-1 into the target neural network respectively. The attention parameter of each second vehicle parameter of the target vehicle at time t-1 is obtained based on the hidden state vector of each second vehicle parameter of the target vehicle at time t-1.

[0190] It should be noted that the hidden state vector of each second vehicle parameter of the target vehicle at time t-1 and the attention parameter of each second vehicle parameter of the target vehicle at time t-1 are generated when predicting the road surface-related prediction results represented by the road surface object in the road surface image related to time t-1.

[0191] In this example, the hidden state vector of each first vehicle parameter of the target vehicle at time t is obtained by respectively inputting each first vehicle parameter of the target vehicle at time t into the target neural network.

[0192] In this example, the hidden state vector of each second vehicle parameter of the target vehicle at time t-1 is pooled to obtain the pooling result of the hidden state vector of each second vehicle parameter of the target vehicle at time t-1. Based on the pooling result of the hidden state vector of each second vehicle parameter of the target vehicle at time t-1 and the hidden state vector of each first vehicle parameter of the target vehicle at t, the attention hidden state vector of each first vehicle parameter of the target vehicle at time t is obtained.

[0193] In this example, a target hidden state vector is generated based on the attention hidden state vector of each first vehicle parameter of the target vehicle at time t and the hidden state vector of each second vehicle parameter of the target vehicle at time t-1.

[0194] In a possible implementation of step S40223, the target hidden state vector is multiplied by the vehicle parameter feature weight to obtain a first weighted vector, and the road surface image feature vector of the road surface image related to the target moment is multiplied by the road surface image feature weight to obtain a second weighted vector; based on the first weighted vector and the second weighted vector, features for prediction are generated.

[0195] The i-th component in the first weighted vector is: the product of the i-th component in the target hidden state vector and the vehicle parameter feature weight.

[0196] The i-th component in the second weighted vector is: the product of the i-th component in the road surface image feature vector of the road surface image related to the target moment and the road surface image feature weight.

[0197] In a possible implementation, generating a feature for prediction according to the first weighted vector and the second weighted vector includes: adding the first weighted vector and the second weighted vector to obtain the feature for prediction.

[0198] In one possible implementation, before step S401, the vehicle parameter feature weights and the road surface image feature weights are determined based on the accuracy of the vehicle parameter feature extraction network in the target neural network on the validation set of the target neural network and the accuracy of the road surface image feature extraction network in the target neural network on the validation set of the target neural network.

[0199] The number of times the vehicle parameter feature extraction network's prediction output on the validation data is accurate divided by the number of times the validation data is input into the vehicle parameter feature extraction network.

[0200] The number of times the road image feature extraction network's prediction output on the validation data is accurate divided by the number of times the validation data is input into the vehicle parameter feature extraction network.

[0201] In one possible implementation, it is determined whether the similarity between the prediction result corresponding to the verification data output by the vehicle parameter feature extraction network in the target neural network and the vehicle parameter feature annotation output corresponding to the verification data is greater than a similarity threshold when the verification data is input into the target neural network. If so, the prediction of the vehicle parameter feature extraction network on the verification data is accurate. If not, the prediction of the vehicle parameter feature extraction network on the verification data is inaccurate. The vehicle parameter feature annotation output corresponding to the verification data is the result of the expected vehicle parameter feature extraction network output when the verification data is input into the vehicle parameter feature extraction network. It is determined whether the similarity between the prediction result corresponding to the verification data output by the road surface image feature extraction network in the target neural network and the road surface image feature annotation output corresponding to the verification data is greater than a similarity threshold when the verification data is input into the target neural network. If so, the prediction of the road surface image feature extraction network on the verification data is accurate. If not, the prediction of the road surface image feature extraction network on the verification data is inaccurate.

[0202] The vehicle parameter feature weight θ can be calculated using the following formula:

[0203]

[0204] Among them, A S represents the accuracy of the vehicle parameter feature extraction network on the validation set of the target neural network, A i Represents the accuracy of the road image feature extraction network on the validation set of the target neural network.

[0205] The road surface image feature weight μ can be calculated using the following formula:

[0206]

[0207] In step S403, each item in the prediction result is provided to the corresponding system of the target vehicle, so that the corresponding system of the target vehicle makes corresponding adjustments according to the corresponding item in the prediction result.

[0208] The process of step S403 refers to the process of step S303.

[0209] An information providing device is also provided in the embodiment of the present invention. The information providing device is installed on the vehicle. The information providing device is used to implement the above-mentioned method embodiment and preferred implementation mode. The descriptions that have been made will not be repeated here. As used below, the term "unit" can implement a combination of software and / or hardware for a predetermined function. Although the device described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceived. The device in the embodiment of the present invention is presented in the form of a functional unit, where the functional unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that executes one or more software or fixed programs, and / or other devices that can provide the above-mentioned functions.

[0210] The information providing device includes:

[0211] an acquisition unit, configured to acquire information for prediction corresponding to a target time, the information for prediction comprising: a plurality of first vehicle parameters of the target vehicle at the target time, and a road surface image related to the target time, the road surface image comprising a road surface object representing the target road surface;

[0212] A prediction unit, used to predict a prediction result related to the target road surface by using a target neural network according to the information for prediction, wherein the prediction result includes: a road surface adhesion coefficient of the target road surface, a bumpiness of the target road surface, and a road surface slope of the target road surface;

[0213] A providing unit is used to provide each item in the prediction result to a corresponding system of the target vehicle, so that the corresponding system can make corresponding adjustments according to the corresponding item in the prediction result.

[0214] In one possible implementation, a unit is provided, which is further used to provide each item in the prediction results to a corresponding system of the target vehicle, so that the corresponding system makes corresponding adjustments according to the corresponding items in the prediction results, including: providing the road adhesion coefficient of the target road to the traction control system of the target vehicle so that the traction control system of the target vehicle adjusts the output power of the engine of the target vehicle according to the road adhesion coefficient of the target road; providing the bumpiness of the target road to the suspension system of the target vehicle so that the suspension system of the target vehicle adjusts the damping rate according to the bumpiness of the target road; providing the road slope of the target road to the torque distribution system of the target vehicle so that the torque distribution system of the target vehicle adjusts the engine torque of the target vehicle according to the road slope of the target road.

[0215] In one possible implementation, the prediction unit is further used to generate a hidden state vector of each first vehicle parameter among the multiple first vehicle parameters, and to extract a road surface image feature vector of the road surface image; to generate a target hidden state vector based on the hidden state vector of each first vehicle parameter among the multiple first vehicle parameters and the hidden state vector of each second vehicle parameter among the multiple second vehicle parameters of the target vehicle at the previous moment before the target moment; to generate features for prediction based on the target hidden state vector and the road surface image feature vector, and to predict the prediction result based on the features for prediction.

[0216] In one possible implementation, the prediction unit is further used to determine the attention parameter of each first vehicle parameter among the multiple first vehicle parameters based on the hidden state vector of each first vehicle parameter; generate the attention hidden state vector of each first vehicle parameter according to the hidden state vector of each first vehicle parameter among the multiple first vehicle parameters and the attention parameter of each first vehicle parameter; generate the target hidden state vector according to the attention hidden state vector of each first vehicle parameter and the hidden state vector of each second vehicle parameter.

[0217] In one possible implementation, the prediction unit is further used to determine a first attention score of each first vehicle parameter among the multiple first vehicle parameters based on a hidden state vector of each first vehicle parameter; for each first vehicle parameter, determine a second attention score of the first vehicle parameter based on the hidden state vector of the first vehicle parameter and a hidden state vector of a second vehicle parameter corresponding to the first vehicle parameter, wherein the type of the first vehicle parameter is the same as the type of the second vehicle parameter corresponding to the first vehicle parameter; determine the attention parameter of each first vehicle parameter based on the first attention score of each first vehicle parameter and the second attention score of each first vehicle parameter.

[0218] In one possible implementation, the prediction unit is further used to generate an attention hidden state vector of each second vehicle parameter based on the hidden state vector of each second vehicle parameter and the attention parameter of each second vehicle parameter; and to generate a target hidden state vector based on the attention hidden state vector of each first vehicle parameter and the attention hidden state vector of each second vehicle parameter.

[0219] In one possible implementation, the prediction unit is further used to generate a first vehicle parameter combination vector including an attention hidden state vector of each first vehicle parameter, and to generate a second vehicle parameter combination vector including an attention hidden state vector of each second vehicle parameter; and to add the first vehicle parameter combination vector to the second vehicle parameter combination vector to obtain a target hidden state vector.

[0220] In one possible implementation, the prediction unit is further used to multiply the target hidden state vector by the vehicle parameter feature weight to obtain a first weighted vector, and to multiply the road surface image feature vector by the road surface image feature weight to obtain a second weighted vector; and generate features for prediction based on the first weighted vector and the second weighted vector.

[0221] In a possible implementation manner, the prediction unit is further configured to add the first weighted vector and the second weighted vector to obtain a feature for prediction.

[0222] In a possible implementation, the information providing device includes:

[0223] The weight determination unit is used to determine the vehicle parameter feature weight and the road surface image feature weight according to the accuracy of the vehicle parameter feature extraction network in the target neural network on the verification set of the target neural network and the accuracy of the road surface image feature extraction network in the target neural network on the verification set before obtaining the information for prediction corresponding to the target time.

[0224] refer to Figure 6 , Figure 61 is a schematic diagram of the hardware structure of a computer device provided by an embodiment of the present invention. The computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. The various components are connected to each other using different buses for communication, and can be installed on a common motherboard or installed in other ways as needed. The processor can process instructions executed in the computer device, including instructions stored in or on the memory to display graphical information of the GUI on an external input / output device (such as a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used with multiple memories and multiple memories. Similarly, multiple vehicles can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). The processor 10 can be a central processing unit, a network processor, or a combination thereof. Among them, the processor 10 can further include a hardware chip. The above-mentioned hardware chip can be a dedicated integrated circuit, a programmable logic device, or a combination thereof. The above-mentioned programmable logic device can be a complex programmable logic device, a field programmable gate array, a general array logic, or any combination thereof. The memory 20 stores instructions executable by at least one processor 10 so that the at least one processor 10 executes the method shown in the above embodiment. The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required by at least one function; the data storage area may store data created according to the use of the vehicle, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage devices. In some optional embodiments, the memory 20 may optionally include a memory remotely arranged relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof. The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid-state hard disk; the memory 20 may also include a combination of the above-mentioned types of memory. The computer device also includes an input device 30 and an output device 40. The processor 10, the memory 20, the input device 30, and the output device 40 may be connected via a bus or in other ways. The input device 30 can receive input digital or character information, and generate key signal input related to the user settings and function control of the computer device, such as a touch screen, a keypad, a mouse, a trackpad, a touch pad, an indicator stick, one or more mouse buttons, a trackball, a joystick, etc.The output device 40 may include a display device, an auxiliary lighting device (e.g., LED), a tactile feedback device (e.g., a vibration motor), etc. The above-mentioned display device includes but is not limited to a liquid crystal display, a light emitting diode, a display, and a plasma display. In some optional embodiments, the display device may be a touch screen.

[0225] The embodiment of the present invention also provides a computer-readable storage medium. The method according to the embodiment of the present invention can be implemented in hardware, firmware, or can be implemented as a computer code that can be recorded in a storage medium, or can be implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and will be stored in a local storage medium through a network download, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state hard disk, etc.; further, the storage medium can also include a combination of the above types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor, or hardware, the method shown in the above embodiment is implemented.

[0226] A portion of the embodiments of the present invention may be applied as a computer program product, such as a computer program instruction, which, when executed by a computer, can call or provide the method and / or technical solution according to the present invention through the operation of the computer. Those skilled in the art should understand that the existence of computer program instructions in a computer-readable medium includes, but is not limited to, source files, executable files, installation package files, etc., and accordingly, the way in which the computer program instructions are executed by the computer includes, but is not limited to: the computer directly executes the instruction, or the computer compiles the instruction and then executes the corresponding compiled program, or the computer reads and executes the instruction, or the computer reads and installs the instruction and then executes the corresponding installed program. Here, the computer-readable medium may be any available computer-readable storage medium or communication medium accessible to the computer.

[0227] The above embodiments are only preferred embodiments for fully illustrating the present invention, and the protection scope of the present invention is not limited thereto. Any equivalent substitution or change made by a person skilled in the art based on the present invention is within the protection scope of the present invention.

Claims

1. A method for providing information, characterized in that: The method comprises: Acquiring information for prediction corresponding to a target time, the information for prediction comprising: a plurality of first vehicle parameters of the target vehicle at the target time, and a road surface image related to the target time, the road surface image comprising a road surface object representing the target road surface; Using a target neural network, based on the information for prediction, predicting a prediction result related to the target road surface, the prediction result including: a road surface adhesion coefficient of the target road surface, a bumpiness of the target road surface, and a road surface slope of the target road surface; Each item in the prediction results is provided to a corresponding system of the target vehicle, so that the corresponding system can make corresponding adjustments according to the corresponding item in the prediction results.

2. The method according to claim 1, characterized in that: Providing each item in the prediction result to a corresponding system of the target vehicle respectively, so that the corresponding system performs corresponding adjustments according to the corresponding item in the prediction result, includes: providing the road surface adhesion coefficient of the target road surface to the traction control system of the target vehicle so that the traction control system of the target vehicle adjusts the output power of the engine of the target vehicle according to the road surface adhesion coefficient of the target road surface; Providing the bumpiness of the target road surface to a suspension system of a target vehicle so that the suspension system of the target vehicle adjusts a damping rate according to the bumpiness of the target road surface; The road surface gradient of the target road surface is provided to a torque distribution system of the target vehicle, so that the torque distribution system of the target vehicle adjusts the engine torque of the target vehicle according to the road surface gradient of the target road surface.

3. The method according to claim 1, characterized in that: Using the target neural network, based on the information for prediction, predicting a prediction result related to the target road surface, including: generating a hidden state vector of each of the plurality of first vehicle parameters, and extracting a road surface image feature vector of the road surface image; Generate a target hidden state vector according to the hidden state vector of each first vehicle parameter in the plurality of first vehicle parameters and the hidden state vector of each second vehicle parameter in the plurality of second vehicle parameters of the target vehicle at a previous moment before the target moment; A feature for prediction is generated according to the target hidden state vector and the road surface image feature vector, and the prediction result is predicted according to the feature for prediction.

4. The method according to claim 3, characterized in that: Generating a target hidden state vector according to the hidden state vector of each first vehicle parameter in the plurality of first vehicle parameters and the hidden state vector of each second vehicle parameter in the plurality of second vehicle parameters of the target vehicle at a previous moment before the target moment comprises: Determining an attention parameter of each first vehicle parameter according to a hidden state vector of each first vehicle parameter among the plurality of first vehicle parameters; Generate an attention hidden state vector of each first vehicle parameter according to the hidden state vector of each first vehicle parameter in the plurality of first vehicle parameters and the attention parameter of each first vehicle parameter; A target hidden state vector is generated based on the attention hidden state vector of each first vehicle parameter and the hidden state vector of each second vehicle parameter.

5. The method according to claim 4, characterized in that: Determining, according to the hidden state vector of each first vehicle parameter among the plurality of first vehicle parameters, the attention parameter of each first vehicle parameter comprises: Determining a first attention score for each of the plurality of first vehicle parameters according to a hidden state vector of the first vehicle parameter; For each of the first vehicle parameters, determining a second attention score of the first vehicle parameter according to a hidden state vector of the first vehicle parameter and a hidden state vector of a second vehicle parameter corresponding to the first vehicle parameter, wherein a type of the first vehicle parameter is the same as a type of the second vehicle parameter corresponding to the first vehicle parameter; An attention parameter of each first vehicle parameter is determined according to the first attention score of each first vehicle parameter and the second attention score of each first vehicle parameter.

6. The method according to claim 5, characterized in that: Generating a target hidden state vector according to the attention hidden state vector of each first vehicle parameter and the hidden state vector of each second vehicle parameter comprises: Generate an attention hidden state vector of each second vehicle parameter according to the hidden state vector of each second vehicle parameter and the attention parameter of each second vehicle parameter; A target hidden state vector is generated according to the attention hidden state vector of each first vehicle parameter and the attention hidden state vector of each second vehicle parameter.

7. The method according to claim 6, characterized in that: Generating a target hidden state vector according to the attention hidden state vector of each first vehicle parameter and the attention hidden state vector of each second vehicle parameter comprises: generating a first vehicle parameter combination vector including an attention hidden state vector of each of the first vehicle parameters, and generating a second vehicle parameter combination vector including an attention hidden state vector of each of the second vehicle parameters; The first vehicle parameter combination vector is added to the second vehicle parameter combination vector to obtain a target hidden state vector.

8. The method according to any one of claims 3 to 7, characterized in that: Generating features for prediction according to the target hidden state vector and the road surface image feature vector includes: Multiplying the target hidden state vector by the vehicle parameter feature weight to obtain a first weighted vector, and multiplying the road surface image feature vector by the road surface image feature weight to obtain a second weighted vector; A feature for prediction is generated according to the first weighting vector and the second weighting vector.

9. The method according to claim 8, characterized in that: Generating features for prediction according to the first weighted vector and the second weighted vector includes: The first weighted vector and the second weighted vector are added together to obtain features for prediction.

10. The method according to claim 8, characterized in that: Before obtaining the information for prediction corresponding to the target time, the method further includes: The vehicle parameter feature weights and the road surface image feature weights are determined according to the accuracy of the vehicle parameter feature extraction network in the target neural network on the verification set of the target neural network and the accuracy of the road surface image feature extraction network in the target neural network on the verification set.

11. An information providing device, characterized in that: The device comprises: an acquisition unit, configured to acquire information for prediction corresponding to a target time, the information for prediction comprising: a plurality of first vehicle parameters of the target vehicle at the target time, and a road surface image related to the target time, the road surface image comprising a road surface object representing the target road surface; A prediction unit, used to predict a prediction result related to the target road surface by using a target neural network according to the information for prediction, wherein the prediction result includes: a road surface adhesion coefficient of the target road surface, a bumpiness of the target road surface, and a road surface slope of the target road surface; A providing unit is used to provide each item in the prediction result to a corresponding system of the target vehicle, so that the corresponding system can make corresponding adjustments according to the corresponding item in the prediction result.

12. A computer device installed in a vehicle, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the method according to any one of claims 1 to 10 by executing the computer instructions.

13. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the method according to any one of claims 1 to 10.

14. A computer program product, characterized in that The method comprises computer instructions for causing a computer to execute the method according to any one of claims 1 to 10.

Images