Configuration modeling system and method for parameters of vehicle-mounted head-up display under big data of Internet of Vehicles

Abstract

The invention discloses a configuration modeling system for parameters of a vehicle-mounted head-up display under big data of Internet of Vehicles, and the system comprises a data preprocessing module which carries out the preprocessing of vehicle operation state data, vehicle-mounted equipment log data, environment information data and key arrangement parameter data of the vehicle-mounted head-up display, and obtains the preprocessed data; the multiple linear regression model training module divides a training set from the preprocessed data, and randomly sets initial model parameters of a multiple linear regression model; iteratively updating the initial model parameters of the multiple linear regression model by using the training set through a gradient descent algorithm to obtain the multiple linear regression model after model parameter optimization so as to obtain a trained multiple linear regression model; and predicting arrangement parameters of the vehicle-mounted head-up display in different driving scenes of the vehicle through the trained multiple linear regression model. The visual interference risk in the driving process is reduced, and the driving safety and the man-machine interaction comfort are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of Internet of Vehicles and big data analysis, and particularly relates to a system and method for configuring and modeling parameters of a head-up display in Internet of Vehicles big data. BACKGROUND

[0002] At present, fixed parameter setting is generally used in the arrangement of the head-up display, which is difficult to adapt to complex driving scenarios. Specifically, the position parameters and display parameters of the head-up display cannot be adjusted according to the dynamic changes of the real-time running state of the vehicle, external environmental factors and driving behavior characteristics, which leads to the fact that the driver is difficult to obtain clear and appropriate display information in some complex driving scenarios, thereby affecting the driving safety and human-computer interaction comfort. In addition, the rich information sources of Internet of Vehicles big data cannot be fully utilized to realize intelligent and accurate configuration of the parameters of the head-up display, so that the overall performance of the head-up display system cannot be fully exerted. SUMMARY

[0003] The present application aims to provide a system and method for configuring and modeling parameters of a head-up display in Internet of Vehicles big data, which reduces the risk of visual interference during driving, and improves the driving safety and human-computer interaction comfort.

[0004] To achieve this object, the present application provides a system for configuring and modeling parameters of a head-up display in Internet of Vehicles big data, which comprises: A data preprocessing module is configured to preprocess vehicle running state data, vehicle-mounted device log data, environmental information data and key arrangement parameter data of the head-up display, so as to obtain preprocessed data. A multivariate linear regression model training module is configured to divide a training set from the preprocessed data, randomly set initial model parameters of the multivariate linear regression model, and use the training set to iteratively update the initial model parameters of the multivariate linear regression model by using a gradient descent algorithm to obtain a model parameter optimized multivariate linear regression model, thereby obtaining a trained multivariate linear regression model. The trained multivariate linear regression model is used to predict the arrangement parameters of the head-up display in different driving scenarios of the vehicle.

[0005] Preferably, a model experiment and verification module is configured to design experiments for different driving scenarios of the vehicle, and to verify the effectiveness of the trained multivariate linear regression model by collecting subjective evaluation data and objective index data of the test objects.

[0006] Preferably, the vehicle running state data, vehicle-mounted device log data, environmental information data and key arrangement parameter data of the head-up display are one-to-one corresponding through the same timestamp. The vehicle operating state data includes real-time vehicle speed, acceleration, braking state, fuel consumption, rotation speed, load state, light intensity, light direction, vehicle body pitch angle and vehicle body roll angle collected by the vehicle-mounted sensor; the vehicle-mounted device log data includes user operation record data and navigation system operating state data collected by the vehicle-mounted device; the environmental information data includes weather condition data, traffic condition data and vehicle maintenance record data collected by the Internet; and the vehicle-mounted head-up display key arrangement parameter data includes horizontal offset, vertical offset, pitch angle, roll angle, field of view angle, resolution and refresh rate of the head-up display. The vehicle operating state data, the vehicle-mounted device log data, the environmental information data and the vehicle-mounted head-up display key arrangement parameter data under the same timestamp form an original data, and all original data in a set time period is collected as an original data set.

[0007] Preferably, the vehicle operating state data, the vehicle-mounted device log data, the environmental information data and the vehicle-mounted head-up display key arrangement parameter data are preprocessed to obtain preprocessed data, and the specific process is as follows: The original data set is subjected to data cleaning, and error data caused by device failure or human operation error is deleted, and the missing values in the original data set are filled by using the Lagrange interpolation method to obtain data cleaned data. The data cleaned data is subjected to data screening, and data with a correlation coefficient with the vehicle-mounted head-up display key arrangement parameter data lower than a set threshold is removed by setting a screening condition to obtain data screened data. The data screened data is subjected to data conversion, and features in the vehicle operating state data, the vehicle-mounted device log data and the environmental information data with a correlation degree greater than a set correlation degree threshold with the vehicle-mounted head-up display key arrangement parameter data are extracted, and the features corresponding data and the vehicle-mounted head-up display key arrangement parameter data are normalized to obtain preprocessed data.

[0008] Preferably, the training set and the validation set are divided from the preprocessed data, the initial model parameters of the multiple linear regression model are randomly set, the initial model parameters of the multiple linear regression model are iteratively updated by using the training set through the gradient descent algorithm to obtain the optimized model parameters of the multiple linear regression model, and the multiple linear regression model with the optimized model parameters is verified through the validation set to obtain the trained multiple linear regression model, and the specific process is as follows: The training set and the validation set are divided from the preprocessed data, the initial model parameters of the multiple linear regression model are randomly set, and the multiple linear regression model with the initial model parameters is constructed: wherein, The visual distraction score predicted by the multiple linear regression model; ~ A set of independent variables consisting of the features in the training set and the layout parameters of the vehicle head-up display; is the intercept term; ~ is the set of regression coefficients; is the preset error term; The training set is used to iteratively update the initial model parameters of the multivariate linear regression model through the gradient descent algorithm, and a set number of samples in the training set are used in each iteration to calculate the descent gradient of the model parameters to obtain the updated model parameters, where a set number of samples in the training set are used to calculate the initial model parameters. The descent gradient is used to obtain the updated model parameters The calculation formula is: in, are the initial model parameters, including the initial intercept term and the initial regression coefficients ~ ; is the learning rate; is the loss function; are the updated model parameters, including the updated intercept term and the updated regression coefficient; is the loss function For the initial model parameters gradient; The loss function By mean square error Calculated, where is the number of samples in the training set used in each iteration; For the The true visual distraction score of samples; For the The predicted visual distraction score of samples; By continuously iteratively updating the model parameters of the multiple linear regression model, the mean square error between the predicted value and the true value of the multiple linear regression model on the samples in the training set is Gradually decreases, when the mean square error of this iteration The mean square error compared to the previous iteration When the change between the values ​​is less than the preset threshold, the loss function Convergence obtains optimized model parameters, which include the optimized intercept term and the optimized regression dataset , the optimized multiple linear regression model after the model parameter of this iteration is verified by the validation set, and a trained multiple linear regression model is obtained, and the expression is: wherein, is the optimized intercept term; is the optimized regression data set; is the visual disturbance score predicted by the trained multiple linear regression model.

[0009] Preferably, the specific process of predicting the arrangement parameters of the vehicle head-up display in different driving scenarios of the vehicle by the trained multiple linear regression model is as follows: The arrangement parameters of the vehicle head-up display in different driving scenarios of the vehicle are predicted by the trained multiple linear regression model, and the expression of the trained multiple linear regression model is as follows: In a given driving scenario, ~ The features extracted in the driving scenario are fixed parameters, ~ The arrangement parameters of the vehicle head-up display in the driving scenario are adjustable parameters, and the arrangement parameters of the vehicle head-up display in the driving scenario are changed ~ The arrangement parameters of the vehicle head-up display in the driving scenario are changed to minimize ~ The corresponding arrangement parameters of the vehicle head-up display in the driving scenario are the best arrangement parameters of the vehicle head-up display in the driving scenario predicted by the trained multiple linear regression model.

[0010] Preferably, the multiple linear regression model is optimized during the training process, and the correlation between each independent variable and the visual disturbance score predicted by the multiple linear regression model is analyzed to calculate the Pearson correlation coefficient , and the calculation formula is as follows: If the Pearson correlation coefficient is lower than the set correlation coefficient threshold, the independent variable is an irrelevant feature, which is removed from the multiple linear regression model to reduce the feature dimension of the multiple linear regression model.

[0011] Preferably, the specific process of designing experiments for different driving scenarios of the vehicle to verify the effectiveness of the trained multiple linear regression model by collecting subjective evaluation data and objective index data of the test objects is as follows: The different driving scenes of the vehicle include highway driving, urban road driving, tunnel driving and night driving; the experimental subjects are drivers of different ages and driving ages driving different vehicle types; the subjective evaluation data includes visual interference degree scores and comfort degrees; and the objective index data includes emergency braking times, lane deviation times and eye tracking data, wherein the eye tracking data includes a proportion of time that the driver gazes at the road and a length of time that the driver gazes at the head-up display. The experimental subjects are divided into an experimental group and a control group, wherein the experimental group adjusts the parameters of the vehicle head-up display through the trained multiple linear regression model, the vehicle head-up display of the control group adopts fixed parameters, the subjective evaluation data and the objective index data of the experimental group and the control group are collected, and if the visual interference degree score of the experimental group is lower than that of the control group, the comfort degree of the experimental group is higher than that of the control group, the emergency braking times and the lane deviation times of the experimental group are less than those of the control group, and the proportion of time that the driver gazes at the road and the length of time that the driver gazes at the head-up display of the experimental group are longer than those of the control group, it is proved that the trained multiple linear regression model can effectively predict and adapt the layout parameters of the vehicle head-up display according to different driving scenes of the vehicle.

[0012] A configuration modeling method of a vehicle head-up display parameter under Internet of Vehicles big data, comprising the following steps: Pretreat vehicle operating state data, vehicle equipment log data, environmental information data and key layout parameter data of the vehicle head-up display to obtain pretreated data; Divide the pretreated data into a training set and a validation set, randomly set initial model parameters of a multiple linear regression model, use the training set to iteratively update the initial model parameters of the multiple linear regression model by a gradient descent algorithm to obtain optimized model parameters of the multiple linear regression model, verify the multiple linear regression model with optimized model parameters by the validation set to obtain a trained multiple linear regression model, and predict layout parameters of the vehicle head-up display under different driving scenes of the vehicle by the trained multiple linear regression model.

[0013] A computer program product comprising a computer program, which, when executed by a processor, implements the steps of the above method.

[0014] The beneficial effects of the present application are: The present invention can dynamically adjust the layout parameters of the vehicle's head-up display by real-time sensing the vehicle's operating status, external environment and driver behavior, ensuring that the driver can quickly and clearly obtain key driving information in various driving scenarios, reducing distraction caused by line of sight, blurred information or display discomfort, and thus reducing the probability of traffic accidents; the present invention can provide personalized vehicle-mounted head-up display solutions based on individual differences of drivers and the needs of different driving scenarios, thereby improving user comfort and satisfaction; the present invention can fully tap the value of vehicle network big data, establish a dynamic correlation model between data and vehicle-mounted head-up display parameters, realize precise configuration of head-up display layout parameters, change the limitations of traditional fixed parameter settings, enable the vehicle-mounted head-up display system to perform optimally in various scenarios, and improve the overall efficiency of the vehicle-mounted head-up display system; the present invention can promote the deep integration of vehicle network big data and vehicle-mounted head-up display equipment, provide innovative ideas and technical support for the optimization of the human-computer interaction system of smart cars, and help promote the development of the automotive industry towards intelligence and personalization. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a structural schematic diagram of the present invention; Figure 2 Flowchart of the present invention. DETAILED DESCRIPTION

[0016] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments: Example 1 A configuration modeling system for vehicle head-up display parameters under Internet of Vehicles big data, such as Figure 1 As shown, it includes: The data preprocessing module is used to preprocess vehicle operating status data, on-board equipment log data, environmental information data, and key layout parameter data of the on-board head-up display to obtain preprocessed data. This design improves the quality and consistency of the input data of the multivariate linear regression model through data preprocessing. The multivariate linear regression model training module is configured to divide a training set from the preprocessed data, randomly set initial model parameters of the multivariate linear regression model, and use the training set to iteratively update the initial model parameters of the multivariate linear regression model by a gradient descent algorithm to obtain model parameters of the multivariate linear regression model after optimization, thereby obtaining a trained multivariate linear regression model. The trained multivariate linear regression model is used to predict the arrangement parameters of the vehicle head-up display in different driving scenarios. The design can effectively reduce the visual interference score by predicting the optimal arrangement parameters of the vehicle head-up display in different driving scenarios. The model parameters of the multivariate linear regression model are optimized by the gradient descent algorithm, and the generalization ability of the multivariate linear regression model is improved by randomly dividing the training set to suppress overfitting of the multivariate linear regression model. The optimal arrangement parameters of the vehicle head-up display can be obtained by inputting the features of the current driving scenario into the trained multivariate linear regression model.

[0017] In the above technical solution, the model experiment and verification module is configured to design experiments for different driving scenarios of the vehicle, and verify the effectiveness of the trained multivariate linear regression model by collecting subjective evaluation data and objective index data of the test subjects. The design collects subjective evaluation data and objective index data of the test subjects, directly reflects the discomfort of the driver through the visual interference score, verifies the effectiveness of the trained multivariate linear regression model, and ensures the reliability of the optimal arrangement parameters of the vehicle head-up display.

[0018] In the above technical solution, the vehicle operating state data, the vehicle-mounted device log data, the environmental information data, and the key arrangement parameters of the vehicle head-up display are one-to-one corresponding through the same timestamp. The vehicle operating state data includes real-time vehicle speed, acceleration, braking state, fuel consumption, speed, load state, light intensity, light direction, body pitch angle, and body roll angle collected by vehicle-mounted sensors (including vehicle speed sensor, fuel consumption sensor, engine speed sensor, and light sensor). The vehicle-mounted device log data includes user operation record data and navigation system operating state data collected by vehicle-mounted devices (including navigation and entertainment systems). The environmental information data includes weather condition data, traffic condition data, and vehicle maintenance record data collected from the Internet. The key arrangement parameters of the vehicle head-up display include horizontal offset, vertical offset, pitch angle, roll angle, field of view angle, resolution, and refresh rate of the head-up display. The vehicle running state data, the vehicle-mounted device log data, the environment information data and the vehicle-mounted head-up display key arrangement parameter data under the same timestamp are bound to ensure the spatiotemporal consistency of the data and avoid feature correlation distortion.

[0019] In the technical solution, the vehicle running state data, the vehicle-mounted device log data, the environment information data and the vehicle-mounted head-up display key arrangement parameter data are preprocessed to obtain preprocessed data, and the specific process is as follows: The original data set is subjected to data cleaning, and error data caused by device failure or human operation error is deleted. The missing values in the original data set are filled by using the Lagrange interpolation method to obtain data after cleaning. The data after cleaning is subjected to data screening, and data with a correlation coefficient lower than a set threshold (the set threshold can be set to 0.4) with the vehicle-mounted head-up display key arrangement parameter data is removed to obtain data after screening. The data after screening is subjected to data conversion, and features (such as vehicle speed, road condition complexity (calculated according to road type and traffic flow), weather condition (numerically coded by sunny, rainy and snowy) and driver fatigue degree (calculated by heart rate variability and blink frequency)) with a correlation degree greater than a set correlation degree threshold (the set correlation degree threshold is usually set to 0.3 to 0.7, and can be 0.4) with the vehicle-mounted head-up display key arrangement parameter data are extracted from the vehicle running state data, the vehicle-mounted device log data and the environment information data, and the data corresponding to the features and the vehicle-mounted head-up display key arrangement parameter data are normalized to obtain preprocessed data. The above design retains the continuity of the data by deleting error data and filling missing values by interpolation, removes low correlation degree features according to the set correlation degree threshold, and eliminates dimension differences by normalization, thereby generating high-purity training data and improving the convergence speed and accuracy of the multiple linear regression model.

[0020] In the technical solution, a training set and a validation set are divided from the preprocessed data, initial model parameters of a multiple linear regression model are randomly set, the initial model parameters of the multiple linear regression model are iteratively updated by using the training set through a gradient descent algorithm to obtain optimized model parameters of the multiple linear regression model, the multiple linear regression model with the optimized model parameters is verified by using the validation set, and a trained multiple linear regression model is obtained, and the specific process is as follows: The training set and validation set are divided from the preprocessed data, the initial model parameters of the multiple linear regression model are randomly set, and the multiple linear regression model with the initial model parameters is constructed: in, is the visual distraction score predicted by the multiple linear regression model (visual distraction score is 1 to 10 points); ~ The independent variable set is composed of the features described in the training set (the features described in the training set are the features extracted from the vehicle operation status data, the vehicle equipment log data, and the environmental information data, the correlation of which with the key layout parameter data of the vehicle head-up display is greater than a set correlation threshold) and the layout parameters of the vehicle head-up display (including the horizontal offset, vertical offset, pitch angle, roll angle, field of view, resolution, and refresh rate of the head-up display); is the intercept term (the intercept term When all the independent variables in the independent variable set are 0, The predicted value of , in gradient descent optimization, the intercept term With the regression coefficient set ~ Synchronous updates); ~ is the set of regression coefficients; is the preset error term (the preset error term The error term follows a normal distribution and is used to represent random fluctuations that cannot be explained by the multiple linear regression model, including unmeasured variables, sensor measurement errors, and instantaneous interference from road emergencies); The training set is used to iteratively update the initial model parameters of the multivariate linear regression model through the gradient descent algorithm, and a set number of samples in the training set are used in each iteration to calculate the descent gradient of the model parameters to obtain the updated model parameters, where a set number of samples in the training set are used to calculate the initial model parameters. The descent gradient is used to obtain the updated model parameters The calculation formula is: in, are the initial model parameters, including the initial intercept term and the initial regression coefficients ~ ; is the learning rate; is the loss function; are the updated model parameters, including the updated intercept term and the updated regression coefficient; is the loss function For the initial model parameters gradient; The loss function By mean square error Calculated, where is the number of samples in the training set used in each iteration; For the The real visual distraction score of samples (the real visual distraction score is the subjective ergonomic data obtained through driving experiments designed according to different driving scenarios); For the The predicted visual distraction score of samples; By continuously iteratively updating the model parameters of the multiple linear regression model, the mean square error between the predicted value and the true value of the multiple linear regression model on the samples in the training set is Gradually decreases, when the mean square error of this iteration The mean square error compared to the previous iteration When the change between the values ​​is less than the preset threshold, the loss function Convergence to obtain optimized model parameters (loss function The corresponding model parameters at convergence is the optimized model parameter), the optimized model parameters include the optimized intercept term and the optimized regression dataset , the multiple linear regression model after the model parameters are optimized after this iteration is verified through the validation set, and the trained multiple linear regression model is obtained. The expression is: in, is the optimized intercept term; is the optimized regression dataset; The visual disturbance score predicted by the trained multiple linear regression model. The above design updates and optimizes the model parameters of the multiple linear regression model through gradient descent. The multiple linear regression model with optimized model parameters after this iteration is verified through the validation set, which can prevent overfitting and improve the prediction accuracy of the trained multiple linear regression model.

[0021] In the above technical solution, the specific process of predicting the layout parameters of the vehicle head-up display under different driving scenarios by the trained multivariate linear regression model is as follows: The trained multiple linear regression model is used to predict the layout parameters of the vehicle head-up display under different driving scenarios. The expression of the trained multiple linear regression model is as follows: In a given driving scenario, The feature extracted in the driving scenario is a fixed parameter, The vehicle head-up display arrangement parameter is an adjustable parameter, and the vehicle head-up display arrangement parameter is changed The vehicle head-up display arrangement parameter is an adjustable parameter, and the vehicle head-up display arrangement parameter is changed The minimum value, at this time The corresponding vehicle head-up display arrangement parameter is the optimal vehicle head-up display arrangement parameter predicted by the trained multiple linear regression model in the driving scenario; the above design inputs the fixed parameter in a given driving scenario into the trained multiple linear regression model, adjusts the vehicle head-up display arrangement parameter, and minimizes the visual interference score to obtain the optimal vehicle head-up display arrangement parameter.

[0022] In the above technical solution, the multiple linear regression model is optimized during the training process, and each independent variable The visual interference score predicted by the multiple linear regression model Correlation analysis is performed to calculate the Pearson correlation coefficient The calculation formula is as follows: If Lower than the set correlation coefficient threshold, the independent variable Is an irrelevant feature, which is removed from the multiple linear regression model to reduce the feature dimension of the multiple linear regression model; the above design can reduce the complexity of the multiple linear regression model and improve the real-time performance by reducing the feature dimension.

[0023] In the above technical solution, the specific process of verifying the effectiveness of the trained multiple linear regression model by designing experiments for different driving scenarios of the vehicle and collecting subjective evaluation data and objective index data of the test object is as follows: The different driving scenarios of the vehicle include highway driving, urban road driving, tunnel driving, and night driving; the test object is a driver of different ages and driving ages driving different vehicle types (which can include SUV, sedan, and new energy vehicle); the subjective evaluation data includes visual interference score and comfort level; the objective index data includes emergency braking times, lane deviation times, and eye tracking data, wherein the eye tracking data includes the proportion of time that the driver gazes at the road and the duration that the driver gazes at the head-up display; ​​​​The experimental subjects are divided into an experimental group (the vehicle-mounted head-up display uses a traditional fixed parameter setting) and a control group (the vehicle-mounted head-up display arrangement parameters are dynamically adjusted based on a trained multiple linear regression model). The experimental group adjusts the vehicle-mounted head-up display parameters through the trained multiple linear regression model, and the control group uses fixed parameters. The subjective evaluation data and objective index data of the experimental group and the control group are collected. If the visual interference degree score of the experimental group is lower than that of the control group, the comfort level of the experimental group is higher than that of the control group, the number of emergency braking times (AEB trigger frequency) and the number of lane deviation times (LDW trigger frequency) of the experimental group are less than those of the control group, and the proportion of driver's gaze on the road and the length of time when the driver gazes at the head-up display of the experimental group are longer than those of the control group, it is proved that the trained multiple linear regression model can effectively predict the adaptive vehicle-mounted head-up display arrangement parameters according to different driving scenarios of the vehicle. The above design verifies the best vehicle-mounted head-up display arrangement parameters predicted by the trained multiple linear regression model according to different driving scenarios of the vehicle by designing the experimental group and the control group, and proves the effectiveness and practicality of the trained multiple linear regression model.

[0024] Embodiment 2 A configuration modeling method of vehicle-mounted head-up display parameters under Internet of Vehicles big data, as shown in Figure 2 The method comprises the following steps: Pretreat vehicle operating state data, vehicle-mounted device log data, environmental information data, and key arrangement parameter data of the vehicle-mounted head-up display to obtain pretreated data. Divide the pretreated data into a training set and a validation set, randomly set initial model parameters of a multiple linear regression model, use the training set to iteratively update the initial model parameters of the multiple linear regression model through a gradient descent algorithm to obtain optimized model parameters of the multiple linear regression model, verify the multiple linear regression model with optimized model parameters through the validation set, obtain a trained multiple linear regression model, and predict vehicle-mounted head-up display arrangement parameters in different driving scenarios of the vehicle through the trained multiple linear regression model.

[0025] Embodiment 3 A computer program product comprises a computer program which, when executed by a processor, implements the steps of the method described in embodiment 2.

[0026] Those skilled in the art will appreciate that embodiments of the present application can be readily used as software, hardware, or a combination of software and hardware. In a software embodiment, the methods can be tangibly embodied in a machine-readable storage medium having stored thereon instructions that can be used to program a computer to perform any of the methods. The software implementation can be initialized by loading and executing a set of instructions arranged to perform one of the methods into the computer's memory. Alternatively, hard-wired circuitry can be used in place of, or in combination with, software instructions. Thus, the

[0027] The present application is described in reference to the drawings using a flowchart and / or a block diagram of the method, apparatus (system) and computer program product according to embodiments of the application. It will be understood that each block of the flowchart and / or block diagram, and combinations of blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart and / or block diagram block or blocks. Figure 1 one or more functions specified in one or more of the flowchart or block diagram block or blocks. Figure 1 one or more functions specified in one or more of the flowchart or block diagram block or blocks.

[0028] These computer program instructions can also be stored in a computer- readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instructions which implement the flowchart and / or block diagram block or blocks. Figure 1 one or more functions specified in one or more of the flowchart or block diagram block or blocks. Figure 1 one or more functions specified in one or more of the flowchart or block diagram block or blocks.

[0029] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the flowchart and / or block diagram block or blocks. Figure 1 one or more functions specified in one or more of the flowchart or block diagram block or blocks. Figure 1 one or more functions specified in one or more of the flowchart or block diagram block or blocks.

[0030] Finally, it should be noted that the above-described embodiments are merely intended for describing the technical solutions of the present application, but not to limit the scope of protection of the present application. Although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can make various changes and modifications to the specific embodiments, or equivalents to the technical solutions of the present application after they have read the specification. However, these changes and modifications, or equivalents, do not depart from the scope of protection of the present application.

[0031] The contents of the specification not described in detail are the prior art known to those skilled in the art.

Claims

1. A configuration modeling system for vehicle head-up display parameters under Internet of Vehicles big data, characterized by: It includes: The data preprocessing module is used to preprocess the vehicle operation status data, vehicle equipment log data, environmental information data and key layout parameter data of the vehicle head-up display to obtain preprocessed data; The multivariate linear regression model training module is used to divide the training set from the preprocessed data, randomly set the initial model parameters of the multivariate linear regression model, and use the training set to iteratively update the initial model parameters of the multivariate linear regression model through the gradient descent algorithm to obtain the multivariate linear regression model with optimized model parameters, thereby obtaining the trained multivariate linear regression model. The trained multivariate linear regression model is used to predict the layout parameters of the vehicle's head-up display under different driving scenarios.

2. The system for configuring and modeling vehicle head-up display parameters based on IoV big data according to claim 1 is characterized in that: It also includes: The model experiment and verification module is used to design experiments for different vehicle driving scenarios and verify the effectiveness of the trained multivariate linear regression model by collecting subjective evaluation data and objective indicator data from the test subjects.

3. The system for configuring and modeling vehicle head-up display parameters based on IoV big data according to claim 1, characterized in that: The vehicle operation status data, vehicle equipment log data, environmental information data and vehicle head-up display key layout parameter data are one-to-one correspondence through the same timestamp; Vehicle operating status data includes real-time vehicle speed, acceleration, braking status, fuel consumption, speed, load status, light intensity, light direction, vehicle pitch angle, and vehicle roll angle collected by on-board sensors; on-board device log data includes user operation record data and navigation system operating status data collected by on-board devices; environmental information data includes weather condition data, traffic condition data, and vehicle maintenance record data collected from the Internet; key layout parameter data of the on-board head-up display includes the horizontal offset, vertical offset, pitch angle, roll angle, field of view, resolution, and refresh rate of the head-up display; A piece of raw data is composed of vehicle operation status data, vehicle equipment log data, environmental information data and key layout parameter data of the vehicle head-up display at the same timestamp. All raw data within a set time period are collected as the raw data set.

4. The system for configuring and modeling vehicle head-up display parameters based on IoV big data according to claim 3, characterized in that: The specific process of preprocessing vehicle operation status data, vehicle equipment log data, environmental information data, and key layout parameter data of the vehicle head-up display to obtain the preprocessed data is as follows: Performing data cleaning on the original data set, deleting erroneous data caused by equipment failure or human error, and filling missing values ​​in the original data set using Lagrange interpolation to obtain cleaned data; Performing data screening on the cleaned data, by setting screening conditions to remove data with a correlation coefficient lower than a set threshold with respect to key layout parameter data of the vehicle head-up display, thereby obtaining the filtered data; Data conversion is performed on the data after data screening, and features whose correlation with the key layout parameter data of the vehicle head-up display is greater than a set correlation threshold are extracted from the vehicle operation status data, the vehicle-mounted equipment log data and the environmental information data, and the data corresponding to the features and the key layout parameter data of the vehicle head-up display are normalized to obtain preprocessed data.

5. The system for configuring and modeling vehicle head-up display parameters based on IoV big data according to claim 4, characterized in that: The training set and validation set are divided from the preprocessed data, and the initial model parameters of the multiple linear regression model are randomly set. The initial model parameters of the multiple linear regression model are iteratively updated using the training set through the gradient descent algorithm to obtain the optimized model parameters of the multiple linear regression model. The multiple linear regression model after the model parameter optimization is verified by the validation set to obtain the trained multiple linear regression model. The specific process is as follows: The training set and validation set are divided from the preprocessed data, the initial model parameters of the multiple linear regression model are randomly set, and the multiple linear regression model with the initial model parameters is constructed: in, The visual distraction score predicted by the multiple linear regression model; ~ A set of independent variables consisting of the features in the training set and the layout parameters of the vehicle head-up display; is the intercept term; ~ is the set of regression coefficients; is the preset error term; The training set is used to iteratively update the initial model parameters of the multivariate linear regression model through the gradient descent algorithm, and a set number of samples in the training set are used in each iteration to calculate the descent gradient of the model parameters to obtain the updated model parameters, where a set number of samples in the training set are used to calculate the initial model parameters. The descent gradient is used to obtain the updated model parameters The calculation formula is: in, are the initial model parameters, including the initial intercept term and the initial regression coefficients ~ ; is the learning rate; is the loss function; are the updated model parameters, including the updated intercept term and the updated regression coefficient; is the loss function For the initial model parameters gradient; The loss function By mean square error Calculated, where is the number of samples in the training set used in each iteration; For the The true visual distraction score of samples; For the The predicted visual distraction score of samples; By continuously iteratively updating the model parameters of the multiple linear regression model, the mean square error between the predicted value and the true value of the multiple linear regression model on the samples in the training set is Gradually decreases, when the mean square error of this iteration The mean square error compared to the previous iteration When the change between the two values ​​is less than the preset threshold, the loss function Convergence obtains optimized model parameters, which include the optimized intercept term and the optimized regression dataset , the multiple linear regression model after the model parameters are optimized after this iteration is verified through the validation set, and the trained multiple linear regression model is obtained. The expression is: in, is the optimized intercept term; is the optimized regression dataset; Visual distraction scores predicted by the trained multiple linear regression model.

6. The system for configuring and modeling vehicle head-up display parameters based on IoV big data according to claim 5, characterized in that: The specific process of predicting the layout parameters of the vehicle head-up display under different driving scenarios using the trained multivariate linear regression model is as follows: The trained multiple linear regression model is used to predict the layout parameters of the vehicle head-up display under different driving scenarios. The expression of the trained multiple linear regression model is as follows: In a given driving scenario, ~ The features extracted in this driving scenario are fixed parameters. ~ The layout parameters of the vehicle head-up display are adjustable parameters. ~ The layout parameters of the vehicle head-up display make Minimize, then ~ The corresponding vehicle head-up display layout parameters are the optimal vehicle head-up display layout parameters predicted by the trained multiple linear regression model under this driving scenario.

7. The system for configuring and modeling vehicle head-up display parameters based on IoV big data according to claim 5, characterized in that: The multivariate linear regression model is optimized during the training process. Visual distraction scores predicted by the multiple linear regression model Perform correlation analysis and calculate Pearson correlation coefficient , the calculation formula is as follows: like If the correlation coefficient is lower than the set threshold, the independent variable As irrelevant features, they are removed from the multiple linear regression model to reduce the feature dimension of the multiple linear regression model.

8. The system for configuring and modeling vehicle head-up display parameters based on IoV big data according to claims 2 and 6, characterized in that: The specific process of designing experiments for different vehicle driving scenarios and verifying the effectiveness of the trained multivariate linear regression model by collecting subjective evaluation data and objective indicator data from the test subjects is as follows: The different driving scenarios for the vehicles included highway driving, urban road driving, tunnel driving, and nighttime driving. The experimental subjects were drivers of different ages and driving experience driving different types of vehicles. The subjective evaluation data included visual distraction scores and comfort levels. The objective indicator data included the number of emergency braking attempts, the number of lane departure attempts, and eye-tracking data, where the eye-tracking data included the percentage of time the driver looked at the road and the length of time the driver looked at the head-up display. The experimental subjects are divided into an experimental group and a control group, wherein the experimental group adjusts the parameters of the vehicle head-up display using the trained multivariate linear regression model, and the vehicle head-up display of the control group uses fixed parameters. Subjective evaluation data and objective indicator data of the experimental group and the control group are collected. If the visual interference score of the experimental group is lower than that of the control group, the comfort of the experimental group is higher than that of the control group, the number of emergency braking and the number of lane departures of the experimental group are less than those of the control group, and the proportion of time the driver looks at the road and the length of time the driver looks at the head-up display in the experimental group are longer than those in the control group, then it is proved that the trained multivariate linear regression model can effectively predict the adaptive vehicle head-up display layout parameters according to different vehicle driving scenarios.

9. A configuration modeling method for vehicle head-up display parameters under Internet of Vehicles big data, characterized in that: It includes the following steps: Preprocessing vehicle operation status data, on-board equipment log data, environmental information data, and key layout parameter data of the on-board head-up display to obtain preprocessed data; A training set and a validation set are divided from the preprocessed data, and the initial model parameters of the multiple linear regression model are randomly set. The initial model parameters of the multiple linear regression model are iteratively updated using the training set through the gradient descent algorithm to obtain the optimized model parameters of the multiple linear regression model. The multiple linear regression model with optimized model parameters is verified using the validation set to obtain the trained multiple linear regression model. The trained multiple linear regression model is used to predict the layout parameters of the vehicle's head-up display under different driving scenarios.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to claim 9 are implemented.

Images