Vehicle lamp control model training method and vehicle lamp control method

Through the training method of the headlight control model, the support vector machine model and MEMS micro-galvanometer technology are used to realize intelligent control of the headlight system, solving the problem that the headlight system cannot intelligently adjust the light beam in the existing technology, and improving driving safety Sex and comfort.

CN118940867APending Publication Date: 2024-11-12CHONGQING SELIS PHOENIX INTELLIGENT INNOVATION TECH CO LTD
View PDF 8 Cites 0 Cited by

Patent Information

Application Number
CN202411264479.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-10
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

The existing headlight system requires manual operation or can only simply adjust the beam brightness according to the vehicle speed, and cannot intelligently adjust the beam direction and range according to the actual road conditions and driving conditions, resulting in insufficient field of view at night or on complex roads.

Method used

The headlight control model training method is adopted. By obtaining sample data including vehicle speed, steering wheel angle, vehicle information in front, road information, weather information and navigation information, using the support vector machine model to adjust the deflection angle of the MEMS micro-galvanometer reflector, The light luminous intensity and emission frequency of the light source realize intelligent and diversified headlight control.

Benefits of technology

It realizes automatic adjustment of the illumination range, brightness, color and form of the headlights according to the vehicle's own driving conditions and environment, improves driving comfort and safety, and improves the accuracy and accuracy of the headlight control by updating the model parameters.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118940867A_ABST
    Figure CN118940867A_ABST
Patent Text Reader

Abstract

The invention relates to the field of vehicle lamp control, in particular to a vehicle lamp control model training method and a vehicle lamp control method. The vehicle lamp control model training method comprises the steps that vehicle lamp control sample data are acquired, the sample data comprise driving data and vehicle lamp control information, and the driving data comprise the vehicle speed, the steering wheel angle, front vehicle information, current road information, weather information and navigation information; the vehicle lamp control information comprises a deflection angle of a reflecting lens in the MEMS micro vibrating mirror, light source luminous intensity, light source emission frequency and a light source emission form; in response to a to-be-trained model based on a support vector machine, obtaining prediction data of the sample data; model parameters of the to-be-trained model are updated to reduce loss of the prediction data and the sample data, and a vehicle lamp control model is obtained and used for responding to the driving data and predicting vehicle lamp control information. According to the invention, intelligent, accurate and diversified control of the vehicle lamp can be realized.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of vehicle light control, and in particular to a vehicle light control model training method and a vehicle light control method. Background Art

[0002] With the development of the automobile industry, the traditional fixed beam headlight system can no longer meet the needs of modern driving. Traditional headlights usually have problems such as insufficient beam coverage and inability to automatically adjust when driving at night, which results in drivers not being able to obtain the best vision when driving on winding roads or at high speeds. Although existing headlight systems have certain adjustment functions, most of them rely on manual operation or can only simply adjust the beam brightness according to the vehicle speed, and cannot intelligently adjust the beam direction and range according to actual road conditions and driving conditions.

[0003] In view of this, the present invention is proposed. Summary of the invention

[0004] The purpose of the present invention is to provide a vehicle light control model training method and a vehicle light control method to solve the problem that the vehicle light system in the prior art needs to rely on manual operation or can only simply adjust the light beam brightness according to the vehicle speed.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] In a first aspect, the present invention provides a vehicle light control model training method, comprising:

[0007] Acquire sample data for headlight control, wherein the sample data includes driving data and headlight control information, wherein the driving data includes vehicle speed, steering wheel angle, information about the vehicle ahead, current road information, weather information, and navigation information, and the headlight control information includes a deflection angle of a reflective lens in a MEMS micro-vibration mirror, light source luminous intensity, light source emission frequency, and light source emission form;

[0008] In response to the model to be trained based on the support vector machine, obtaining prediction data of the sample data;

[0009] The model parameters of the model to be trained are updated to reduce the loss of the predicted data and the sample data, and a vehicle light control model is obtained, wherein the vehicle light control model is used to respond to the driving data and predict vehicle light control information.

[0010] As a further preferred technical solution, the step of obtaining the predicted data of the sample data in response to the model to be trained based on a support vector machine includes:

[0011] Inputting the sample feature vector of the sample data into a model to be trained based on a support vector machine;

[0012] Mapping the sample feature vector to a high-dimensional space according to the kernel function in the model to be trained;

[0013] The sample feature vector is processed based on the hyperplane equation, the objective function and the kernel function, and the predicted data is output.

[0014] As a further preferred technical solution, before inputting the sample feature vector of the sample data into the model to be trained based on the support vector machine, it also includes:

[0015] Determining a sample feature set according to the driving data in the sample data;

[0016] The sample feature set is serialized, standardized and normalized in sequence to obtain a sample feature vector of the sample data.

[0017] As a further preferred technical solution, the standardized mathematical formula is: Among them, X′ is the result after standardization, x min is the minimum value of the sample, x max is the maximum value of the sample, and x is the original data;

[0018] The normalized mathematical formula is: X″″=X″*(x max -x min )+x min , where X″″ is the normalized result.

[0019] As a further preferred technical solution, the updating of the model parameters of the model to be trained to reduce the loss of the prediction data and the sample data to obtain the headlight control model includes:

[0020] According to the predicted data and the headlight control information in the sample data, the kernel function is adjusted to reduce the loss of the predicted data and the sample data, thereby obtaining a headlight control model.

[0021] As a further preferred technical solution, adjusting the kernel function according to the predicted data and the headlight control information in the sample data includes:

[0022] Determining an evaluation index of the predicted data according to the predicted data and the vehicle light control information in the sample data; the evaluation index is a harmonic mean of the accuracy and recall of the predicted data;

[0023] The kernel function is adjusted according to the evaluation index.

[0024] In a second aspect, the present invention provides a vehicle light control method, comprising:

[0025] In response to the vehicle lights being turned on, the driving data is input into the vehicle light control model to obtain vehicle light control information; wherein the driving data includes vehicle speed, steering wheel angle, information of the vehicle ahead, current road information, weather information and navigation information, and the vehicle light control model is trained by the vehicle light control model training method according to any one of claims 1 to 6;

[0026] A vehicle light control operation is performed according to the vehicle light control information.

[0027] In a third aspect, the present invention provides a vehicle light control model training device, comprising:

[0028] A sample data acquisition module is used to acquire sample data for headlight control, wherein the sample data includes driving data and headlight control information, wherein the driving data includes vehicle speed, steering wheel angle, information about the vehicle ahead, current road information, weather information, and navigation information, and the headlight control information includes the deflection angle of the reflective lens in the MEMS micro-vibration mirror, the luminous intensity of the light source, the emission frequency of the light source, and the emission form of the light source;

[0029] A sample data prediction module, used for obtaining predicted data of the sample data in response to a model to be trained based on a support vector machine;

[0030] The headlight control model determination module is used to update the model parameters of the model to be trained to reduce the loss of the prediction data and the sample data, and obtain the headlight control model, wherein the headlight control model is used to respond to the driving data and predict the headlight control information.

[0031] In a fourth aspect, the present invention provides an electronic device, comprising:

[0032] at least one processor, and a memory communicatively coupled to at least one of the processors;

[0033] The memory stores instructions that can be executed by at least one of the processors, and the instructions are executed by at least one of the processors so that at least one of the processors can perform the above method.

[0034] In a fifth aspect, the present invention provides a computer-readable storage medium, on which computer instructions are stored, and the computer instructions are used to enable a computer to execute the above method.

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] The driving data in the headlight control model training method provided by the present invention includes vehicle speed, steering wheel angle, information of the vehicle ahead, current road information, weather information and navigation information. The headlight control information includes the deflection angle of the reflective lens in the MEMS micro-vibration mirror, the luminous intensity of the light source, the emission frequency of the light source and the emission form of the light source. The model to be trained is trained based on the driving data and the headlight control information. Compared with the prior art method that can only adjust the light beam brightness according to the vehicle speed, the driving data is more diversified and the headlight control information is richer. Therefore, the illumination range, brightness, color and form (continuous light or flashing light) of the headlight can be automatically and intelligently controlled based on the vehicle's own driving conditions and environmental conditions, thereby improving driving comfort and safety. Moreover, during the training process, the present invention reduces the loss of prediction data and sample data by updating the model parameters, and can obtain a headlight control model with higher accuracy. In addition, the present invention combines the MEMS micro-vibration mirror with the support vector machine to further improve the accuracy of headlight control. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0038] Figure 1 is a flow chart of the vehicle light control model training method provided in Example 1;

[0039] Figure 2 is a structural schematic diagram of a vehicle light control model training device provided in Example 3;

[0040] Figure 3 It is a schematic diagram of the structure of the electronic device provided in Example 5. DETAILED DESCRIPTION

[0041] The following is a description of exemplary embodiments of the present application in conjunction with the accompanying drawings, including various details of the embodiments of the present application to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present application. Similarly, for the sake of clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0042] As mentioned in the background technology, the existing technology has the problem that the headlight system needs to rely on manual operation or can only simply adjust the light beam brightness according to the vehicle speed. In this regard, the present invention first designs a headlight control model training method, which trains a model to be trained based on a support vector machine based on sample data composed of specific driving data and specific headlight control information to obtain a headlight control model. Further, the present invention uses the headlight control model to control the headlight, thereby realizing intelligent, precise and diversified control of the headlight. The present invention is further described in detail below in conjunction with the embodiments.

[0043] Example 1

[0044] Figure 1 It is a flow chart of a vehicle light control model training method provided in this embodiment. The method can be executed by a vehicle light control model training device. The device can be composed of software and / or hardware and is generally integrated in an electronic device. The electronic device can be a vehicle computer or other mobile terminal (such as a computer, tablet computer, etc.). For the convenience of understanding, each step in the control method of this embodiment is executed by the vehicle computer.

[0045] like Figure 1 As shown, this embodiment provides a vehicle light control model training method, comprising the following steps:

[0046] S110. Acquire sample data for headlight control, wherein the sample data includes driving data and headlight control information. The driving data includes vehicle speed, steering wheel angle, information about the vehicle ahead, current road information, weather information, and navigation information. The headlight control information includes a deflection angle of a reflective lens in a MEMS micro-vibration mirror, light intensity of a light source, emission frequency of the light source, and emission form of the light source.

[0047] Among them, the vehicle speed refers to the current speed of the vehicle, which can be obtained by the vehicle speed sensor. The steering wheel angle can be obtained by the steering wheel angle sensor. The vehicle information in front refers to the information of the vehicle in front of the vehicle, including the position of the vehicle in front and / or the speed of the vehicle in front. Weather information includes sunny days, rainy days or haze days. The weather information can be obtained from the weather forecast software at the location of the vehicle, or obtained after analysis by the relevant sensors of the vehicle load (such as light sensors, rain sensors, visibility sensors, etc.). Navigation information includes road information ahead, which can be obtained from the information sent by the vehicle navigation or the mobile terminal (such as a smart phone) navigation. The road information ahead and the current road information each independently include at least one of the road type, road slope, road curvature or road surface flatness, and the road type includes urban roads, rural roads or highways, etc. The reason why the driving data includes the current road information and navigation information is that on the highway, the light beam will focus on the distant road, while on the urban road, the light beam will be adjusted to a wider range; when the navigation prompts that there is a curve ahead, it is necessary to adjust the direction of the light beam in advance to ensure that the driver has enough vision when entering the curve.

[0048] MEMS (Micro-Electro-Mechanical Systems) micro-vibration mirror is a miniature reflector manufactured using micro-machining technology, which is mainly used for guiding and controlling light beams. In the present invention, the MEMS micro-vibration mirror includes a movable reflective lens, which is designed to be able to perform tiny mechanical swings on two axes. The swing of the reflective lens is driven by an actuator integrated under the lens. When voltage or current is applied, the actuator generates force to deflect the reflective lens around a hinge or support point, thereby changing the reflection path of the incident light. The present invention can adjust the deflection angle of the reflective lens in real time based on driving data. This method allows the light beam to be accurately scanned in space to form the desired beam distribution pattern, thereby achieving control of the illumination range of the LED low beam, the LED high beam, and the laser headlight.

[0049] Under different weather information, the required luminous intensity of the light source is different. For example, the luminous intensity of the light source is higher on rainy days or foggy days than on sunny days. For different headlights, the luminous intensity is also different. For example, the luminous intensity of the low beam is relatively low, and the luminous intensity of the high beam is relatively high. The luminous intensity is also different for the two-lamp headlight system and the four-lamp headlight system. Therefore, it is necessary to control the luminous intensity of the light source in combination with the driving data. It should be noted that the luminous intensity of each light source can be as low as 0, which means that the light source does not emit light, thereby achieving the effect of controlling which headlights emit light and which headlights do not emit light. For example, in extremely severe weather conditions, such as dense fog or heavy rain, the rear fog lights can be turned on to improve the visibility of the vehicle to the rear vehicles, and the rear fog lights can be turned off on sunny days.

[0050] The emission frequency of the light source corresponds to the color of the light emitted by the light source, and the color of the light is different at different frequencies. In general driving environments, the color of the light is white, that is, white light. However, in some special cases, the color of the light needs to be changed to achieve a warning effect, such as non-emergency but attention-demanding situations, such as a system is working (such as ESP is preventing skidding) or there is a potential problem (such as ABS system failure). In this case, the headlights and taillights will display yellow; when a serious fault or dangerous situation requires immediate action, such as a speed sensor failure that causes engine control problems, it may trigger the headlights and taillights to display red, requiring the driver to stop as soon as possible and seek help.

[0051] The light source emission forms include continuous light or flashing light (i.e. light emitted at certain intervals). Continuous light is generally used for lighting, and flashing light is generally used for warning. For example, double flash lights are a form of flashing light.

[0052] Optionally, the vehicle light control information in the sample data may be obtained based on actual operation calibration of the driver under different driving data.

[0053] S120: In response to the model to be trained based on the support vector machine, obtain prediction data of the sample data.

[0054] The model to be trained in step S120 is based on a support vector machine, which is a binary classification model. Its purpose is to find a hyperplane to segment the sample. The principle of segmentation is to maximize the interval. The support vector machine maps the vector to a higher dimensional space, in which a maximum interval hyperplane is established. Two parallel hyperplanes are built on both sides of the hyperplane that separates the data. The separation hyperplane with a suitable direction is established to maximize the distance between the two parallel hyperplanes.

[0055] Optionally, obtaining predicted data of the sample data in response to the model to be trained based on a support vector machine includes:

[0056] Inputting the sample feature vector of the sample data into a model to be trained based on a support vector machine;

[0057] Mapping the sample feature vector to a high-dimensional space according to the kernel function in the model to be trained;

[0058] The sample feature vector is processed based on the hyperplane equation, the objective function and the kernel function, and the predicted data is output.

[0059] For binary classification problems, we can give a training sample data set D = {(x1, y1), (x2, y2), ..., (x m,y m )}y i ={-1,+1} In the sample space, the partitioning hyperplane can be described by the following linear equation: T x+b=0;

[0060] Where w is the normal vector, which specifies the direction of the hyperplane; b is the displacement, which specifies the distance from the origin to the hyperplane. If you want SVM to be able to effectively train the data set, then for any (x i ,y i )∈D, all have constraints: y i (w T x i +b)≥1. If the hyperplane to the nearest sample points on both sides also makes this formula valid, they are called support vectors.

[0061] The sum of the distances from the hyperplane to the support vectors on both sides is called the margin.

[0062] For better generalization, we hope that the sample points are on the correct side of the hyperplane and the margin is as large as possible. At this time, the hyperplane that maximizes the interval is the best separation plane. The optimization problem is:

[0063]

[0064] This method is also known as the hard margin support vector machine algorithm. When the original samples are inseparable, the soft margin is introduced to relax the hard margin constraint and reflect it in the objective function in the form of a penalty term. The optimization problem becomes:

[0065] w T x+b=0

[0066]

[0067] sty i ( <w T ,x i >+b)≥1-ζ i

[0068] ζ i ≥0,i=1,...,m,

[0069] Among them, C is the regularization parameter, which indicates the penalty intensity; ζ i Represents the amount of separation error. The larger it is, the greater the penalty value is.

[0070] However, for higher-dimensional data sets, the problem of two-dimensional plane classification is unsolvable and needs to be converted into high dimensions. Therefore, in this embodiment, a kernel function is used to map the sample feature vector to a high-dimensional space, and then the sample feature vector is processed based on the hyperplane equation, the objective function and the kernel function to output the predicted data.

[0071] Optionally, before inputting the sample feature vector of the sample data into the model to be trained based on a support vector machine, the method further includes:

[0072] Determining a sample feature set according to the driving data in the sample data;

[0073] The sample feature set is serialized, standardized and normalized in sequence to obtain a sample feature vector of the sample data.

[0074] In this embodiment, after the sample features are serialized, since the data have great differences in quantity, which can easily affect (dominate) the results, the data are standardized and normalized to facilitate subsequent calculations.

[0075] Optionally, the standardized mathematical formula is: Among them, X′ is the result after standardization, x min is the minimum value of the sample, x max is the maximum value of the sample, and x is the original data;

[0076] The normalized mathematical formula is: X″=X′*(x max -x min )+x min , where X″ is the normalized result.

[0077] The standardization in this implementation is actually to calculate the ratio of the distance from the element to the minimum value in each column to the distance between the maximum value and the minimum value of the column, which is actually scaling the data to the [0,1] interval; normalization is actually mapping the standardized data to the given [min,max] interval.

[0078] S130. Update the model parameters of the model to be trained to reduce the loss of the prediction data and the sample data, and obtain a headlight control model, wherein the headlight control model is used to respond to the driving data and predict the headlight control information.

[0079] Optionally, updating the model parameters of the model to be trained to reduce the loss of the prediction data and the sample data to obtain the vehicle light control model includes:

[0080] According to the predicted data and the headlight control information in the sample data, the kernel function is adjusted to reduce the loss of the predicted data and the sample data, thereby obtaining a headlight control model.

[0081] In S120, it is mentioned that the present embodiment preferably adopts a kernel function to map the sample feature vector to a high-dimensional space. Therefore, the selection of the kernel function has a great influence on the accuracy of the model. The present embodiment adopts a method of adjusting the kernel function to optimize the model.

[0082] In this implementation, the kernel function includes a linear kernel function, a polynomial kernel function, a Gaussian kernel function and a sigmoid kernel function. The sample data is trained by using the above different kernel functions to determine the selection of the kernel function.

[0083] Optionally, adjusting the kernel function according to the predicted data and the headlight control information in the sample data includes:

[0084] Determining an evaluation index of the predicted data according to the predicted data and the vehicle light control information in the sample data; the evaluation index is a harmonic mean of the accuracy and recall of the predicted data;

[0085] The kernel function is adjusted according to the evaluation index.

[0086] This implementation manner determines an evaluation index of the prediction data according to the prediction data and the vehicle light control information in the sample data, and then adjusts the kernel function according to the evaluation index.

[0087] Among them, the accuracy Precision is calculated using the following formula: TP is the number of positive samples correctly predicted by the model, and FP is the number of negative samples correctly predicted by the model. The recall rate Recall is calculated using the following formula: FN is the number of positive samples that are incorrectly predicted as negative samples by the model. The evaluation index F1 is calculated using the following formula:

[0088] The experimental results show that the evaluation index F1 of the linear kernel function is about 0.89, the evaluation index F1 of the polynomial kernel function is about 0.92, the evaluation index F1 of the Gaussian kernel function is about 0.82, and the evaluation index F1 of the sigmoid kernel function is about 0.59. Therefore, the adjusted kernel function is determined to be the polynomial kernel function.

[0089] The expression of the polynomial kernel function is: K(x,z)=(<x,z> +c) d, where d is a constant, indicating the degree of the polynomial, that is, the coefficient of the highest order term, and the default value is set to 3. c is also a constant, which is used to adjust the contribution of high-order terms and low-order terms to the final kernel function. When the constant term is 0, the contribution of high-order terms and low-order terms to the kernel function is equal; when the constant term is not 0, the contribution of high-order terms is more significant than that of low-order terms. In this embodiment, since the sample feature set is relatively small, the contribution of high-order terms and low-order terms to the kernel function is equal, so it is set to 0.

[0090] In this embodiment, cross operation is performed on the data of each feature set. For example, for two vectors of vehicle speed a=(a1, a2, ..., an) and steering wheel angle b=(b1, b2, ..., bn), their dot product is defined as: K(x, z)=(<a,b> +c) 3 =(a1b1+a2b2+....+anbn) 3 ; For example, for two vectors of vehicle speed a = (a1, a2, ..., an) and the position of the vehicle ahead d = (d1, d2, ..., dn), their dot product is defined as: K(x, z) = (<a,d> +c) 3 =(a1d1+a2d2+....+andn) 3 ; and so on.

[0091] The driving data in the above-mentioned headlight control model training method includes vehicle speed, steering wheel angle, information of the vehicle ahead, current road information, weather information and navigation information. The headlight control information includes the deflection angle of the reflective lens in the MEMS micro-vibration mirror, the luminous intensity of the light source, the emission frequency of the light source and the emission form of the light source. The model to be trained is trained based on the driving data and the headlight control information. Compared with the prior art method that can only adjust the brightness of the light beam according to the vehicle speed, the driving data is more diversified and the headlight control information is richer. Therefore, based on the vehicle's own driving conditions and environmental conditions, the illumination range, brightness, color and form (continuous light or flashing light) of the headlight can be automatically and intelligently controlled, thereby improving driving comfort and safety. Moreover, during the training process, the present invention reduces the loss of prediction data and sample data by updating the model parameters, and can obtain a headlight control model with higher accuracy. In addition, the present invention combines the MEMS micro-vibration mirror with the support vector machine to further improve the accuracy of headlight control.

[0092] Example 2

[0093] This embodiment provides a vehicle light control method, including:

[0094] In response to the vehicle lights being turned on, the driving data is input into the vehicle light control model to obtain vehicle light control information; wherein the driving data includes vehicle speed, steering wheel angle, information of the vehicle ahead, current road information, weather information and navigation information, and the vehicle light control model is trained using the vehicle light control model training method described in Example 1;

[0095] A vehicle light control operation is performed according to the vehicle light control information.

[0096] The vehicle light control method of this embodiment is based on the vehicle light control model training method described in Example 1 to obtain vehicle light control information, and then execute vehicle light control operations, which has the advantages of intelligent, diversified and precise control.

[0097] Example 3

[0098] like Figure 2 As shown, this embodiment provides a vehicle light control model training device, including:

[0099] The sample data acquisition module 201 is used to acquire sample data for headlight control, wherein the sample data includes driving data and headlight control information, wherein the driving data includes vehicle speed, steering wheel angle, information about the vehicle ahead, current road information, weather information, and navigation information, and the headlight control information includes the deflection angle of the reflective lens in the MEMS micro-vibration mirror, the luminous intensity of the light source, the emission frequency of the light source, and the emission form of the light source;

[0100] A sample data prediction module 202, configured to obtain prediction data of the sample data in response to a model to be trained based on a support vector machine;

[0101] The headlight control model determination module 203 is used to update the model parameters of the model to be trained to reduce the loss of the prediction data and the sample data, and obtain the headlight control model, wherein the headlight control model is used to respond to the driving data and predict the headlight control information.

[0102] The device is used to execute the method described in Example 1, and thus has at least functional modules and beneficial effects corresponding to the above method.

[0103] Example 4

[0104] This embodiment provides a vehicle light control system, including:

[0105] A data collection unit, used to collect driving data of the vehicle, wherein the driving data includes vehicle speed, steering wheel angle, information of the vehicle ahead, current road information, weather information and navigation information;

[0106] A computing unit, configured to input the driving data into a headlight control model in response to the headlight being turned on, to obtain headlight control information; the headlight control model is trained using the headlight control model training method described in Example 1;

[0107] The control unit is used to perform a vehicle light control operation according to the vehicle light control information.

[0108] The vehicle light control model in this embodiment is trained using the vehicle light control model training method described in Embodiment 1, and thus has at least the same advantages as Embodiment 1.

[0109] Among them, the above-mentioned data acquisition unit (mostly a sensor) is an indispensable unit in various functions of the vehicle. This embodiment does not require additional sensors suitable for data acquisition of car lights. Basic functions can be achieved using the vehicle's own speed, angular velocity and other sensors.

[0110] Furthermore, the above-mentioned headlight control system also includes: an intelligent interaction unit, which is used to display the headlight control information and headlight status information, respond to the headlight control instruction input by the user, and send the headlight control instruction to the control unit so that the control unit executes the headlight control instruction, and the headlight control instruction has a higher priority than the headlight control information. This method further improves user interactivity on the basis of automatic control of the headlights, and can feed back the headlight status information to the user, so that the user can accurately grasp the headlight status.

[0111] Optionally, the user can input the vehicle light control command by triggering the control module of the intelligent interaction unit or inputting voice. Optionally, the intelligent interaction unit can be integrated into an IVI (In-Vehicle Infotainment) system.

[0112] Furthermore, the LIN signal is used to report the status information of the vehicle lights to the intelligent interaction unit. Since the reporting of the status information of the vehicle lights does not require higher bandwidth and larger network capacity like the CAN protocol, the less cost-effective LIN can be used. LIN uses a single-wire serial communication, which reduces the complexity and cost of the wiring harness compared to the twisted pair of CAN. At the same time, the LIN transceiver is simpler than that of CAN, which reduces the hardware cost, and its single-wire architecture also means lower power consumption.

[0113] Furthermore, the data transmission from the computing unit to the control unit, and the data transmission from the intelligent interaction unit to the control unit, are carried out using the CAN protocol. The headlights play a great role in automobile safety, so in this embodiment, the CAN protocol is used for data transmission, which can effectively resist electromagnetic interference and noise and ensure the reliability of data transmission.

[0114] Example 5

[0115] like Figure 3 As shown, this embodiment provides an electronic device, including:

[0116] at least one processor; and

[0117] a memory communicatively connected to at least one of the processors; wherein,

[0118] The memory stores instructions executable by at least one of the processors, and the instructions are executed by at least one of the processors to enable at least one of the processors to perform the above method. At least one processor in the electronic device can perform the above method, and thus has at least the same advantages as the above method.

[0119] Optionally, the electronic device also includes an interface for connecting various components, including a high-speed interface and a low-speed interface. Various components are interconnected using different buses and can be installed on a common mainboard or installed in other ways as needed. The processor can process instructions executed in the electronic device, including instructions stored in or on a memory to display graphical information of a GUI (Graphical User Interface) on an external input / output device (such as a display device coupled to an interface). In other embodiments, if necessary, multiple processors can be used together with multiple memories, and / or multiple buses can be used together with multiple memories. Similarly, multiple electronic devices can be connected (for example, as a server array, a group of blade servers, or a multi-processor system), and each device provides some necessary operations. Figure 3 A processor 301 is taken as an example.

[0120] The memory 302, as a computer-readable storage medium, can be used to store software programs, computer executable programs and modules, such as program instructions / modules corresponding to the headlight control model training method in the embodiment of the present invention (for example, the sample data acquisition module, sample data prediction module, and headlight control model determination module in the headlight control model training device). The processor 301 executes various functional applications and data processing of the device by running the software programs, instructions and modules stored in the memory 302, that is, implements the above-mentioned headlight control model training method.

[0121] The memory 302 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system and at least one application required for a function; the data storage area may store data created according to the use of the terminal, etc. In addition, the memory 302 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some instances, the memory 302 may further include a memory remotely arranged relative to the processor 301, and these remote memories may be connected to the device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0122] The electronic device may further include: an input device 303 and an output device 304. The processor 301, the memory 302, the input device 303 and the output device 304 may be connected via a bus or other means. Figure 3 The example of connecting through bus is taken in the following.

[0123] The input device 303 can receive input digital or character information, and the output device 304 can include a display device, an auxiliary lighting device (e.g., LED), and a tactile feedback device (e.g., a vibration motor), etc. The display device can include, but is not limited to, a liquid crystal display (LCD), a light emitting diode (LED) display, and a plasma display. In some embodiments, the display device can be a touch screen.

[0124] Example 6

[0125] This embodiment provides a computer-readable storage medium, on which computer instructions are stored, and the computer instructions are used to enable a computer to execute the above method. The computer instructions on the computer-readable storage medium are used to enable a computer to execute the above method, and thus have at least the same advantages as the above method.

[0126] The medium in the present invention may adopt any combination of one or more computer-readable media. The medium may be a computer-readable signal medium or a computer-readable storage medium. The medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination thereof. More specific examples of the medium (a non-exhaustive list) include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In this document, the medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, device, or device.

[0127] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0128] The program code contained on the computer-readable medium can be transmitted using any appropriate medium, including but not limited to wireless, wire, optical cable, RF (Radio Frequency), etc., or any suitable combination of the above.

[0129] Computer program code for performing the operations of the present invention may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0130] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this application can be executed in parallel, sequentially or in different orders, as long as the expected results of the technical solution disclosed in this application can be achieved, and this document is not limited here.

[0131] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. A vehicle light control model training method, It is characterized in that include: Acquire sample data for headlight control, wherein the sample data includes driving data and headlight control information, wherein the driving data includes vehicle speed, steering wheel angle, information about the vehicle ahead, current road information, weather information, and navigation information, and the headlight control information includes a deflection angle of a reflective lens in a MEMS micro-vibration mirror, light source luminous intensity, light source emission frequency, and light source emission form; In response to the model to be trained based on the support vector machine, obtaining prediction data of the sample data; The model parameters of the model to be trained are updated to reduce the loss of the predicted data and the sample data, and a vehicle light control model is obtained, wherein the vehicle light control model is used to respond to the driving data and predict vehicle light control information.

2. The vehicle light control model training method according to claim 1, It is characterized in that The step of obtaining predicted data of the sample data in response to the model to be trained based on the support vector machine includes: Inputting the sample feature vector of the sample data into a model to be trained based on a support vector machine; Mapping the sample feature vector to a high-dimensional space according to the kernel function in the model to be trained; The sample feature vector is processed based on the hyperplane equation, the objective function and the kernel function, and the predicted data is output.

3. The vehicle light control model training method according to claim 2, It is characterized in that Before inputting the sample feature vector of the sample data into the model to be trained based on the support vector machine, the method further includes: Determining a sample feature set according to the driving data in the sample data; The sample feature set is serialized, standardized and normalized in sequence to obtain a sample feature vector of the sample data.

4. The vehicle light control model training method according to claim 3, It is characterized in that The mathematical formula for the standardization is: Among them, X′ is the result after standardization, x min is the minimum value of the sample, x max is the maximum value of the sample, and x is the original data; The normalized mathematical formula is: X″=X′*(x max -x min )+x min , where X″ is the normalized result.

5. The vehicle light control model training method according to claim 1, It is characterized in that The updating of the model parameters of the model to be trained to reduce the loss of the prediction data and the sample data to obtain the vehicle light control model includes: According to the predicted data and the headlight control information in the sample data, the kernel function is adjusted to reduce the loss of the predicted data and the sample data, thereby obtaining a headlight control model.

6. The vehicle light control model training method according to claim 5, It is characterized in that The step of adjusting the kernel function according to the prediction data and the vehicle light control information in the sample data includes: Determining an evaluation index of the predicted data according to the predicted data and the vehicle light control information in the sample data; the evaluation index is a harmonic mean of the accuracy and recall of the predicted data; The kernel function is adjusted according to the evaluation index.

7. A vehicle light control method, It is characterized in that include: In response to the vehicle lights being turned on, the driving data is input into the vehicle light control model to obtain vehicle light control information; wherein the driving data includes vehicle speed, steering wheel angle, information of the vehicle ahead, current road information, weather information and navigation information, and the vehicle light control model is trained by the vehicle light control model training method according to any one of claims 1 to 6; A vehicle light control operation is performed according to the vehicle light control information.

8. A vehicle light control model training device, It is characterized in that include: A sample data acquisition module is used to acquire sample data for headlight control, wherein the sample data includes driving data and headlight control information, wherein the driving data includes vehicle speed, steering wheel angle, information about the vehicle ahead, current road information, weather information, and navigation information, and the headlight control information includes the deflection angle of the reflective lens in the MEMS micro-vibration mirror, the luminous intensity of the light source, the emission frequency of the light source, and the emission form of the light source; A sample data prediction module, used for obtaining predicted data of the sample data in response to a model to be trained based on a support vector machine; The headlight control model determination module is used to update the model parameters of the model to be trained to reduce the loss of the prediction data and the sample data, and obtain the headlight control model, wherein the headlight control model is used to respond to the driving data and predict the headlight control information.

9. An electronic device, It is characterized in that include: at least one processor, and a memory communicatively coupled to at least one of the processors; The memory stores instructions executable by at least one of the processors, and the instructions are executed by at least one of the processors so that the at least one processor can execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, It is characterized in that The 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 7.

Citation Information

Patent Citations

  • Support vector machine kernel function selection method under sparse representation and application thereof

    CN104462019A

  • Light control method and device based on illuminance clustering and support vector machine

    CN111669881A

  • Automobile head lamp adjusting model creating method, automobile head lamp adjusting method and automobile

    CN113370884A

  • Control method of dot-matrix LED intelligent car lamp curve lighting system based on BP network model

    CN115526288A

  • Model training method and device for automated driving and electronic equipment

    CN116263876A