Vehicle warning method, computer-readable storage medium, electronic device, and vehicle

Abstract

The present application discloses a vehicle warning method, a computer-readable storage medium, an electronic device, and a vehicle. The method comprises: acquiring a driving state of a driver at a current moment; and on the basis of the driving state, controlling a first warning apparatus to give a warning, wherein the first warning apparatus is a warning apparatus of which the warning effect satisfies a set requirement in the driving state. The vehicle warning method, the computer-readable storage medium, the electronic device, and the vehicle of the present application can significantly improve the warning effect.

Description

Vehicle reminding method, computer readable storage medium, electronic device and vehicle

[0001] The present application claims priority to the Chinese patent publication with the application number 202410867054.1, the title of "Vehicle reminding method, computer readable storage medium, electronic device and vehicle", which was filed on June 28, 2024, in the China Patent Office, the entire content of which is incorporated herein by reference. TECHNICAL FIELD

[0002] The present application relates to the technical field of vehicles, in particular to a vehicle reminding method, a computer readable storage medium, an electronic device and a vehicle. BACKGROUND

[0003] At present, there are many in-vehicle safety reminding devices on vehicles, such as indicator lights (reminding by light), steering wheels (reminding by vibration), etc. Vehicles usually remind users through fixed reminding devices. With the increase of use time, the sensitivity of users to some reminding devices may decrease, resulting in poor reminding effect.

[0004] Therefore, it is necessary to improve to at least partially solve the above problems. TECHNICAL SOLUTION

[0005] In the summary section of the application, a series of simplified concepts are introduced, which will be further described in detail in the specific embodiment section. The summary section of the present application does not mean to attempt to limit the key features and necessary technical features of the claimed technical solution, nor to attempt to determine the protection scope of the claimed technical solution.

[0006] In order to at least partially solve the above problems, according to the first aspect of the present application, a vehicle reminding method is provided, comprising:

[0007] Obtaining the driving state of the driver at the current time;

[0008] According to the driving state at the current time, controlling the first reminding device to remind, wherein the first reminding device is a reminding device whose reminding effect meets the set requirements under the driving state at the current time.

[0009] Exemplarily, the first reminding device is randomly selected according to the expected benefits of a plurality of types of reminding devices under the driving state at the current time after normalization processing.

[0010] Exemplarily, the randomly selecting one after normalization processing according to the expected benefits of a plurality of types of reminding devices under the driving state at the current time comprises:

[0011] determining expected benefits of the plurality of types of reminding devices in the driving state at the current time according to reminding device expected benefit data corresponding to the driving state;

[0012] forming a one-dimensional vector of the expected benefits of the plurality of types of reminding devices in the driving state at the current time;

[0013] normalizing the one-dimensional vector;

[0014] taking the value obtained after normalization as the adoption probability of the plurality of types of reminding devices;

[0015] randomly selecting one of the plurality of types of reminding devices as the first reminding device according to the adoption probability.

[0016] Exemplarily, the method further comprises:

[0017] determining an original expected benefit corresponding to the driving state of the driver and the first reminding device at the previous time based on reminding device expected benefit data corresponding to the driving state;

[0018] obtaining an actual benefit of the driver at the current time;

[0019] updating the reminding device expected benefit data based on the original expected benefit and the actual benefit.

[0020] Exemplarily, the obtaining of the actual benefit of the driver at the current time comprises:

[0021] determining a fixed benefit corresponding to the driving state at the current time according to reminding device fixed benefit data corresponding to the driving state;

[0022] determining a maximum expected benefit corresponding to the driving state at the current time according to the reminding device expected benefit data;

[0023] determining the actual benefit at the current time according to the following formula:

[0024] Q 实际 = reward + γ × max(Q(State k+1 ))

[0025] wherein Q 实际 is the actual benefit at the current time, reward is the fixed benefit, γ is a decay rate, and max(Q(State k+1 )) is the maximum expected benefit.

[0026] Exemplarily, the decay rate γ is determined through the following steps:

[0027] obtaining landmark data in various driving states;

[0028] The key data is input into a linear processing unit and compressed into a one-dimensional vector;

[0029] Calculate the Euclidean distance between the vectors output by the linear processing unit, and take the reciprocal to obtain the connectivity matrix between different driving states;

[0030] The connected matrix is ​​row-normalized, and the attenuation rate matrix is ​​output.

[0031] The attenuation rate γ is determined based on the driving state at the previous moment, the driving state at the current moment, and the attenuation rate matrix.

[0032] For example, updating the expected revenue data of the reminder device based on the original expected revenue and the actual revenue includes:

[0033] The expected benefits of the updated driving status and alert device corresponding to the previous moment are calculated using the following formula:

[0034] Q(State k Reminder device k )′=Q(State k Reminder device k )+α×(Q 实际 -Q(State k Reminder device k ))

[0035] Where Q(State) k Reminder device k Q(State)' represents the driving state at the previous moment and the expected updated benefit corresponding to the warning device. k Reminder device k ) represents the original expected benefit corresponding to the driving status and the reminder device at the previous moment, and α is a coefficient, 0 < α < 1;

[0036] Replace the original expected revenue in the expected revenue data of the reminder device with the updated expected revenue.

[0037] For example, the coefficient α is a fixed value or the coefficient α gradually decreases with time.

[0038] For example, before obtaining the driver's current driving state, the method further includes:

[0039] Obtain the driver's identity information;

[0040] The expected revenue data of the reminder device corresponding to the identity information is retrieved based on the identity information.

[0041] Exemplarily, the method further includes:

[0042] obtaining a current state of the vehicle;

[0043] determining an unavailable reminding device corresponding to the current state;

[0044] wherein the expected benefits of the plurality of types of reminding devices in the driving state at the current time, as determined according to the expected benefit data of the reminding devices, do not include the expected benefit of the unavailable reminding device in the driving state at the current time.

[0045] Exemplarily, the driving state is normal driving, driver fatigue, driver smoking, or driver making a phone call.

[0046] Exemplarily, the plurality of types of reminding devices include at least two of a fragrance device, an indicator light, a steering wheel, an air conditioner, a multimedia device, a window, a seat, and a brake device.

[0047] According to a second aspect of the present application, there is provided a computer readable storage medium having stored thereon computer program instructions which, when executed by a processor, implement the steps of the method as described above.

[0048] According to a third aspect of the present application, there is provided an electronic device comprising:

[0049] a memory having stored thereon computer instructions;

[0050] a processor configured to execute the computer instructions in the memory to implement the steps of the method as described above.

[0051] According to a fourth aspect of the present application, there is provided a vehicle comprising the electronic device as described above.

[0052] According to the vehicle reminding method, the computer readable storage medium, the electronic device, and the vehicle of the present application, by controlling the reminding effect of the corresponding reminding device according to the driving state to meet the set requirements, the sensitivity of the user to the fixed reminding device can be effectively avoided from being reduced, and a better reminding effect can be achieved. BRIEF DESCRIPTION OF DRAWINGS

[0053] The following drawings of the present application are hereby incorporated into this application as part of the present application for understanding the present application. The embodiments of the present application and the description thereof shown in the drawings are used to explain the devices and principles of the present application. In the drawings,

[0054] FIG. 1 is a flowchart of a vehicle reminding method according to an embodiment of the present application;

[0055] FIG. 2 is a flowchart of a vehicle reminding method according to an embodiment of the present application;

[0056] FIG. 3 is a flow diagram of a vehicle alerting method according to an embodiment of the present application;

[0057] FIG. 4 is a data flow diagram of a vehicle alerting method according to an embodiment of the present application;

[0058] FIG. 5 is a diagram of a process of obtaining a decay rate matrix.

[0059] Embodiments of the present application

[0060] In the following description, numerous specific details are set forth to provide a more thorough understanding of the present application. However, it will be apparent to one of skill in the art upon

[0061] It should be understood that the present application can be carried out in many ways, and that the application should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the application to those skilled in the art. In the drawings, the size and relative sizes of layers and regions can be exaggerated for clarity. Like reference numerals can be used to denote like elements throughout.

[0062] It will be understood that, although the terms first, second, third, etc. can be used herein to describe various elements, components, regions, layers and / or sections, these elements, components, regions, layers and / or sections should not be limited by these terms. These terms are only used to distinguish one element, component, region, layer or section from another element, component, region, layer or section. Thus, a first element, component, region, layer or section discussed below could be termed a second element, component, region, layer or section without departing from the teachings of the present application.

[0063] Spatially relative terms, such as "beneath", "below", "lower", "under", "above", "upper" and the like, can be used herein for ease of description to describe one element or feature's relationship to another element(s) or feature(s) as illustrated in the figures. It will be understood that the spatially relative terms are intended to encompass different orientations of the device in use or operation in addition to the orientation depicted in the figures.

[0064] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting thereof. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises" and / or "comprising", when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. As used herein the term "and / or" includes any and all combinations of associated items.

[0065] A vehicle reminding method according to an embodiment of the present application is exemplarily illustrated with reference to FIG. 1. The method comprises:

[0066] S100: obtaining a driving state of a driver at a current time.

[0067] S200: controlling a first reminding device to remind according to the driving state at the current time.

[0068] The first reminding device is a reminding device whose reminding effect meets the set requirement in the driving state.

[0069] The vehicle reminding method according to the present application can effectively avoid the decrease of the sensitivity of the user to the fixed reminding device by controlling the corresponding reminding device whose reminding effect meets the set requirement to remind according to the driving state, and can achieve better reminding effect.

[0070] Specifically, in step S100, after the DMS (Driver Monitor System) on the vehicle is started, the image information of the driver can be collected by the camera on the vehicle, and the driving state of the driver at the current time can be obtained based on the image information, wherein the driving state can be normal driving, driver fatigue, driver smoking or driver making a phone call, etc., wherein the driver fatigue can be further divided into very tired driver, relatively tired driver and slightly tired driver, etc., which can be configured by the skilled person as needed. Exemplarily, based on the image information collected by the camera, the driving state of the driver at the current time can be determined by means such as image recognition.

[0071] In step S200, the first reminding device is randomly selected according to the expected benefits of the plurality of types of reminding devices in the driving state at the current time after normalization processing, that is, step S200 is to randomly select one as the first reminding device according to the expected benefits of the plurality of types of reminding devices in the driving state at the current time after normalization processing, and control the first reminding device to remind. Wherein, the expected benefit is a pre-configured benefit value, which is used to represent the expected reminding effect, the higher the expected benefit, the better the expected reminding effect. Normalization processing is a common data preprocessing method, which is used to map the original data to a specific interval, usually a decimal between [0, 1]. After normalization processing of the expected benefits of the plurality of types of reminding devices in the driving state at the current time, the result can be used as the adoption probability of the corresponding reminding device, and the higher the expected benefit of the reminding device, the greater the corresponding adoption probability.

[0072] Therefore, according to the adoption probability, a reminding device is randomly selected as the first reminding device to remind, on the one hand, the reminding device with good reminding effect (high expected benefit) has a high probability of being selected, thereby effectively guaranteeing the basic effect of reminding and avoiding the use of reminding devices with poor reminding effect for reminding; on the other hand, there is an opportunity to select different reminding devices for reminding, which can provide a fresh feeling to the user, avoid the sensitivity of the user to the same reminding device from being reduced, and achieve better reminding effect.

[0073] Referring to FIG. 2, after normalization processing of the expected benefits of the plurality of types of reminding devices in the driving state at the current time, one is randomly selected, which specifically includes the following steps:

[0074] S210: determining the expected benefits of the plurality of types of reminding devices in the driving state at the current time according to the reminding device expected benefit data corresponding to the driving state.

[0075] Specifically, the driving state corresponding reminder device expected benefit data can be a driving state-reminder device expected benefit table, which shows the expected benefits of various types of reminder devices under various driving states. It should be noted that the driving state corresponding reminder device expected benefit data can also be other forms of data that can represent the expected benefits of various types of reminder devices under various driving states. Table 1 below is a schematic driving state-reminder device expected benefit table, wherein State1, State2, State3 are three different driving states, and exemplary State1, State2, State3 can be normal driving, driver fatigue and driver smoking, respectively. Reminder device 1 to reminder device 5 can be any of the following five types of reminder devices: fragrance device (which can remind by the strength of fragrance), indicator light (which can remind by light), steering wheel (which can remind by vibration), air conditioner (which can remind by adjusting temperature and / or air volume), multimedia device (which can remind by image and / or sound), window (which can remind by lifting), seat (which can remind by vibration) and brake device (which can remind by brake). It should be noted that the specific number of reminder devices is not limited to five, as long as it includes at least two of the above reminder devices. After obtaining the driving state of the driver at the current time through step S100, the expected benefits of various reminder devices corresponding to the driving state can be determined by looking up the driving state corresponding reminder device expected benefit data, i.e., the driving state-reminder device expected benefit table. For example, when the driving state is State3, the expected benefits of the five reminder devices under the current driving state are 5, 2, 3, 1 and 4 respectively according to the driving state-reminder device expected benefit table. It should be noted that the higher the expected benefit, the better the reminder effect. It should be noted that the driving state-reminder device expected benefit table can be formed by big data, expert opinions, experimental tests and the like by those skilled in the art.

[0076] Table 1: Driving state-reminder device expected benefit table

[0077] S220: Form a one-dimensional vector of the expected benefits of various types of reminder devices under the current driving state.

[0078] The one-dimensional vector refers to an array or list with a single dimension, that is, an ordered real number sequence, which can be represented as (x1, x2,..., xn), where each xi is a real number. Such a vector has a fixed dimension, referred to as "one-dimensional", because it has only one "row of numbers" or "column of numbers". In step S220, the expected benefits of the plurality of types of reminder devices in the driving state at the current time are combined into a one-dimensional vector, that is, an array, each real number in the array being the expected benefit of a reminder device. For example, when the driving state at the current time is State3, the expected benefits of the five types of reminder devices corresponding thereto can form a one-dimensional vector (5, 2, 3, 1, 4).

[0079] S230: Normalizing the one-dimensional vector.

[0080] Specifically, the one-dimensional vector can be normalized using a Softmax function or other suitable function. The Softmax function is an activation function that can normalize a numerical vector into a probability distribution vector, and the sum of the probabilities is 1. For example, after normalizing the one-dimensional vector (5, 2, 3, 1, 4) using the Softmax function, the probability distribution vector obtained is (0.636, 0.0317, 0.0861, 0.0117, 0.2345).

[0081] S240: Taking the value obtained after normalization as the adoption probability of the plurality of types of reminder devices.

[0082] Specifically, each value in the probability distribution vector obtained after normalization is taken as the adoption probability of the corresponding reminder device. For example, when the probability distribution vector is (0.636, 0.0317, 0.0861, 0.0117, 0.2345), 0.636 is taken as the adoption probability of reminder device 1, 0.0317 is taken as the adoption probability of reminder device 2, 0.0861 is taken as the adoption probability of reminder device 3, 0.0117 is taken as the adoption probability of reminder device 4, and 0.2345 is taken as the adoption probability of reminder device 5.

[0083] S250: Randomly selecting one of the plurality of types of reminder devices as the first reminder device according to the adoption probability.

[0084] For example, when the driving state at the current moment is State 3, the expected benefits of the reminder devices 1-5 are 5, 2, 3, 1, and 4 respectively, and the adoption probabilities of the reminder devices 1-5 are 0.636 (63.6%), 0.0317 (3.17%), 0.0861 (8.61%), 0.0117 (1.17%), and 0.2345 (23.45%) respectively. As can be seen, the better the reminder effect (the greater the expected benefit) of the reminder device, the greater the probability of being selected as the first reminder device to perform the reminder, thereby effectively guaranteeing the basic effect of the reminder and avoiding frequent reminders using reminder methods with poor reminder effects. At the same time, the reminder devices with relatively poor reminder effects also have a relatively low chance of being selected as the first reminder device to perform the reminder, thereby providing a fresh feeling to the user, avoiding the user's sensitivity to the same reminder device from decreasing, and achieving a better reminder effect.

[0085] Exemplarily, in some embodiments, before step S100, the vehicle reminder method further includes the following steps: first, obtaining the identity information of the driver, and then calling the reminder device expected benefit data corresponding to the identity information according to the identity information.

[0086] Specifically, after the DMS (Driver Monitor System, driver monitoring system) on the vehicle is turned on, the driver can be face-recognized by the camera, and the identity information of the driver can be determined according to the face information of the driver. Then, the driver data corresponding to the identity information is obtained from the local or cloud, wherein the driver data includes the reminder device expected benefit data, such as the driving state-reminder device expected benefit table. Thus, the reminder device expected benefit data matches the habits of the driver and the acceptance degree of various reminders, and selecting the reminder device to perform the reminder based on the reminder device expected benefit data can achieve a better reminder effect.

[0087] ​Exemplarily, when the vehicle is in a certain state, certain reminding devices can not be able to be normally used or can bring certain safety risks after being used. Thus, in some embodiments, the vehicle reminding method further comprises the following steps: obtaining a current state of the vehicle, and determining an unusable reminding device corresponding to the current state. Wherein, the expected benefits of the plurality of types of reminding devices in the driving state at the current time determined according to the reminding device expected benefit data table do not include the expected benefit of the unusable reminding device in the driving state at the current time. That is, in steps S210 and S220, the expected benefits corresponding to the unusable reminding device are excluded. For example, when the driving state at the current time is State3 and the unusable reminding device is reminding device 1, the expected benefits of the plurality of types of reminding devices in the driving state at the current time determined according to the driving state-reminding device expected benefit table are respectively the expected benefits 2, 3, 1, 4 of reminding devices 2-5, and correspondingly, the one-dimensional vector composed of the expected benefits of the four reminding devices in the driving state State3 at the current time is (2, 3, 1, 4). The value obtained by normalizing the one-dimensional vector (2, 3, 1, 4) is used as the adoption probability of the four reminding devices 2-5, and then one of the four reminding devices is randomly selected as the first reminding device according to the adoption probability.

[0088] Specifically, the current state of the vehicle can include a driving state of the vehicle (such as a driving speed of the vehicle, an automatic driving mode or an auxiliary driving mode of the vehicle, etc.) or a state of a vehicle component (such as a state of an in-vehicle fragrance). For example, when the driving speed of the car exceeds 100 km / h, the corresponding unusable reminding device is the window; for example, when the vehicle is in ACC overtake mode, the corresponding unusable reminding device is the brake device; for example, when the fragrance material is insufficient, the corresponding unusable reminding device is the fragrance device. Exemplarily, in some embodiments, a mapping table of the current state of the vehicle and the unusable reminding device can be established, and after obtaining the current state of the vehicle, the corresponding unusable reminding device can be determined according to the mapping table.

[0089] Referring to FIG. 3, in the present embodiment, the vehicle reminding method further comprises the following steps:

[0090] S310: determining the driving state of the driver at the previous time and the original expected benefit corresponding to the first reminding device based on the reminding device expected benefit data corresponding to the driving state;

[0091] S320: obtaining the actual benefit of the driver at the current time;

[0092] S330: updating the reminding device expected benefit data corresponding to the driving state based on the original expected benefit and the actual benefit.

[0093] Next, the above steps will be specifically described with reference to FIG. 4.

[0094] Assume that the current time is the k+1 time, and the driving state at its last time, i.e. the k time, is State k , and the corresponding first reminding device is reminding device k . Reminding device k may be determined through the above steps S210 to S250.

[0095] In step S310, the driving state-reminding device expected reward table (for example, the above-mentioned table 1) is searched to determine the driving state of the driver at the last time and the original expected reward Q(State k , reminding device k ) of the reminding device corresponding thereto. Q(State k , reminding device k ) is the reward expected to be obtained at the k+1 time, which represents the expected reminding effect. The original expected reward is the expected reward corresponding to the driving state State k and the reminding device k before the update.

[0096] In step S320, first, the fixed reward corresponding to the driving state (State k+1 ) at the current time (the k+1 time) is determined according to the reminding device fixed reward data corresponding to the driving state. The reminding device fixed reward data corresponding to the driving state can be a driving state fixed reward table, which shows the fixed reward corresponding to various driving states. It should be noted that the reminding device fixed reward data corresponding to the driving state can also be other forms of data capable of representing the fixed reward that can be obtained in each driving state. Table 2 below is a schematic driving state fixed reward table, which shows the fixed reward (reward) corresponding to various driving states. The fixed reward indicates the beneficial degree of the driver in this state, which is subjectively rated by engineers, such as a lower value when the driver is in an abnormal state such as a fatigue state, and a maximum value when the driver is in a normal driving state.

[0097] Table 2: Driving state fixed reward table

[0098] Then, the maximum expected reward max(Q(State k+1 )) corresponding to the driving state (State k+1 ) at the current time is determined according to the reminding device expected reward data corresponding to the driving state. That is, the driving state-reminding device expected reward table (for example, the above-mentioned table 1) is searched to determine the maximum value among the expected rewards of various reminding devices corresponding to the driving state (State k+1 ) at the current time.

[0099] Then, the actual payoff at the current moment is determined according to the following formula:

[0100] Q 实际 =reward + γ × max(Q(State) k+1 ))

[0101] Among them, Q 实际 This represents the actual reward at the current time (time k+1). The actual reward characterizes the actual reminder effect achieved; a higher actual reward indicates a better reminder effect. `reward` is a fixed reward, and `γ` is the decay rate. `max(Q(State)` is the maximum reward value. k+1 )) 为最大预期收益 Maximum expected return max(Q(State)) k+1 The product of the expected future return and the decay rate γ represents the degree of influence of the current decision on the expected future return when in this state. The larger the value of γ, the greater the influence of the future return.

[0102] In step S330, firstly, the expected updated benefit corresponding to the driving state and the reminder device at the previous moment is calculated according to the following formula:

[0103] Q(State k Reminder device k )′=Q(State k Reminder device k )+α×(Q 实际 -Q(State k Reminder device k ))

[0104] Where Q(state) k Reminder device k Q(State)' represents the driving state at the previous moment and the expected updated benefit corresponding to the warning device. k Reminder device k )′ represents the original expected benefit corresponding to the driving state and the reminder device at the previous moment, and α is a coefficient, 0<α<1.

[0105] Then, use the updated expected return Q(State) k Reminder device k Replace the original expected revenue in the expected revenue data of the reminder device, that is, replace the original expected revenue Q(Statek, reminder device) in the driving state-reminder device expected revenue table. k ).

[0106] It should be noted that a is transcendental, which determines the strength of iteration. The greater the value of a, the greater the strength. For example, the value of a gradually decreases with the change of time, that is, the value of a corresponding to the current time is greater than the value of a corresponding to the last time, which can promote the machine to converge stably to the most effective strategy for a specific user. In some embodiments, a can be a fixed value, that is, the expected return in the driving state-reminder device expected return table is iterated with a fixed strength.

[0107] The vehicle reminder method of the present application can maintain the randomness of the selection of the first reminder device to adapt to the changing real scene and different driver habits. Its superiority is reflected in that: in a certain state, if the return of using a certain reminder device is always the largest, it can be predicted that the probability of using this device will approach 100% in the iteration, at this time the strategy collapses into the existing fixed transfer strategy. Therefore, the present application has the ability to expand the randomness of the selection of the reminder device on the basis of retaining the existing fixed reminder device, and realizes the adaptive adjustment of the fatigue monitoring system. And it can adaptively learn the most effective reminder device for different users in different scenes and at different times to improve the efficiency of the reminder.

[0108] For example, when the current state of the driver is a fatigue state, according to the driving state-reminder device expected return table, it is found that the expected return of the reminder device indicator light in this state is the largest and accounts for 80% of the total expected return of the reminder device. Therefore, the vehicle has an 80% probability of selecting the indicator light as the first reminder device for reminding. Assuming that the machine selects the indicator light, and the actual return of the state of the driver monitored at the next time is greater than the expected return at the last time, the selection of the indicator light will be reinforced. Specifically, when the driver is in a fatigue state next time, the probability of selecting the indicator light will be increased, such as to 85%.

[0109] Considering that there is a certain similarity between different driving states, if it is monitored that the state of the driver is similar to the state at the last time after a certain reminder device is adopted (such as from a very fatigue state to a relatively fatigue state), it indicates that the reminder device adopted at this moment is not effective enough, and the decay rate γ should be increased accordingly to select a more effective reminder device.

[0110] In this embodiment, the decay rate γ can be determined based on the Markov principle. The determination steps of the decay rate γ are exemplarily described below with reference to FIG. 5.

[0111] Firstly, the characteristic data (State0, State1, State2, etc.) in various driving states is obtained, which can be the picture or video data of the driver, which can be collected by the camera on the vehicle.

[0112] Then, the landmark data is input into a linear processing unit and compressed into a one-dimensional vector. A one-dimensional vector refers to an array or list with a single dimension, and each element typically represents a numerical value. Compressing landmark data (such as images or videos) into a one-dimensional vector can reduce the dimensionality of the data while preserving key information, reducing storage space, and improving processing speed. The linear processing unit typically refers to a module in a neural network that performs linear transformations on input data. This linear processing unit can be a long matrix used to embed landmark data (such as images or videos) into a fixed-dimensional vector (such as 1*512 dimensions), with one vector corresponding to the landmark data in a driving state. Specifically, Vit (prior art) and CLIP (prior art) can be used to achieve this.

[0113] Then, the Euclidean distance between each vector output by the linear processing unit is calculated, and the reciprocal is taken to obtain a connectivity matrix between different driving states. The Euclidean distance refers to the real distance between two points or vectors in an n-dimensional space, which can be used to compare the similarity (the closer the distance, the higher the similarity) of two points or vectors. The connectivity matrix is a matrix representing the similarity between different data points, with each value in the connectivity matrix representing the distance between the vector corresponding to a driving state and the vector corresponding to another driving state.

[0114] Then, the connectivity matrix is normalized by row, and a transition probability matrix, i.e., a decay rate matrix, is output. Table 3 below is an exemplary decay rate matrix. Each row of the decay rate matrix represents the probability (i.e., decay rate) of transitioning from the driving state represented by the row to other states, and the probability value is inversely proportional to the Euclidean distance between the embedding vectors corresponding to each driving state, i.e., the closer the distance between two driving states, the greater the probability of mutual transition between them. Taking the transition probability as the decay rate γ, when the current state is similar to the previous state, γ will consider the future reward more, which meets the expected requirements.

[0115] Table 3 Decay rate matrix

[0116] Finally, the decay rate γ is determined based on the driving state at the previous time, the driving state at the current time, and the decay rate matrix. That is, the decay rate matrix is searched based on the driving state at the previous time and the driving state at the current time to obtain the corresponding decay rate γ. Taking Table 3 as an example, when the driving state at the previous time is State2 and the driving state at the current time is State3, the corresponding decay rate γ is 0.4.

[0117] By abstracting the characteristics of different driving states as a Markov chain in state space distance, the application can avoid the machine being limited to the current state during adaptive learning. Thus, differentiated reminders can be effectively implemented, and the reminder effect can be improved.

[0118] In some other embodiments, the attenuation rate γ can not be determined based on the Markov principle, but a matrix of attenuation rates (such as the matrix shown in Table 3) can be directly established based on big data or empirical accumulation, and then the attenuation rate γ can be determined according to the driving state at the last time, the driving state at the current time and the matrix of attenuation rates.

[0119] It should be noted that when the user turns off the vehicle, the DMS stops working, and the latest driving state-reminder device expected benefit table can be saved to the local, uploaded to the cloud if the network is connected, or uploaded to the cloud after waiting for network connection.

[0120] The application also provides a computer readable storage medium, which stores computer program instructions, and the program instructions are executed by a processor to implement the steps of the vehicle reminding method.

[0121] The application also provides an electronic device, which comprises:

[0122] a memory, which stores computer instructions;

[0123] a processor, which is configured to execute the computer instructions in the memory to implement the steps of the vehicle reminding method.

[0124] The application also provides a vehicle, which comprises the electronic device.

[0125] Although the example embodiments have been described herein with reference to the accompanying drawings, it should be understood that the above example embodiments are merely exemplary and are not intended to limit the scope of the application. Those of ordinary skill in the art can make various changes and modifications without departing from the scope and spirit of the application. All such changes and modifications are intended to be included within the scope of the application as claimed in the appended claims.

[0126] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in connection with the embodiments disclosed herein can be realized in electronic hardware, or a combination of computer software and electronic hardware. Whether the functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Those of ordinary skill in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the application.

[0127] In several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the above-described device embodiments are merely illustrative, and the division of the units is merely a logical function division. In actual implementation, another division manner can be used, for example, a plurality of units or components can be combined or integrated into another device, or some features can be omitted or not executed.

[0128] In the specification provided herein, a large number of specific details are described. However, it can be understood that the embodiments of the present application can be practiced without these specific details. In some examples, well-known methods, structures and techniques are not described in detail in order not to obscure the understanding of the present specification.

[0129] Similarly, it should be appreciated that the various features of the application are sometimes grouped together in the description, example embodiments and claims of the application for the purpose of streamlining the disclosure and aiding in the understanding of one or more of the various aspects and embodiments of the application. In addition, it should be understood that broadest scope of the application is not necessarily limited to any specific embodiments, examples, or implementations set forth herein. In some aspects, the methods provided can be

[0130] Those skilled in the art can understand that all the features disclosed in the specification (including the accompanying claims, abstract and drawings) and all the processes or units of any method or device disclosed in this way can be combined in any combination, except that the features are mutually exclusive. Unless explicitly stated otherwise, each feature disclosed in the specification (including the accompanying claims, abstract and drawings) can be replaced by an alternative feature that provides the same, equivalent or similar purpose.

[0131] In addition, those skilled in the art can understand that although some embodiments described herein include certain features and not others included in other embodiments, the combination of features of different embodiments means within the scope of the present application and forms different embodiments. For example, in the claims, any one of the claimed embodiments can be used in any combination.

[0132] It should be noted that the above embodiments illustrate the present application rather than limit the present application, and those skilled in the art can design alternative embodiments without departing from the scope of the appended claims.

Claims

1. A vehicle alert method, wherein, include: Obtain the driver's current driving status; as well as, Based on the current driving status, the first reminder device is controlled to issue a reminder, wherein the first reminder device is a reminder device whose reminder effect meets the set requirements under the current driving status.

2. The method according to claim 1, wherein, The first reminder device is obtained by randomly selecting one type after normalizing the expected benefits of various types of reminder devices under the current driving state.

3. The method according to claim 2, wherein, The step of normalizing the expected benefits of various types of reminder devices under the current driving state and then randomly selecting one includes: Based on the expected benefit data of the reminder devices corresponding to the driving state, determine the expected benefits of various types of reminder devices under the current driving state; The expected benefits of various types of alert devices under the current driving condition are combined into a one-dimensional vector; Normalize the one-dimensional vector; The normalized values ​​are used as the probability of using various types of reminder devices; and According to the above, one of several types of reminder devices is randomly selected as the first reminder device based on probability.

4. The method according to claim 3, wherein, The method further includes: The driver's driving status and the original expected revenue corresponding to the first reminder device are determined based on the expected revenue data of the reminder device corresponding to the driving status at the previous moment. Obtain the driver's actual benefit at the current moment; and, The expected revenue data of the reminder device is updated based on the original expected revenue and the actual revenue.

5. The method according to claim 4, wherein, The step of obtaining the driver's actual benefit at the current moment includes: The fixed income corresponding to the driving status at the current moment is determined based on the fixed income data of the reminder device corresponding to the driving status. The maximum expected benefit corresponding to the current driving state is determined based on the expected benefit data from the reminder device; and, The actual profit at the current moment is determined using the following formula: Q 实际 =reward+γ×max(Q(State k+1 )) Among them, Q 实际 The current actual reward is given, reward is the fixed reward, γ is the decay rate, and max(Q(State)) is the maximum return. k+1 The maximum expected return is denoted as ).

6. The method according to claim 5, wherein, The attenuation rate γ is determined through the following steps: Acquire key data under various driving conditions; The key data is input into a linear processing unit and compressed into a one-dimensional vector; Calculate the Euclidean distance between the vectors output by the linear processing unit, and take the reciprocal to obtain the connectivity matrix between different driving states; The connected matrix is ​​row-normalized, and the attenuation rate matrix is ​​output. as well as, The attenuation rate γ is determined based on the driving state at the previous moment, the driving state at the current moment, and the attenuation rate matrix.

7. The method according to claim 5 or 6, wherein, The process of updating the expected revenue data of the reminder device based on the original expected revenue and the actual revenue includes: The expected benefits of the updated driving status and alert device corresponding to the previous moment are calculated using the following formula: Q(State k Reminder device k )′=Q(State k Reminder device k )+α×(Q 实际 -Q(State k Reminder device k )) Where Q(State) k Reminder device k Q(State)' represents the driving state at the previous moment and the expected updated benefit corresponding to the warning device. k Reminder device k ) represents the original expected benefit corresponding to the driving status and the reminder device at the previous moment, and α is a coefficient, 0 < α < 1; Replace the original expected revenue in the expected revenue data of the reminder device with the updated expected revenue.

8. The method according to claim 7, wherein, The coefficient α is a fixed value or the coefficient α gradually decreases with time.

9. The method according to any one of claims 1-8, wherein, Before obtaining the driver's current driving status, the method further includes: Obtain the driver's identity information; and, The expected revenue data of the reminder device corresponding to the identity information is retrieved based on the identity information.

10. The method according to any one of claims 1-8, wherein, The method further includes: Obtain the current status of the vehicle; and, Determine the unavailable alert device corresponding to the current state; The expected benefits of various types of reminder devices in the current driving state, determined based on the expected benefit data of the reminder devices, do not include the expected benefits of the unavailable reminder devices in the current driving state.

11. The method according to any one of claims 1 to 10, wherein, The driving status is defined as normal driving, driver fatigue, driver smoking, or driver making a phone call.

12. The method according to any one of claims 1 to 11, wherein, Various types of alert devices include at least two of the following: fragrance devices, indicator lights, steering wheel, air conditioning, multimedia devices, windows, seats, and braking devices.

13. A computer-readable storage medium having stored thereon computer instructions, wherein, When executed by a processor, the computer instructions implement the steps of the method described in any one of claims 1-12.

14. An electronic device, wherein, include: Memory, on which computer instructions are stored; A processor for executing the computer instructions in the memory to implement the steps of the method according to any one of claims 1-12.

15. A vehicle, wherein, Includes the electronic device as described in claim 14.

Images