WO2021027410A1

WO2021027410A1 - Driving behavior detection method and apparatus, computer device, and storage medium

Info

Publication number: WO2021027410A1
Application number: PCT/CN2020/098827
Authority: WO
Inventors: 成潜; 张刘立
Original assignee: 平安科技（深圳）有限公司
Priority date: 2019-08-14
Filing date: 2020-06-29
Publication date: 2021-02-18
Also published as: CN110550035A; CN110550035B

Abstract

A driving behavior detection method and apparatus, a computer device, and a storage medium. The method comprises: obtaining first acceleration data, first angular acceleration data, and first azimuth angle data in a vehicle running process in real time; calculating an acceleration reference value, an angular acceleration reference value, and an azimuth angle reference value under each running state according to the first acceleration data, the first angular acceleration data, the first azimuth angle data, preset first characteristic value data, preset second characteristic value data, preset third characteristic value data, and a first algorithm; respectively comparing the acceleration reference value with a preset acceleration threshold, the angular acceleration reference value with a preset angular acceleration threshold, and the azimuth angle reference value with a preset azimuth angle threshold under the same running state; and determining whether a user currently conducts a dangerous driving behavior according to all the comparative results. The driving behavior detection method can detect whether a driver conducts a dangerous driving behavior.

Description

Driving behavior detection method, device, computer equipment and storage medium

This application claims the priority of a Chinese patent application filed with the Chinese Patent Office on August 14, 2019, the application number is 201910750554.6, and the invention title is "driving behavior detection method, device, computer equipment and storage medium", the entire content of which is incorporated by reference Incorporated in this application.

Technical field

This application relates to the technical field of traffic safety in smart cities, and in particular to a driving behavior detection method, device, computer equipment and storage medium.

Background technique

In recent years, with the development of the automobile industry and the improvement of people’s living standards, the number of residents’ car ownership has continued to increase, but subsequent traffic accidents have also occurred frequently, such as rear-end collisions and rollovers of vehicles, and these traffic accidents Most of the reasons behind are due to the dangerous driving behavior of drivers.

The inventor realizes that for drivers, drivers with dangerous driving behaviors often seldom realize that they have dangerous driving behaviors, which is likely to bring greater traffic safety hazards. Therefore, to avoid dangerous driving behaviors In the coming traffic safety hazards, it is particularly necessary to detect the driving behavior of drivers in real time.

technical problem

The main purpose of this application is to provide a driving behavior detection method, device, computer equipment, and storage medium, which are designed to detect whether drivers have dangerous driving behaviors during the driving of the vehicle, so as to avoid traffic caused by dangerous driving behaviors. Security risks.

Technical solutions

In the first aspect, this application proposes a driving behavior detection method applied to a mobile terminal, and the method includes:

Real-time acquisition of driving data during vehicle driving, where the driving data includes first acceleration data, first angular acceleration data, and first azimuth data;

According to the first acceleration data, the preset first characteristic value data corresponding to the first acceleration data in each driving state, and the preset first algorithm, the acceleration reference value in each driving state is calculated; and, according to the first angular acceleration data , The preset second characteristic value data and the first algorithm corresponding to the first angular acceleration data in each driving state, calculate the angular acceleration reference value in each driving state; and, according to the first azimuth angle data, the corresponding driving state The preset third characteristic value data of the first azimuth angle data and the first algorithm calculate the azimuth reference value in each driving state, where the driving state includes a sharp turn state, a sharp acceleration state, and a sudden braking state;

Compare the acceleration reference value with the preset acceleration threshold value, the angular acceleration reference value and the preset angular acceleration threshold value, and the azimuth angle reference value and the preset azimuth angle threshold value respectively in the same driving state;

If the acceleration reference value in the same driving state is greater than the preset acceleration threshold, and the angular acceleration reference value in the same driving state is greater than the preset angular acceleration threshold, and the azimuth reference value in the same driving state is greater than the preset azimuth threshold, then It is determined that the user currently has a dangerous driving behavior, where the dangerous driving behavior includes one or more of a sharp turn, a sharp acceleration, and a sharp brake.

In the second aspect, this application also proposes a driving behavior detection device applied to a mobile terminal, and the device includes:

The first acquisition module is configured to acquire driving data during the driving of the vehicle in real time, where the driving data includes first acceleration data, first angular acceleration data, and first azimuth data;

The first calculation module is configured to calculate the acceleration reference value in each driving state according to the first acceleration data, the preset first characteristic value data corresponding to the first acceleration data in each driving state, and the preset first algorithm; and Calculate the angular acceleration reference value in each driving state according to the first angular acceleration data, the preset second characteristic value data corresponding to the first angular acceleration data in each driving state, and the first algorithm; and, according to the first azimuth angle Data, the preset third characteristic value data corresponding to the first azimuth angle data in each driving state, and the first algorithm to calculate the azimuth reference value in each driving state. The driving state includes a sharp turn state, a sharp acceleration state, and Sudden braking state;

The comparison module is used to compare the acceleration reference value and the preset acceleration threshold value in the same driving state, the angular acceleration reference value and the preset angular acceleration threshold value, and the azimuth angle reference value and the preset azimuth angle threshold value respectively;

The determination module is used for when the acceleration reference value in the same driving state is greater than the preset acceleration threshold, and the angular acceleration reference value in the same driving state is greater than the preset angular acceleration threshold, and the azimuth angle reference value in the same driving state is greater than the preset When the azimuth angle threshold is set, it is determined that the user currently has a dangerous driving behavior, where the dangerous driving behavior includes one or more of a sharp turn, a sharp acceleration, and a sharp brake.

In a third aspect, this application also proposes a computer device, including a memory and a processor, the memory stores a computer program, and the processor implements a driving behavior detection method when the computer program is executed. The steps of the driving behavior detection method include:

Real-time acquisition of driving data during the driving of the vehicle, where the driving data includes first acceleration data, first angular acceleration data, and first azimuth angle data;

Calculating the acceleration reference value in each driving state according to the first acceleration data, the preset first characteristic value data corresponding to the first acceleration data in each driving state, and the preset first algorithm; and, According to the first angular acceleration data, the preset second characteristic value data corresponding to the first angular acceleration data in each of the driving states, and the first algorithm, the angular acceleration reference in each of the driving states is calculated Value; and, according to the first azimuth angle data, the preset third characteristic value data corresponding to the first azimuth angle data in each of the driving states, and the first algorithm, calculate each of the driving states The reference value of the azimuth angle of, wherein the driving state includes a sharp turn state, a sharp acceleration state and a sudden braking state;

Comparing the acceleration reference value with a preset acceleration threshold value, the angular acceleration reference value and the preset angular acceleration threshold value, and the azimuth angle reference value and the preset azimuth angle threshold value respectively in the same driving state;

If the acceleration reference value in the same driving state is greater than the preset acceleration threshold, and the angular acceleration reference value in the same driving state is greater than the preset angular acceleration threshold, and the azimuth angle reference value in the same driving state is greater than the preset acceleration threshold. Setting the azimuth angle threshold, it is determined that the user currently has a dangerous driving behavior, where the dangerous driving behavior includes one or more of a sharp turn, a sharp acceleration, and a sharp brake.

In a fourth aspect, the present application also proposes a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, a driving behavior detection method is implemented. The steps of the driving behavior detection method include:

Beneficial effect

The driving behavior detection method, device, computer equipment, and computer-readable storage medium provided by the embodiments of the present application obtain driving data of a vehicle during driving in real time through a mobile terminal, and then combine preset characteristic value data of the vehicle in different driving states Through calculation, the acceleration reference value, angular acceleration reference value and azimuth angle reference value corresponding to each driving state can be obtained, and then the acceleration reference value, angular acceleration reference value, and azimuth angle reference value obtained by calculation are respectively and the corresponding preset Set the acceleration threshold, the preset angular acceleration threshold, and the preset azimuth threshold to compare one by one. It can detect in real time whether the driver has any dangerous driving behavior during the driving of the vehicle, and then when a dangerous driving behavior occurs, the Voice prompts and other methods are used to remind users to drive safely, thereby avoiding traffic safety hazards caused by dangerous driving behaviors.

Description of the drawings

FIG. 1 is a schematic flowchart of a driving behavior detection method in an implementation of this application;

2 is a schematic diagram of the structure of a driving behavior detection device in an implementation of the present application;

Fig. 3 is a schematic diagram of the structure of a computer device in an implementation of the present application.

The best mode of the invention

1, an embodiment of the present application proposes a driving behavior detection method, which relates to the technical field of traffic safety in a smart city, and is applied to a mobile terminal. The method includes:

S11: Acquire driving data during the driving of the vehicle in real time, where the driving data includes first acceleration data, first angular acceleration data, and first azimuth data;

S12: Calculate the acceleration reference value in each driving state according to the first acceleration data, the preset first characteristic value data corresponding to the first acceleration data in each driving state, and the preset first algorithm; and, according to the first angle Acceleration data, preset second characteristic value data corresponding to the first angular acceleration data in each driving state, and the first algorithm to calculate the angular acceleration reference value in each driving state; and, according to the first azimuth data, each driving state Download the preset third characteristic value data corresponding to the first azimuth data and the first algorithm to calculate the azimuth reference value in each driving state, where the driving state includes a sharp turn state, a sharp acceleration state, and a sudden braking state;

S13, comparing the acceleration reference value with the preset acceleration threshold value, the angular acceleration reference value and the preset angular acceleration threshold value, and the azimuth angle reference value and the preset azimuth angle threshold value respectively in the same driving state;

S14, if the acceleration reference value in the same driving state is greater than the preset acceleration threshold, and the angular acceleration reference value in the same driving state is greater than the preset angular acceleration threshold, and the azimuth reference value in the same driving state is greater than the preset azimuth threshold , It is determined that the user currently has a dangerous driving behavior, where the dangerous driving behavior includes one or more of sharp turning, sharp acceleration, and sharp braking.

In the above S11, specifically, the mobile terminal may obtain the first acceleration data during the vehicle driving through the built-in linear acceleration sensor, obtain the first angular acceleration data during the vehicle travel through the built-in gyroscope, and use the built-in direction sensor Acquire the first azimuth angle data during the driving process of the vehicle, where the first acceleration data includes acceleration values in the three axial directions of X, Y, and Z, and the first angular acceleration data includes angular acceleration in the three axial directions of X, Y, and Z The first azimuth angle data includes pitch angle (ie pitch value), yaw angle (ie yaw value) and roll angle (ie roll value).

In the above S12, the function of the preset first feature value data is to modify the collected first acceleration data in combination with the first algorithm to obtain the acceleration reference value for driving behavior detection, the driving state, and the first acceleration data There is a one-to-one mapping relationship with the first eigenvalue data; in the same way, the function of the preset first eigenvalue data is to correct the collected first angular acceleration data in combination with the first algorithm to obtain The acceleration reference value of the driving behavior detection, the driving state, the first angular acceleration data and the first characteristic value data have a one-to-one mapping relationship; the function of the preset first characteristic value data is to combine the first algorithm to the collected data The first azimuth data is corrected to obtain an acceleration reference value used for driving behavior detection, and the driving state, the first azimuth data and the first characteristic value data have a one-to-one mapping relationship.

In the above S13 and S14, specifically, after obtaining the acceleration reference value, angular acceleration reference value, and azimuth reference value in each driving state, in order to detect whether the user currently has a dangerous driving behavior, the acceleration in the sharp turn state The reference value is compared with the preset acceleration threshold in the emergency turning state, the angular acceleration reference value in the sharp turning state is compared with the preset angular acceleration threshold in the emergency turning state, the acceleration reference value in the emergency braking state is compared with the emergency braking The preset acceleration threshold value of the state is compared, if the acceleration reference value in the sharp turn state is greater than the preset acceleration threshold value in the emergency turning state, and the angular acceleration reference value in the sharp turn state is greater than the preset angular acceleration threshold value in the emergency turning state , And the acceleration reference value in the emergency braking state is greater than the preset acceleration threshold for the emergency braking state, the mobile terminal can determine that the user currently has a dangerous driving behavior (that is, a dangerous driving behavior in a sharp turn); Use the known data of the emergency acceleration state or use the known data of the emergency braking state to determine whether the user currently has a dangerous driving behavior (such as whether there is a dangerous driving behavior of sudden acceleration or a dangerous driving behavior of sudden braking), its judgment The process is similar to the above-mentioned process of judging by using the known data of the emergency turning state, and will not be repeated here.

In this embodiment, the driving behavior detection method acquires the driving data of the vehicle in real time through the mobile terminal, and then combines the presets of the vehicle in different driving states (a sharp turn state, a sharp acceleration state, and a sudden braking state). The characteristic value data is calculated to obtain the acceleration reference value, angular acceleration reference value, and azimuth angle reference value corresponding to each driving state, and then the calculated acceleration reference value, angular acceleration reference value, and azimuth angle reference value are respectively and The corresponding preset acceleration threshold, preset angular acceleration threshold, and preset azimuth threshold are compared one by one, and it can detect in real time whether the driver has dangerous driving behavior during the driving of the vehicle, and then when dangerous driving behavior occurs, Voice prompts can be used to remind users to drive safely, thereby avoiding traffic safety hazards caused by dangerous driving behaviors.

In an optional embodiment, the foregoing step of acquiring driving data during the vehicle driving process in real time, wherein the driving data includes the first acceleration data, the first angular acceleration data, and the first azimuth angle data, includes:

S111, real-time collection of acceleration data, angular acceleration data, and azimuth angle data during vehicle driving;

S112, performing low-pass filtering processing on the acceleration data to obtain first acceleration data, and performing band-pass filtering processing on the angular acceleration data and the azimuth angle data respectively, to obtain the first angular acceleration data and the first azimuth angle data.

In this embodiment, by performing low-pass filtering processing on acceleration data collected in real time, and band-pass filtering processing on angular acceleration data and azimuth angle data collected in real time, abnormal signals in driving data can be processed to avoid abnormalities. The interference caused by the signal can improve the accuracy of data and the accuracy of driving behavior detection.

In an optional embodiment, the preset first characteristic value data includes a first characteristic value coefficient corresponding to an acceleration value of the X axis in a specific driving state, and a first characteristic value coefficient corresponding to an acceleration value of the Y axis in a specific driving state. The second eigenvalue coefficient of, the third eigenvalue coefficient corresponding to the acceleration value of the Z axis in a specific driving state, and the preset first characteristic value corresponding to the first acceleration data in each driving state according to the first acceleration data The steps of calculating the reference value of acceleration in each driving state by using the data and the preset first algorithm include:

Use the preset first calculation formula a _parameter = Aa _x1 +Ba _y1 +Ca _z1 to calculate the first acceleration data to obtain the first acceleration reference value for the emergency turning state and the second acceleration reference value for the emergency acceleration state And the third acceleration reference value for the emergency braking state, where in the first calculation formula, a _parameter is the acceleration reference value under a specific driving state, a _x1 is the acceleration value of the X axis in the first acceleration data, and a _y1 is The acceleration value of the Y axis in the first acceleration data, a ₂₁ is the acceleration value of the Z axis in the first acceleration data, A is the first characteristic value coefficient, B is the second characteristic value coefficient, and C is the third characteristic value coefficient.

In this embodiment, the above-mentioned specific driving state is a sharp turn state, or a rapid acceleration state, or a sudden braking state; when the specific driving state is a sharp turn state, the above-mentioned first calculation formula can be used to calculate the emergency turning state The acceleration reference value (ie, the first acceleration reference value). When the specific driving state is a rapid acceleration state, the acceleration reference value (ie, the second acceleration reference value) for the emergency acceleration state can be calculated using the above first calculation formula. When the specific driving state is the emergency braking state, the acceleration reference value for the emergency braking state (that is, the third acceleration reference value) can be calculated using the above-mentioned first calculation formula.

In some other embodiments, according to the first angular acceleration data, the preset second characteristic value data corresponding to the first angular acceleration data in each driving state, and the first algorithm, the angular acceleration reference value in each driving state is calculated. The calculation process, and the foregoing calculation process of calculating the reference value of the azimuth angle in each driving state based on the first azimuth angle data, the preset third characteristic value data corresponding to the first azimuth angle data in each driving state, and the first algorithm, Similar to the calculation of the acceleration reference value in each driving state by using the first calculation formula described above (that is, the relevant parameter in the first calculation formula is changed to the parameter corresponding to the angular acceleration data or the corresponding azimuth angle data), those skilled in the art Understandably, this will not be repeated here.

In an optional embodiment, before the step of real-time obtaining driving data during the driving of the vehicle, the method further includes:

S10a: Collect acceleration sample data in each driving state within a predetermined time period according to a predetermined frequency, where there are at least three sets of acceleration sample data in each driving state;

S10b, using the preset second calculation formula

Calculate each group of acceleration sample data in each driving state to obtain multiple sets of corresponding total acceleration values. In the second calculation formula, a is _always the total acceleration value in a specific driving state, and a _x is a specific driving The acceleration value of the X axis in the acceleration sample data in the state, a _y is the acceleration value of the Y axis in the acceleration sample data in the specific driving state, and a _z is the acceleration value of the Z axis in the acceleration sample data in the specific driving state;

S10c: Perform Fourier transform on each group of acceleration sample data in each driving state and multiple sets of corresponding total acceleration values to obtain multiple sets of first frequency domain data in each driving state and multiple sets of corresponding total acceleration values. Two frequency domain data;

S10d: Perform spectrum analysis on each group of first frequency domain data and each group of second frequency domain data to obtain multiple sets of first feature values corresponding to acceleration sample data in each driving state and multiple sets of second features corresponding to total acceleration values value;

S10e: Use the preset second calculation formula Ax+By+Cz=Max to calculate each group of first eigenvalues and each group of second eigenvalues to obtain the first eigenvalue coefficient, the second eigenvalue coefficient, and the first eigenvalue coefficient. Three eigenvalue coefficients, where in the second calculation formula, A is the first eigenvalue coefficient, B is the second eigenvalue coefficient, C is the third eigenvalue coefficient, and x is the acceleration of the X axis in a specific driving state The first characteristic value corresponding to the value, y is the first characteristic value corresponding to the acceleration value of the Y axis in a specific driving state, and z is the first characteristic value corresponding to the acceleration value of the Z axis in a specific driving state , Max is the second characteristic value corresponding to the total acceleration value in a specific driving state.

In the above S10a, before using the first acceleration data, first angular acceleration data, and first azimuth angle data collected in real time during the driving process of the vehicle for driving behavior detection, it can be in a sharp turn, a sharp acceleration, and a sharp brake. In the state, use the built-in acceleration sensor of the mobile terminal to collect acceleration sample data. The above-mentioned predetermined time can be 2 seconds, 3 seconds, 4 seconds, etc., and there is no specific restriction on this; the predetermined frequency can be 1 time per 0.5 second, 1 time per second There are no specific restrictions on this; for example, three sets of acceleration sample data of the vehicle in a sharp turn and three sets of acceleration samples in a sharp acceleration state can be collected at a frequency of 1 time per second within 3 seconds. Data and three sets of acceleration sample data in the sudden braking state.

In the above S10b, by way of example, the three sets of acceleration sample data in the sharp turn state are respectively substituted into the above second formula for calculation, and the three sets of total acceleration values in the sharp turn state can be obtained. The process of the total acceleration value in the braking state is similar to the foregoing, which can be understood by those skilled in the art and will not be repeated here.

In the above S10c, the specific representation form of the first frequency domain data and the second frequency domain data is a spectrogram. Illustratively, the acceleration sample data collected in a sharp turn state is used for description, and the acceleration sample in the emergency turn state The data includes the acceleration values of X, Y, and Z axes. Through the Fourier transform of the acceleration values of X, Y, and Z axes, the corresponding spectrograms of X, Y, and Z axes can be obtained. (I.e. the first frequency domain data), in the same way, by performing Fourier transform on the total acceleration value in a sharp turn, the spectrogram corresponding to the total acceleration value (ie the second frequency domain data) can be obtained; The relevant process of the Fourier transformation of the acceleration sample data and the total acceleration value in the state and in the sudden braking state is similar to the foregoing, and will not be repeated here.

In the above S10d, exemplarily, the first frequency domain data and the second frequency domain data in the sharp turning state are respectively analyzed by the frequency spectrum as an example. The frequency spectrum corresponding to the three axes of X, Y, and Z The graph performs spectrum analysis, and the eigenvalues corresponding to the acceleration values of the X, Y, and Z axes (ie the first eigenvalue) can be extracted, and the total acceleration can be extracted by performing spectrum analysis on the spectrogram corresponding to the total acceleration value The eigenvalue corresponding to the value (i.e., the second eigenvalue), where the above-mentioned first eigenvalue and the second eigenvalue refer to the maximum energy density (ie the peak value in the spectrogram). The related processes of performing spectrum analysis on the first frequency domain data and the second frequency domain data below are similar to the foregoing, and will not be repeated here.

In the above S10e, the specific driving state is a sharp turn state, or a sharp acceleration state, or a sudden braking state; when the specific driving state is a sharp turn state, the three sets of first characteristic values and the three sets of the emergency turning state The second eigenvalue is substituted into the above-mentioned first calculation formula to solve the equation group, and the first eigenvalue coefficient, the second eigenvalue coefficient and the third eigenvalue coefficient for the emergency turning state can be calculated; when the specific driving state is rapid acceleration In the state, the three sets of first eigenvalues and three sets of second eigenvalues for the emergency acceleration state are substituted into the above-mentioned first calculation formula to solve the equation set, and the first eigenvalue coefficients for the emergency acceleration state, The second characteristic value coefficient and the third characteristic value coefficient; when the specific driving state is a sudden braking state, the three sets of first characteristic values and three sets of second characteristic values for the emergency braking state are substituted into the above first calculation formula By solving the equations, the first eigenvalue coefficient, the second eigenvalue coefficient, and the third eigenvalue coefficient for the emergency braking state can be calculated. It should be noted here that in some other embodiments, the same as this embodiment can be used. Example of the same inventive concept to obtain characteristic value coefficients in other states (such as the following fourth characteristic value coefficient, fifth characteristic value coefficient and sixth characteristic value coefficient).

S101. Continuously collect multiple sets of motion data of a user according to a preset frequency within a specified time period, where the motion data includes second acceleration data, second angular acceleration data, and second azimuth data;

S102: Calculate a fourth acceleration reference value corresponding to the start state of the stroke according to the second acceleration data, the preset fourth characteristic value data corresponding to the second acceleration data in the stroke start state, and the preset second algorithm; and, according to The second angle acceleration data and the preset step counting algorithm calculate the number of exercise steps of the user; and, according to the second azimuth angle data, the preset fifth characteristic value data corresponding to the second azimuth angle data in the state of using the mobile terminal And the second algorithm to calculate the azimuth reference value corresponding to the state of the mobile terminal;

S103. Compare the fourth acceleration reference value with a preset acceleration threshold, compare the number of exercise steps with the preset step threshold, compare the azimuth reference value with the preset azimuth threshold, and compare The result of the comparison determines whether the user is currently driving;

If the user is currently in a driving state, the above S11 is executed to obtain driving data of the vehicle in real time, where the driving data includes first acceleration data, first angular acceleration data, and first azimuth data.

In the above S101, the specified time length can be 15 seconds, 16 seconds, 17 seconds, etc., and there is no specific limitation on this; specifically, the mobile terminal can use the built-in linear acceleration sensor, gyroscope, and direction sensor within the specified time period ( Such as 15 seconds) continuously collect multiple groups (such as 15 groups) of second acceleration data, second angular acceleration data, and second azimuth angle data of the user according to a preset frequency (such as 1 time per second), where the second acceleration data includes X, Y, and Z axis acceleration values, the second angular acceleration data includes the X, Y, Z axis angular acceleration values, the second azimuth angle data includes the pitch angle (ie pitch value), yaw angle (Ie yaw value) and roll angle (ie roll value).

In S102, specifically, the calculation process of the fourth acceleration reference value is as follows: The function of the preset fourth characteristic value data is to correct the collected second acceleration data in combination with the second algorithm to obtain Whether the user is in the fourth acceleration reference value at the beginning of the stroke, the function of the above-mentioned preset fifth characteristic value data is to correct the collected second azimuth angle data in combination with the second algorithm to obtain the data used to detect whether the user is in use or not The reference value of the azimuth angle of the terminal state, calculate the user's movement steps within the specified time based on the collected multiple sets of second angle acceleration data, that is, use the preset step counting algorithm to calculate the gyroscope data to calculate the movement steps Since it is a relatively mature existing technology, it will not be repeated here.

In the above S103, specifically, if the fourth acceleration reference value is greater than the preset acceleration threshold, the mobile terminal can determine that the user is in the starting state of driving according to this; if the number of exercise steps is greater than the preset threshold, move The terminal can determine that the user is walking; if the azimuth reference value is greater than the preset azimuth threshold, the mobile terminal can determine that the user is in the state of using the mobile terminal; generally, because the user is driving, it will The mobile terminal is placed in a specific position in the car, such as on the seat, fixed next to the driver's seat with a mobile phone holder, etc., that is, the mobile terminal is in the placed state rather than being operated by the user. Therefore, when the The group exercise data determines that the user is at the start of driving, is not walking, and is not in the state of using the mobile terminal, the mobile terminal can determine that the user is currently driving, that is, if the fourth acceleration reference value is greater than the preset When the acceleration threshold, the number of exercise steps is greater than the preset step threshold, and the azimuth reference value is greater than the preset azimuth threshold, the mobile terminal can determine that the user is currently driving, and then can make subsequent judgments Whether the user has operations related to dangerous driving behavior.

In an optional embodiment, the preset fourth characteristic value data includes a fourth characteristic value coefficient corresponding to the acceleration value of the X axis in the starting state of the stroke, and corresponding to the acceleration value of the Y axis in the starting state of the stroke. The fifth characteristic value coefficient of, the sixth characteristic value coefficient corresponding to the acceleration value of the Z axis in the start state of the stroke, and the preset fourth characteristic value corresponding to the second acceleration data in the start state of the stroke according to the second acceleration data The step of calculating the fourth acceleration reference value corresponding to the start state of the stroke by using the data and the preset second algorithm includes:

S102a, using a preset third calculation formula

Calculate the second acceleration data to obtain the fourth acceleration reference value, where, in the third calculation formula, the _{four parameters a} is the fourth acceleration reference value, a _x2 is the acceleration value of the X axis in the second acceleration data, a _y2 Is the acceleration value of the Y axis in the second acceleration data, a _z2 is the acceleration value of the Z axis in the second acceleration data, D is the fourth eigenvalue coefficient, E is the fifth eigenvalue coefficient, F is the sixth eigenvalue coefficient, n is the specified duration.

As described in S102a above, the mobile terminal uses the built-in linear acceleration sensor to continuously collect multiple sets of second acceleration data (such as 15 consecutive sets) at a preset frequency (such as 1 time per second) within a specified time period (such as 15 seconds). The acceleration value of the X axis, the acceleration value of the Y axis and the acceleration value of the Z axis) are substituted into the above-mentioned third calculation formula to solve the equation group, and then the fourth acceleration reference value can be calculated. In an optional embodiment, the driving data further includes GPS data, and the GPS data includes speed information, acceleration information, and azimuth angle information. After the above step of determining that the user currently has a dangerous driving behavior, it further includes:

S14A, if the user currently has a dangerous driving behavior of a sharp turn, determine whether the speed value in the speed information is greater than a preset speed threshold and whether the azimuth angle value in the azimuth angle information is greater than the preset azimuth angle threshold in a sharp turn;

If the user currently has a dangerous driving behavior of rapid acceleration, it is determined whether the total acceleration value of the acceleration information in the preset time period is greater than the preset acceleration threshold value in the rapid acceleration state;

If the user currently has a dangerous driving behavior of sudden braking, it is determined whether the total acceleration value of the acceleration information in the preset time period is greater than the preset acceleration threshold in the sudden braking state;

S14B: Determine whether the user currently has dangerous driving behavior and the type of dangerous driving behavior according to the judgment result.

In this embodiment, specifically, when the user is driving, the mobile terminal can collect GPS data during the driving process of the vehicle in real time by turning on the built-in GPS sensor. When it is predicted that the user is currently in danger by the method in S13 above When driving behavior, the data collected by the GPS sensor is used for further confirmation. Since the data collected by the GPS sensor is more accurate than the data collected by the acceleration sensor, gyroscope, and direction sensor, when using the acceleration sensor, When the data collected by the gyroscope and the direction sensor predicts that the user currently has dangerous driving behavior, the GPS data can be used for further confirmation, which can improve the accuracy of driving behavior detection;

Specifically, if the speed value in the speed information is greater than the preset speed threshold and the azimuth angle value in the azimuth information is greater than the preset azimuth threshold for the emergency turning state, the mobile terminal can determine that the user currently has a dangerous driving behavior and The type of the dangerous driving behavior is a sharp turn; if the total acceleration value of the acceleration information in the preset time period is greater than the preset acceleration threshold for the emergency acceleration state, the mobile terminal can determine that the user currently has a dangerous driving behavior and the dangerous The type of driving behavior is rapid acceleration; if the total acceleration value of the acceleration information in the preset time period is greater than the preset acceleration threshold for the emergency braking state, the mobile terminal can determine that the user currently has a dangerous driving behavior and the dangerous driving behavior The type is sudden braking.

In an optional embodiment, after the step of determining whether the user currently has dangerous driving behavior and the type of dangerous driving behavior according to the judgment result, the method further includes:

S15, when the user currently has a dangerous driving behavior, obtain current location information and time information;

S16: Associate the type of dangerous driving behavior with location information and time information to generate associated information.

In this embodiment, when the user has a dangerous driving behavior, the type of the dangerous driving behavior is associated with the current location information and time information to generate the associated information, so that the user can see clearly by viewing the associated information after driving. Knowing what kind of dangerous driving behaviors have occurred when and where, so that users can conduct serious self-reflection learning after driving to avoid the recurrence of dangerous driving behaviors.

In an optional embodiment, after the step of associating the type of dangerous driving behavior with location information and time information, and generating the associated information, the method further includes:

S17: Determine whether the user is currently in a parking state according to the first angular acceleration data;

If the user is currently in a parking state, execute S18 to generate a driving track record chart, and mark the associated information on the corresponding position in the driving track record chart.

In this embodiment, specifically, the first angular acceleration data during the driving process of the vehicle can be collected in real time through the built-in gyroscope of the mobile terminal, and then the number of movement steps can be calculated according to the first angular acceleration data, and then it can be judged whether the number of movement steps exceeds The preset number of steps threshold. If it is, it can be judged that the user is walking, so that it can be judged that the user is currently in a parking state. Since the GPS sensor can record the trajectory of the vehicle during driving, when it is judged that the user is currently In the parking state, the driving track record chart can be generated according to the driving track recorded by the GPS sensor, and the associated information can be marked on the corresponding position in the driving track record chart, so that the user can view by viewing after driving. The driving trajectory record chart can be more intuitive and clear to know when and where what dangerous driving behaviors have occurred, so that users can conduct serious self-reflection learning after driving to avoid the recurrence of dangerous driving behaviors .

In a preferred embodiment, after the above step of collecting acceleration sample data in each of the driving states within a predetermined time period according to a predetermined frequency, the method further includes:

S10a1. Input the acceleration sample data, angular acceleration sample data and azimuth angle sample data obtained in each driving state into the preset RBF neural network model for data processing, and output the above-mentioned preset acceleration thresholds in each driving state, The above-mentioned preset angular acceleration threshold and the above-mentioned preset azimuth angle threshold.

In this embodiment, the above-mentioned RBF neural network model is a model that is trained to convergence through training samples in advance. The training process is described as follows: First, obtain training samples, where the training samples include input data and output data, and the input data is Acceleration sample training data, angular acceleration sample training data, and azimuth angle sample training data collected in a sharp turn, rapid acceleration, and sudden braking. The output data is the acceleration threshold and corresponding angle of the corresponding acceleration data in each state The angular acceleration threshold of the acceleration data and the azimuth threshold of the corresponding azimuth angle data. After that, the RBF neural network model is trained through the training samples to obtain the RBF neural network model with better network parameters. Therefore, in the process of using the vehicle to drive Before the first acceleration data, first angular acceleration data, and direction angle data collected in real time are used for driving behavior detection, the built-in acceleration sensor, gyroscope and direction sensor of the mobile terminal can be used in the sharp turn state to collect the sharp turn state. Acceleration sample data, angular acceleration sample data and azimuth angle sample data, similarly, the acceleration sensor, gyroscope and direction sensor built in the mobile terminal can be used to collect acceleration sample data and angular acceleration sample in the rapid acceleration state. Data and azimuth sample data, similarly, the acceleration sensor, gyroscope and direction sensor built in the mobile terminal can be used to collect acceleration sample data, angular acceleration sample data and azimuth sample data in the sudden braking state, respectively. Then the acceleration sample data, angular acceleration sample data, and azimuth angle sample data collected in each state are input into the preset RBF neural network model, and the preset acceleration threshold and preset angle for the emergency turning state in S13 can be obtained. Acceleration threshold, preset azimuth threshold, preset acceleration threshold, preset angular acceleration threshold, preset azimuth threshold for emergency acceleration state, preset acceleration threshold, preset angular acceleration threshold, preset azimuth for emergency braking state The angle threshold, in the same way, the acceleration threshold and the azimuth angle threshold in S103 can be obtained.

This application also proposes a driving behavior detection device applied to a mobile terminal, and the device includes:

The first acquisition module 11 is configured to acquire driving data during the driving process of the vehicle in real time, where the driving data includes first acceleration data, first angular acceleration data, and first azimuth data;

The first calculation module 12 is configured to calculate the acceleration reference value in each driving state according to the first acceleration data, the preset first characteristic value data corresponding to the first acceleration data in each driving state, and the preset first algorithm; And, according to the first angular acceleration data, the preset second characteristic value data corresponding to the first angular acceleration data in each driving state, and the first algorithm, the angular acceleration reference value in each driving state is calculated; and, according to the first orientation Angle data, preset third characteristic value data corresponding to the first azimuth angle data in each driving state and the first algorithm to calculate the reference value of the azimuth angle in each driving state, where the driving state includes a sharp turn state and a sharp acceleration state And sudden braking;

The comparison module 13 is configured to compare the acceleration reference value and the preset acceleration threshold value in the same driving state, the angular acceleration reference value and the preset angular acceleration threshold value, and the azimuth angle reference value and the preset azimuth angle threshold value respectively;

The determination module 14 is used for when the acceleration reference value in the same driving state is greater than the preset acceleration threshold, and the angular acceleration reference value in the same driving state is greater than the preset angular acceleration threshold, and the azimuth angle reference value in the same driving state is greater than the preset acceleration threshold. When the azimuth angle threshold is set, it is determined that the user currently has a dangerous driving behavior, where the dangerous driving behavior includes one or more of a sharp turn, a sharp acceleration, and a sharp brake.

In an optional embodiment, the aforementioned first obtaining module 11 includes:

Acquisition unit for real-time acquisition of acceleration data, angular acceleration data and azimuth angle data during vehicle driving;

The noise reduction processing unit is used to perform low-pass filtering processing on the acceleration data to obtain the first acceleration data, and to perform band-pass filtering processing on the angular acceleration data and the azimuth angle data respectively to obtain the first angular acceleration data and the first azimuth angle data.

In an optional embodiment, the preset first characteristic value data includes a first characteristic value coefficient corresponding to an acceleration value of the X axis in a specific driving state, and a first characteristic value coefficient corresponding to an acceleration value of the Y axis in a specific driving state. The second eigenvalue coefficient of, and the third eigenvalue coefficient corresponding to the acceleration value of the Z axis in a specific driving state. The above-mentioned first calculation module 12 includes:

The first calculation unit is configured to use the preset first calculation formula a _parameter = Aa _x1 +Ba _y1 +Ca _z1 to calculate the first acceleration data to obtain the first acceleration reference value for the emergency turning state and for the emergency acceleration The second acceleration reference value of the state and the third acceleration reference value of the emergency braking state, where, in the first calculation formula, a _parameter is the acceleration reference value in a specific driving state, and a _x1 is the X axis in the first acceleration data A _y1 is the acceleration value of the Y axis in the first acceleration data, a _z1 is the acceleration value of the Z axis in the first acceleration data, A is the first eigenvalue coefficient, B is the second eigenvalue coefficient, and C is The third eigenvalue coefficient.

In an optional embodiment, the above driving behavior detection device further includes:

The first collection module is configured to collect acceleration sample data in each driving state within a predetermined time period according to a predetermined frequency, wherein there are at least three sets of acceleration sample data in each driving state;

The second calculation module is used to use the preset second calculation formula

The data transformation module is used to perform Fourier transform on each set of acceleration sample data and multiple sets of corresponding total acceleration values in each driving state to obtain multiple sets of first frequency domain data and multiple sets of corresponding total acceleration values in each driving state. The second frequency domain data of the acceleration value;

The spectrum analysis module is used to perform spectrum analysis on each group of first frequency domain data and each group of second frequency domain data to obtain multiple groups of first characteristic values corresponding to acceleration sample data in each driving state and multiple groups of corresponding total acceleration values The second characteristic value;

The third calculation module is used to use the preset second calculation formula Ax+By+Cz=Max to calculate each group of first characteristic values and each group of second characteristic values to obtain the first characteristic value coefficients and the second characteristic values. The characteristic value coefficient and the third characteristic value coefficient, where in the second calculation formula, A is the first characteristic value coefficient, B is the second characteristic value coefficient, C is the third characteristic value coefficient, and x is in a specific driving state The first characteristic value corresponding to the acceleration value of the X axis, y is the first characteristic value corresponding to the acceleration value of the Y axis in a specific driving state, and z is the acceleration value corresponding to the Z axis in a specific driving state Max is the second characteristic value corresponding to the resultant acceleration value in a specific driving state.

The second collection module is configured to continuously collect multiple sets of motion data of the user according to a preset frequency within a specified time period, where the motion data includes second acceleration data, second angular acceleration data, and second azimuth angle data;

The fourth calculation module is used to calculate the fourth acceleration reference corresponding to the stroke start state according to the second acceleration data, the preset fourth characteristic value data corresponding to the second acceleration data in the stroke start state, and the preset second algorithm And, according to the second angle acceleration data and the preset step counting algorithm, calculate the user's movement steps; and, according to the second azimuth data, the preset corresponding to the second azimuth data in the state of using the mobile terminal The fifth characteristic value data and the second algorithm calculate the azimuth reference value corresponding to the state of the mobile terminal;

The comparison module is used to compare the fourth acceleration reference value with the preset acceleration threshold, the number of motion steps to compare with the preset step threshold, and the azimuth reference value to compare with the preset azimuth threshold. Compare, and judge whether the user is currently driving according to the result of the comparison;

The above-mentioned first acquisition module 11 is specifically configured to acquire driving data of the vehicle in real time when the user is currently driving, wherein the driving data includes first acceleration data, first angular acceleration data, and first azimuth angle data.

In an optional embodiment, the preset fourth characteristic value data includes a fourth characteristic value coefficient corresponding to the acceleration value of the X axis in the starting state of the stroke, and corresponding to the acceleration value of the Y axis in the starting state of the stroke. The fifth eigenvalue coefficient of, the sixth eigenvalue coefficient corresponding to the acceleration value of the Z axis in the stroke start state, the fourth calculation module mentioned above, includes:

The second calculation unit is used to use a preset third calculation formula

In an optional embodiment, the driving data further includes GPS data, and the GPS data includes speed information, acceleration information, and azimuth angle information. The above-mentioned driving behavior detection device further includes:

The first judgment module is used for judging whether the speed value in the speed information is greater than the preset speed threshold and the azimuth angle value in the azimuth angle information is greater than the predicted value in the sharp turn when the user currently has a dangerous driving behavior of a sharp turn. Set the azimuth threshold; it is also used when the user currently has a dangerous driving behavior of rapid acceleration, to determine whether the total acceleration value of the acceleration information in the preset time period is greater than the preset acceleration threshold in the rapid acceleration state; also used when When the user currently has a dangerous driving behavior of sudden braking, it is judged whether the total acceleration value of the acceleration information in the preset time period is greater than the preset acceleration threshold in the sudden braking state;

The determination module is used to determine whether the user currently has dangerous driving behavior and the type of dangerous driving behavior according to the judgment result.

The second acquisition module is used to acquire current location information and time information when the user currently has dangerous driving behavior;

The association module is used to associate the type of dangerous driving behavior with location information and time information to generate associated information.

The second judgment module is used to judge whether the user is currently in a parking state according to the first angular acceleration data;

The generating module is used to generate a driving track record chart when the user is currently in a parking state, and mark the associated information on the corresponding position in the driving track record chart.

In a preferred embodiment, the above-mentioned driving behavior detection device further includes:

The data processing module is used to input the acceleration sample data, angular acceleration sample data and azimuth sample data obtained in each driving state into the preset RBF neural network model for data processing, and output the above predictions in each driving state Set the acceleration threshold, the above-mentioned preset angular acceleration threshold and the above-mentioned preset azimuth angle threshold.

It should be noted that for the foregoing device embodiment, since it basically corresponds to the foregoing method embodiment, the specific implementation mode can be referred to the description of the method embodiment section, which can be understood by those skilled in the art and will not be repeated here. .

3, an embodiment of the present application also provides a computer device. The computer device may be a server, and its internal structure may be as shown in FIG. 3. The computer equipment includes a processor, a memory, a network interface and a database connected through a system bus. Among them, the computer designed processor is used to provide calculation and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The computer equipment database is used to store driving behavior detection methods and programs. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the previous driving behavior detection method is realized, and the steps of the driving behavior detection method include:

Calculate the acceleration reference value in each driving state according to the first acceleration data, the preset first characteristic value data corresponding to the first acceleration data in each driving state, and the preset first algorithm; and, According to the first angular acceleration data, the preset second characteristic value data corresponding to the first angular acceleration data in each of the driving states, and the first algorithm, the angular acceleration reference in each of the driving states is calculated Value; and, according to the first azimuth angle data, the preset third characteristic value data corresponding to the first azimuth angle data in each of the driving states, and the first algorithm, the calculation of each of the driving states The reference value of the azimuth angle of, wherein the driving state includes a sharp turn state, a sharp acceleration state, and a sharp braking state;

The embodiment of the present application also proposes a computer-readable storage medium. The storage medium is a volatile storage medium or a non-volatile storage medium, and a computer program is stored thereon. When the computer program is executed by a processor, a driving In a behavior detection method, the steps of the driving behavior detection method include:

Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by computer programs instructing relevant hardware. The computer programs can be stored and a non-volatile computer readable storage In the medium, when the computer program is executed, it may include the procedures of the above-mentioned method embodiments. Wherein, any reference to memory, storage, database or other media provided in this application and used in the embodiments may include non-volatile and/or volatile memory. Non-volatile memory may include read only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not a limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual-rate data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous Link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

The above are only the preferred embodiments of this application, and do not limit the scope of this application. Any equivalent structure or equivalent process transformation made using the content of this application description and drawings, or directly or indirectly applied to other related The technical field is equally included in the scope of patent protection of this application.

Claims

A driving behavior detection method, wherein, applied to a mobile terminal, the method includes:

Real-time acquisition of driving data during the driving of the vehicle, where the driving data includes first acceleration data, first angular acceleration data, and first azimuth angle data;

Calculate the acceleration reference value in each driving state according to the first acceleration data, the preset first characteristic value data corresponding to the first acceleration data in each driving state, and the preset first algorithm; and, According to the first angular acceleration data, the preset second characteristic value data corresponding to the first angular acceleration data in each of the driving states, and the first algorithm, the angular acceleration reference in each of the driving states is calculated Value; and, according to the first azimuth angle data, the preset third characteristic value data corresponding to the first azimuth angle data in each of the driving states, and the first algorithm, the calculation of each of the driving states The reference value of the azimuth angle of, wherein the driving state includes a sharp turn state, a sharp acceleration state, and a sharp braking state;

Comparing the acceleration reference value with a preset acceleration threshold value, the angular acceleration reference value and the preset angular acceleration threshold value, and the azimuth angle reference value and the preset azimuth angle threshold value respectively in the same driving state;

If the acceleration reference value in the same driving state is greater than the preset acceleration threshold, and the angular acceleration reference value in the same driving state is greater than the preset angular acceleration threshold, and the azimuth angle reference value in the same driving state is greater than the preset acceleration threshold. Setting the azimuth angle threshold, it is determined that the user currently has a dangerous driving behavior, where the dangerous driving behavior includes one or more of a sharp turn, a sharp acceleration, and a sharp brake.
The driving behavior detection method according to claim 1, wherein the preset first characteristic value data includes a first characteristic value coefficient corresponding to an acceleration value of the X axis in a specific driving state, and a first characteristic value coefficient corresponding to an acceleration value of the X axis in a specific driving state. The second characteristic value coefficient corresponding to the acceleration value of the Y axis, and the third characteristic value coefficient corresponding to the acceleration value of the Z axis in a specific driving state, according to the first acceleration data, corresponding to each driving state The step of calculating the reference value of acceleration in each of the driving states by the preset first characteristic value data of the first acceleration data and the preset first algorithm includes:

Use the preset first calculation formula a parameter = Aa x1 +Ba y1 +Ca z1 to calculate the first acceleration data to obtain the first acceleration reference value corresponding to the sharp turning state and corresponding to the sharp acceleration state The second acceleration reference value corresponding to the sudden braking state and the third acceleration reference value corresponding to the sudden braking state, wherein, in the first calculation formula, a parameter is the acceleration reference value in a specific driving state, and a x1 is the first The acceleration value of the X axis in the acceleration data, a y1 is the acceleration value of the Y axis in the first acceleration data, a z1 is the acceleration value of the Z axis in the first acceleration data, and A is the first characteristic value coefficient , B is the second characteristic value coefficient, and C is the third characteristic value coefficient.
The driving behavior detection method according to claim 2, wherein, before the step of obtaining driving data during the driving of the vehicle in real time, the method further comprises:

Collecting acceleration sample data in each of the driving states for a predetermined time period according to a predetermined frequency, wherein there are at least three sets of acceleration sample data in each of the driving states;

Use the preset second calculation formula
Calculate each group of the acceleration sample data in each of the driving states to obtain multiple sets of corresponding total acceleration values, where in the second calculation formula, a is always the total acceleration value in a specific driving state , A x is the acceleration value of the X axis in the acceleration sample data in a specific driving state, a y is the acceleration value of the Y axis in the acceleration sample data in a specific driving state, and a z is the acceleration sample in the specific driving state The acceleration value of the Z axis in the data;

Fourier transform is performed on each group of said acceleration sample data and multiple sets of corresponding total acceleration values in each said driving state, respectively, to obtain multiple sets of first frequency domain data and multiple sets of each said driving state Second frequency domain data corresponding to the total acceleration value;

Perform spectrum analysis on each group of the first frequency domain data and each group of the second frequency domain data, and obtain multiple sets of first feature values corresponding to the acceleration sample data in each of the driving conditions and multiple sets of corresponding data. The second characteristic value of the total acceleration value;

Using the preset second calculation formula Ax+By+Cz=Max, each group of the first characteristic value and each group of the second characteristic value are respectively calculated to obtain the first characteristic value coefficient and the first characteristic value coefficient. Two eigenvalue coefficients and the third eigenvalue coefficient, wherein, in the second calculation formula, A is the first eigenvalue coefficient, B is the second eigenvalue coefficient, and C is the third The characteristic value coefficient, where x is the first characteristic value corresponding to the acceleration value of the X axis in a specific driving state, and y is the first characteristic value corresponding to the acceleration value of the Y axis in a specific driving state, z is the first characteristic value corresponding to the acceleration value of the Z axis in a specific driving state, and Max is the second characteristic value corresponding to the resultant acceleration value in a specific driving state.
The driving behavior detection method according to claim 1, wherein, before the step of acquiring driving data in the vehicle driving process in real time, the method further comprises:

Continuously collecting multiple sets of motion data of the user according to a preset frequency within a specified time period, where the motion data includes second acceleration data, second angular acceleration data, and second azimuth angle data;

According to the second acceleration data, the preset fourth characteristic value data corresponding to the second acceleration data in the stroke start state, and the preset second algorithm, the fourth acceleration reference value corresponding to the stroke start state is calculated And, calculating the number of exercise steps of the user according to the second angular acceleration data and a preset step-counting algorithm; and, according to the second azimuth data, corresponding to the first in the state of using the mobile terminal The preset fifth characteristic value data of the two-azimuth angle data and the second algorithm calculate the azimuth angle reference value corresponding to the state of the mobile terminal in use;

The fourth acceleration reference value is compared with a preset acceleration threshold value, the number of exercise steps is compared with a preset step threshold value, and the azimuth angle reference value is compared with the preset azimuth angle threshold value. Compare, and determine whether the user is currently driving according to the result of the comparison;

If the user is currently in a driving state, the step of acquiring driving data during the driving of the vehicle in real time is executed.
The driving behavior detection method according to claim 4, wherein the preset fourth characteristic value data includes a fourth characteristic value coefficient corresponding to the acceleration value of the X axis in the starting state of the stroke, and The fifth characteristic value coefficient corresponding to the acceleration value of the Y axis in the starting state, and the sixth characteristic value coefficient corresponding to the acceleration value of the Z axis in the starting state of the stroke, according to the second acceleration data, The step of calculating the fourth acceleration reference value corresponding to the start state of the stroke corresponding to the preset fourth characteristic value data of the second acceleration data and the preset second algorithm in the stroke start state includes:

Use the preset third calculation formula
Calculate the second acceleration data to obtain the fourth acceleration reference value, where in the third calculation formula, a four-parameter is the fourth acceleration reference value, and a x2 is the second acceleration The acceleration value of the X axis in the data, a y2 is the acceleration value of the Y axis in the second acceleration data, a z2 is the acceleration value of the Z axis in the second acceleration data, D is the fourth characteristic value coefficient, E is the fifth characteristic value coefficient, F is the sixth characteristic value coefficient, and n is the specified duration.
The driving behavior detection method according to any one of claims 1 to 5, wherein the driving data further includes GPS data, and the GPS data includes speed information, acceleration information, and azimuth information; the determining that the user currently exists After the dangerous driving behavior steps, it also includes:

If the user currently has a dangerous driving behavior of a sharp turn, it is determined whether the speed value in the speed information is greater than a preset speed threshold and whether the azimuth angle value in the azimuth angle information is greater than the azimuth angle value in the sharp turn. Preset azimuth angle threshold;

If the user currently has a dangerous driving behavior of rapid acceleration, determining whether the total acceleration value of the acceleration information in a preset time period is greater than the preset acceleration threshold in the rapid acceleration state;

If the user currently has a dangerous driving behavior of sudden braking, determining whether the total acceleration value of the acceleration information within a preset period of time is greater than the preset acceleration threshold in the sudden braking state;

According to the judgment result, it is determined whether the user currently has the dangerous driving behavior and the type of the dangerous driving behavior.
The driving behavior detection method according to claim 6, wherein after the step of determining whether the user currently has the dangerous driving behavior and the type of the dangerous driving behavior according to the judgment result, the method further comprises:

When the user currently has the dangerous driving behavior, acquiring current location information and time information;

Associating the type of the dangerous driving behavior with the location information and time information to generate associated information.
A driving behavior detection device, wherein, applied to a mobile terminal, the device includes:

The first acquisition module is configured to acquire driving data during the driving of the vehicle in real time, wherein the driving data includes first acceleration data, first angular acceleration data, and first azimuth data;

The first calculation module is configured to calculate the first acceleration data in each driving state, the preset first characteristic value data corresponding to the first acceleration data in each driving state, and the preset first algorithm And, according to the first angular acceleration data, the preset second characteristic value data corresponding to the first angular acceleration data in each of the driving states, and the first algorithm, calculating each of the The angular acceleration reference value in the driving state; and, calculating according to the first azimuth angle data, the preset third characteristic value data corresponding to the first azimuth angle data in each of the driving states, and the first algorithm The azimuth reference value in each of the driving states, where the driving state includes a sharp turn state, a sharp acceleration state, and a sudden braking state;

The comparison module is configured to compare the acceleration reference value with a preset acceleration threshold under the same driving state, compare the angular acceleration reference value with the preset angular acceleration threshold, and compare the azimuth reference value with the preset azimuth Angle threshold for comparison;

The determination module is configured to: when the acceleration reference value in the same driving state is greater than a preset acceleration threshold, and the angular acceleration reference value in the same driving state is greater than the preset angular acceleration threshold, and the orientation in the same driving state When the angle reference value is greater than the preset azimuth angle threshold, it is determined that the user currently has a dangerous driving behavior, where the dangerous driving behavior includes one or more of a sharp turn, a sharp acceleration, and a sharp brake.
The driving behavior detection device according to claim 8, wherein the first acquisition module comprises:

Acquisition unit for real-time acquisition of acceleration data, angular acceleration data and azimuth angle data during vehicle driving;

The noise reduction processing unit is used to perform low-pass filtering processing on the acceleration data to obtain the first acceleration data, and to perform band-pass filtering processing on the angular acceleration data and the azimuth angle data respectively to obtain the first angular acceleration data and the first azimuth angle data.
The driving behavior detection device according to claim 7, further comprising:

The first collection module is configured to collect acceleration sample data in each driving state within a predetermined time period according to a predetermined frequency, wherein there are at least three sets of acceleration sample data in each driving state;

The second calculation module is used to use the preset second calculation formula
Calculate each group of acceleration sample data in each driving state to obtain multiple sets of corresponding total acceleration values. In the second calculation formula, a is always the total acceleration value in a specific driving state, and a x is a specific driving The acceleration value of the X axis in the acceleration sample data in the state, a y is the acceleration value of the Y axis in the acceleration sample data in the specific driving state, and a z is the acceleration value of the Z axis in the acceleration sample data in the specific driving state;

The data transformation module is used to perform Fourier transform on each set of acceleration sample data and multiple sets of corresponding total acceleration values in each driving state to obtain multiple sets of first frequency domain data and multiple sets of corresponding total acceleration values in each driving state. The second frequency domain data of the acceleration value;

The spectrum analysis module is used to perform spectrum analysis on each group of first frequency domain data and each group of second frequency domain data to obtain multiple groups of first characteristic values corresponding to acceleration sample data in each driving state and multiple groups of corresponding total acceleration values The second characteristic value;

The third calculation module is used to use the preset second calculation formula Ax+By+Cz=Max to calculate each group of first characteristic values and each group of second characteristic values to obtain the first characteristic value coefficients and the second characteristic values. The characteristic value coefficient and the third characteristic value coefficient, where in the second calculation formula, A is the first characteristic value coefficient, B is the second characteristic value coefficient, C is the third characteristic value coefficient, and x is in a specific driving state The first characteristic value corresponding to the acceleration value of the X axis, y is the first characteristic value corresponding to the acceleration value of the Y axis in a specific driving state, and z is the acceleration value corresponding to the Z axis in a specific driving state Max is the second characteristic value corresponding to the resultant acceleration value in a specific driving state.
A computer device includes a memory and a processor, the memory stores a computer program, wherein the processor implements a driving behavior detection method when the computer program is executed, and the steps of the driving behavior detection method include:

Real-time acquisition of driving data during the driving of the vehicle, where the driving data includes first acceleration data, first angular acceleration data, and first azimuth angle data;

Calculate the acceleration reference value in each driving state according to the first acceleration data, the preset first characteristic value data corresponding to the first acceleration data in each driving state, and the preset first algorithm; and, According to the first angular acceleration data, the preset second characteristic value data corresponding to the first angular acceleration data in each of the driving states, and the first algorithm, the angular acceleration reference in each of the driving states is calculated Value; and, according to the first azimuth angle data, the preset third characteristic value data corresponding to the first azimuth angle data in each of the driving states, and the first algorithm, the calculation of each of the driving states The reference value of the azimuth angle of, wherein the driving state includes a sharp turn state, a sharp acceleration state, and a sharp braking state;

Comparing the acceleration reference value with a preset acceleration threshold value, the angular acceleration reference value and the preset angular acceleration threshold value, and the azimuth angle reference value and the preset azimuth angle threshold value respectively in the same driving state;

If the acceleration reference value in the same driving state is greater than the preset acceleration threshold, and the angular acceleration reference value in the same driving state is greater than the preset angular acceleration threshold, and the azimuth angle reference value in the same driving state is greater than the preset acceleration threshold. Setting the azimuth angle threshold, it is determined that the user currently has a dangerous driving behavior, where the dangerous driving behavior includes one or more of a sharp turn, a sharp acceleration, and a sharp brake.
The computer device according to claim 11, wherein the preset first characteristic value data includes a first characteristic value coefficient corresponding to an acceleration value of the X axis in a specific driving state, and a first characteristic value coefficient corresponding to an acceleration value of the Y axis in a specific driving state. The second characteristic value coefficient corresponding to the acceleration value of the vehicle, the third characteristic value coefficient corresponding to the acceleration value of the Z axis in a specific driving state, and the first acceleration data corresponding to the first characteristic value coefficient in each driving state according to the first acceleration data The step of calculating the reference value of acceleration in each of said driving states with preset first characteristic value data of acceleration data and a preset first algorithm includes:

Use the preset first calculation formula a parameter = Aa x1 +Ba y1 +Ca z1 to calculate the first acceleration data to obtain the first acceleration reference value corresponding to the sharp turning state and corresponding to the sharp acceleration state The second acceleration reference value corresponding to the sudden braking state and the third acceleration reference value corresponding to the sudden braking state, wherein, in the first calculation formula, a parameter is the acceleration reference value in a specific driving state, and a x1 is the first The acceleration value of the X axis in the acceleration data, a y1 is the acceleration value of the Y axis in the first acceleration data, a z1 is the acceleration value of the Z axis in the first acceleration data, and A is the first characteristic value coefficient , B is the second characteristic value coefficient, and C is the third characteristic value coefficient.
The computer device according to claim 12, wherein, before the step of obtaining driving data during the driving of the vehicle in real time, it further comprises:

Collecting acceleration sample data in each of the driving states for a predetermined time period according to a predetermined frequency, wherein there are at least three sets of acceleration sample data in each of the driving states;

Use the preset second calculation formula
Calculate each group of the acceleration sample data in each of the driving states to obtain multiple sets of corresponding total acceleration values, where in the second calculation formula, a is always the total acceleration value in a specific driving state , A x is the acceleration value of the X axis in the acceleration sample data in a specific driving state, a y is the acceleration value of the Y axis in the acceleration sample data in a specific driving state, and a z is the acceleration sample in the specific driving state The acceleration value of the Z axis in the data;

Fourier transform is performed on each group of said acceleration sample data and multiple sets of corresponding total acceleration values in each said driving state, respectively, to obtain multiple sets of first frequency domain data and multiple sets of each said driving state Second frequency domain data corresponding to the total acceleration value;

Perform spectrum analysis on each group of the first frequency domain data and each group of the second frequency domain data, and obtain multiple sets of first feature values corresponding to the acceleration sample data in each of the driving conditions and multiple sets of corresponding data. The second characteristic value of the total acceleration value;

Using the preset second calculation formula Ax+By+Cz=Max, each group of the first characteristic value and each group of the second characteristic value are respectively calculated to obtain the first characteristic value coefficient and the first characteristic value coefficient. Two eigenvalue coefficients and the third eigenvalue coefficient, wherein, in the second calculation formula, A is the first eigenvalue coefficient, B is the second eigenvalue coefficient, and C is the third The characteristic value coefficient, where x is the first characteristic value corresponding to the acceleration value of the X axis in a specific driving state, and y is the first characteristic value corresponding to the acceleration value of the Y axis in a specific driving state, z is the first characteristic value corresponding to the acceleration value of the Z axis in a specific driving state, and Max is the second characteristic value corresponding to the resultant acceleration value in a specific driving state.
The computer device according to claim 11, wherein, before the step of acquiring driving data during the driving of the vehicle in real time, it further comprises:

Continuously collecting multiple sets of motion data of the user according to a preset frequency within a specified time period, where the motion data includes second acceleration data, second angular acceleration data, and second azimuth angle data;

According to the second acceleration data, the preset fourth characteristic value data corresponding to the second acceleration data in the stroke start state, and the preset second algorithm, the fourth acceleration reference value corresponding to the stroke start state is calculated And, calculating the number of exercise steps of the user according to the second angular acceleration data and a preset step-counting algorithm; and, according to the second azimuth data, corresponding to the first in the state of using the mobile terminal The preset fifth characteristic value data of the two-azimuth angle data and the second algorithm calculate the azimuth angle reference value corresponding to the state of the mobile terminal in use;

The fourth acceleration reference value is compared with a preset acceleration threshold value, the number of exercise steps is compared with a preset step threshold value, and the azimuth angle reference value is compared with the preset azimuth angle threshold value. Compare, and determine whether the user is currently driving according to the result of the comparison;

If the user is currently in a driving state, the step of acquiring driving data during the driving of the vehicle in real time is executed.
The computer device according to claim 14, wherein the preset fourth characteristic value data includes a fourth characteristic value coefficient corresponding to the acceleration value of the X axis in the stroke start state, and in the stroke start state The fifth characteristic value coefficient corresponding to the acceleration value of the Y axis, the sixth characteristic value coefficient corresponding to the acceleration value of the Z axis in the stroke start state, the second acceleration data, the stroke The step of calculating the fourth acceleration reference value corresponding to the starting state of the stroke by the preset fourth characteristic value data corresponding to the second acceleration data and the preset second algorithm in the starting state includes:

Use the preset third calculation formula
Calculate the second acceleration data to obtain the fourth acceleration reference value, where in the third calculation formula, a four-parameter is the fourth acceleration reference value, and a x2 is the second acceleration The acceleration value of the X axis in the data, a y2 is the acceleration value of the Y axis in the second acceleration data, a z2 is the acceleration value of the Z axis in the second acceleration data, D is the fourth characteristic value coefficient, E is the fifth characteristic value coefficient, F is the sixth characteristic value coefficient, and n is the specified duration.
A computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, a driving behavior detection method is implemented, and the steps of the driving behavior detection method include:

Real-time acquisition of driving data during the driving of the vehicle, where the driving data includes first acceleration data, first angular acceleration data, and first azimuth angle data;

Calculate the acceleration reference value in each driving state according to the first acceleration data, the preset first characteristic value data corresponding to the first acceleration data in each driving state, and the preset first algorithm; and, According to the first angular acceleration data, the preset second characteristic value data corresponding to the first angular acceleration data in each of the driving states, and the first algorithm, the angular acceleration reference in each of the driving states is calculated Value; and, according to the first azimuth angle data, the preset third characteristic value data corresponding to the first azimuth angle data in each of the driving states, and the first algorithm, the calculation of each of the driving states The reference value of the azimuth angle of, wherein the driving state includes a sharp turn state, a sharp acceleration state, and a sharp braking state;

Comparing the acceleration reference value with a preset acceleration threshold value, the angular acceleration reference value and the preset angular acceleration threshold value, and the azimuth angle reference value and the preset azimuth angle threshold value respectively in the same driving state;

If the acceleration reference value in the same driving state is greater than the preset acceleration threshold, and the angular acceleration reference value in the same driving state is greater than the preset angular acceleration threshold, and the azimuth angle reference value in the same driving state is greater than the preset acceleration threshold. Setting the azimuth angle threshold, it is determined that the user currently has a dangerous driving behavior, where the dangerous driving behavior includes one or more of a sharp turn, a sharp acceleration, and a sharp brake.
The computer-readable storage medium according to claim 16, wherein the preset first characteristic value data includes a first characteristic value coefficient corresponding to an acceleration value of the X-axis in a specific driving state, and in a specific driving state The second eigenvalue coefficient corresponding to the acceleration value of the Y axis, and the third eigenvalue coefficient corresponding to the acceleration value of the Z axis in a specific driving state, according to the first acceleration data, corresponding to each driving state The step of calculating the reference value of acceleration in each driving state by the preset first characteristic value data of the first acceleration data and the preset first algorithm includes:

Use the preset first calculation formula a parameter = Aa x1 +Ba y1 +Ca z1 to calculate the first acceleration data to obtain the first acceleration reference value corresponding to the sharp turning state and corresponding to the sharp acceleration state The second acceleration reference value corresponding to the sudden braking state and the third acceleration reference value corresponding to the sudden braking state, wherein, in the first calculation formula, a parameter is the acceleration reference value in a specific driving state, and a x1 is the first The acceleration value of the X axis in the acceleration data, a y1 is the acceleration value of the Y axis in the first acceleration data, a z1 is the acceleration value of the Z axis in the first acceleration data, and A is the first characteristic value coefficient , B is the second characteristic value coefficient, and C is the third characteristic value coefficient.
18. The computer-readable storage medium according to claim 17, wherein, before the step of acquiring driving data during the driving of the vehicle in real time, the method further comprises:

Collecting acceleration sample data in each of the driving states for a predetermined time period according to a predetermined frequency, wherein there are at least three sets of acceleration sample data in each of the driving states;

Use the preset second calculation formula
Calculate each group of the acceleration sample data in each of the driving states to obtain multiple sets of corresponding total acceleration values, where in the second calculation formula, a is always the total acceleration value in a specific driving state , A x is the acceleration value of the X axis in the acceleration sample data in a specific driving state, a y is the acceleration value of the Y axis in the acceleration sample data in a specific driving state, and a z is the acceleration sample in the specific driving state The acceleration value of the Z axis in the data;

Fourier transform is performed on each group of said acceleration sample data and multiple sets of corresponding total acceleration values in each said driving state, respectively, to obtain multiple sets of first frequency domain data and multiple sets of each said driving state Second frequency domain data corresponding to the total acceleration value;

Perform spectrum analysis on each group of the first frequency domain data and each group of the second frequency domain data, and obtain multiple sets of first feature values corresponding to the acceleration sample data in each of the driving conditions and multiple sets of corresponding data. The second characteristic value of the total acceleration value;

Using the preset second calculation formula Ax+By+Cz=Max, each group of the first characteristic value and each group of the second characteristic value are respectively calculated to obtain the first characteristic value coefficient and the first characteristic value coefficient. Two eigenvalue coefficients and the third eigenvalue coefficient, wherein, in the second calculation formula, A is the first eigenvalue coefficient, B is the second eigenvalue coefficient, and C is the third The characteristic value coefficient, where x is the first characteristic value corresponding to the acceleration value of the X axis in a specific driving state, and y is the first characteristic value corresponding to the acceleration value of the Y axis in a specific driving state, z is the first characteristic value corresponding to the acceleration value of the Z axis in a specific driving state, and Max is the second characteristic value corresponding to the resultant acceleration value in a specific driving state.
15. The computer-readable storage medium according to claim 16, wherein before the step of acquiring driving data during the driving of the vehicle in real time, the method further comprises:

Continuously collecting multiple sets of motion data of the user according to a preset frequency within a specified time period, where the motion data includes second acceleration data, second angular acceleration data, and second azimuth angle data;

According to the second acceleration data, the preset fourth characteristic value data corresponding to the second acceleration data in the stroke start state, and the preset second algorithm, the fourth acceleration reference value corresponding to the stroke start state is calculated And, calculating the number of exercise steps of the user according to the second angular acceleration data and a preset step-counting algorithm; and, according to the second azimuth data, corresponding to the first in the state of using the mobile terminal The preset fifth characteristic value data of the two-azimuth angle data and the second algorithm calculate the azimuth angle reference value corresponding to the state of the mobile terminal in use;

The fourth acceleration reference value is compared with a preset acceleration threshold value, the number of exercise steps is compared with a preset step threshold value, and the azimuth angle reference value is compared with the preset azimuth angle threshold value. Compare, and determine whether the user is currently driving according to the result of the comparison;

If the user is currently in a driving state, the step of acquiring driving data during the driving of the vehicle in real time is executed.
The computer-readable storage medium according to claim 19, wherein the preset fourth characteristic value data includes a fourth characteristic value coefficient corresponding to an acceleration value of the X axis in the stroke start state, The fifth characteristic value coefficient corresponding to the acceleration value of the Y axis in the stroke start state, and the sixth characteristic value coefficient corresponding to the acceleration value of the Z axis in the stroke start state, according to the second acceleration data The step of calculating the fourth acceleration reference value corresponding to the start state of the stroke corresponding to the preset fourth characteristic value data of the second acceleration data and the preset second algorithm in the start state of the stroke includes:

Use the preset third calculation formula
Calculate the second acceleration data to obtain the fourth acceleration reference value, where in the third calculation formula, a four-parameter is the fourth acceleration reference value, and a x2 is the second acceleration The acceleration value of the X axis in the data, a y2 is the acceleration value of the Y axis in the second acceleration data, a z2 is the acceleration value of the Z axis in the second acceleration data, D is the fourth characteristic value coefficient, E is the fifth characteristic value coefficient, F is the sixth characteristic value coefficient, and n is the specified duration.