CN112183151A - Driving behavior detection method and device - Google Patents

Abstract

The application discloses a driving behavior detection method and device, wherein the method comprises the following steps: acquiring current road traffic state information, vehicle state information and driver state information; carrying out danger level standardization processing on the information to obtain corresponding road traffic state vectors, vehicle state vectors and driver state vectors; determining a current road traffic state dangerous driving index, a vehicle state dangerous driving index and a driver state dangerous driving index by using the vectors and the corresponding dangerous driving judgment matrix; determining a current dangerous driving comprehensive index by utilizing the dangerous driving indexes and a preset dangerous driving comprehensive judgment matrix; and finally, judging whether the alarm is needed at present according to the dangerous driving comprehensive index and a preset early warning threshold value, and executing a corresponding alarm process when needed. By adopting the method and the device, the accuracy of identifying and alarming the dangerous driving can be improved.

Description

Driving behavior detection method and device

Technical Field

The invention relates to a computer application technology, in particular to a driving behavior detection method and device.

Background

At present, driving behavior data are mainly acquired through a vehicle-mounted DMS system and are analyzed, but with the increase of the types and the number of vehicle-mounted sensors, various vehicle-mounted sensors can acquire environmental information, vehicle state information, driver state information and the like of a vehicle in real time, the data can be used as driving behavior data, and effective guidance information is provided for a driver through fusion and analysis of the driving behavior data. The driving behavior data sources are complex, the data types are various, and the real-time driving behavior recognition technology based on multi-source data fusion becomes a development trend.

At present, most driving behavior recognition and judgment methods are based on fusion analysis of single data sources or data of the same type, the data analysis angle is not comprehensive, and more abundant perception data resources are not fully utilized, so that dangerous driving of a driver cannot be accurately recognized.

Disclosure of Invention

In view of the above, the main object of the present invention is to provide a driving behavior detection method and device, which can improve the accuracy of recognizing and alarming dangerous driving.

In order to achieve the purpose, the technical scheme provided by the invention is as follows:

a driving behavior detection method, comprising:

acquiring current road traffic state information, vehicle state information and driver state information; respectively carrying out danger level standardization processing on the road traffic state information, the vehicle state information and the driver state information to obtain corresponding road traffic state vectors, vehicle state vectors and driver state vectors;

determining a current dangerous driving index of the road traffic state by using the road traffic state vector and a preset dangerous driving judgment matrix of the road traffic state; determining a current dangerous driving index of the vehicle state by using the vehicle state vector and a preset dangerous driving judgment matrix of the vehicle state; determining a current dangerous driving index of the driver state by using the driver state vector and a preset dangerous driving judgment matrix of the driver state;

determining a current dangerous driving comprehensive index by using the road traffic state dangerous driving index, the vehicle state dangerous driving index, the driver state dangerous driving index and a preset dangerous driving comprehensive judgment matrix;

and judging whether the alarm is needed at present or not according to the dangerous driving comprehensive index and a preset early warning threshold value, and executing a corresponding alarm process when needed.

Preferably, the stage normalization process comprises:

determining a danger level corresponding to each attribute in the road traffic state information, the vehicle state information and the driver state information by using a preset dangerous driving level mapping relation; the dangerous driving grade mapping relation comprises a dangerous grade corresponding to each attribute value contained in each attribute of road traffic state information, vehicle state information and driver state information;

for the current road traffic state information, taking the danger level corresponding to each attribute as a vector element, and generating a road traffic state vector corresponding to the road traffic state information;

for the current vehicle state information, generating a vehicle state vector corresponding to the vehicle state information by taking the danger level corresponding to each attribute as a vector element;

and for the current driver state information, generating a driver state vector corresponding to the driver state information by taking the danger level corresponding to each attribute as a vector element.

Preferably, the attributes of the road traffic status information include: road type, road status, and weather status; the attributes of the vehicle state information include: overspeed, distance between front and rear vehicles, whether the vehicles slide in neutral gear, turning turn to turn lights and traffic lights; the attributes of the driver state information include: whether a safety belt is fastened, whether a steering wheel is correctly held, the yawning times, the eye closing time length, whether other equipment is used in the driving process and whether the sight line of a driver keeps watching the front in the driving process of the vehicle.

Preferably, the value range of the road type includes: town roads, urban roads, highways, mountain roads, and other road types;

the value range of the road state comprises: unobstructed, slight congestion, serious congestion and traffic accidents;

the value range of the weather state comprises: normal weather, sand-dust weather, rainy weather, snow weather and heavy fog weather, wherein the normal weather comprises sunny days and cloudy days;

the value range of the vehicle overspeed comprises the following steps: no overspeed, less than 10% overspeed, 10% to 30% overspeed, 30% to 50% overspeed and more than 50% overspeed;

the value range of the front and rear vehicle distances comprises: greater than 100 meters, greater than 80 meters and less than or equal to 100 meters, greater than 60 meters and less than or equal to 80 meters, greater than 50 meters and less than or equal to 60 meters, and less than or equal to 50 meters;

the value range of whether the neutral gear slides comprises the following steps: neutral coasting and no neutral coasting;

the value range of the turning turn signal lamp comprises the following steps: the steering lamp is used correctly, the steering lamp is not turned on, and the steering lamp is turned by mistake;

the value range of the traffic signal lamp comprises: green light running, yellow light running and red light running;

the value range of whether the safety belt is fastened comprises the following steps: belting and unbelting;

the value range of whether the steering wheel is correctly held comprises the following steps: a double-hand-held steering wheel, a single-hand-held steering wheel and a non-held steering wheel;

the value range of the yawning times comprises: the yawning times in 5 minutes are less than or equal to 2, the yawning times in 5 minutes are more than 3 and less than or equal to 5, the yawning times in 5 minutes are more than 5 and less than or equal to 7, and the yawning times in 5 minutes are more than 7;

the value range of the eye closing time comprises the following steps: less than 0.5 seconds, 1 second, 3 seconds, and 5 seconds;

the value range of whether other equipment is used in the driving process comprises the following steps: use and non-use of other devices;

whether the driver's sight keeps watching the value range in the place ahead during the vehicle is gone includes: stay ahead of gaze and not stay ahead of gaze.

Preferably, the danger levels corresponding to the town road, the city road, the expressway, the mountain road and other road types are respectively as follows: 1. 2,3,4, 2;

the corresponding danger levels of the unobstructed road, the slight congestion, the serious congestion and the traffic accident are respectively as follows: 1. 2,3 and 4;

the danger levels corresponding to normal weather, sand and dust weather, rainy weather, snowy weather and heavy fog weather respectively are as follows: 1. 2,3,4, 5;

the corresponding danger grades of the non-overspeed, the overspeed less than 10%, the overspeed 10% to 30%, the overspeed 30% to 50% and the overspeed more than 50% are respectively as follows: 1. 2,3,4, 5;

the danger levels respectively corresponding to more than 100 meters, more than 80 meters and less than or equal to 100 meters, more than 60 meters and less than or equal to 80 meters, more than 50 meters and less than or equal to 60 meters, and less than or equal to 50 meters are: 1. 2,3,4, 5;

the corresponding danger grades of the neutral gear sliding and the non-neutral gear sliding are respectively as follows: 4. 1;

the danger levels corresponding to the correct use of the steering lamp, the no-turn-on of the steering lamp and the wrong turn-on of the steering lamp are respectively as follows: 1. 3, 5;

the danger levels corresponding to the green light running, the yellow light running and the red light running are respectively as follows: 1. 3, 5;

the corresponding danger grades of the belted safety belt and the belted safety belt are respectively as follows: 1. 5;

the danger levels respectively corresponding to the double-hand-held steering wheel, the single-hand-held steering wheel and the steering wheel which is not held are as follows: 1. 3, 5;

the corresponding danger levels of the yawning times within 5 minutes being less than or equal to 2 times, the yawning times within 5 minutes being more than 3 times and less than or equal to 5 times, the yawning times within 5 minutes being more than 5 times and less than or equal to 7 times, and the yawning times within 5 minutes being more than 7 times are respectively as follows: 1. 2,3 and 5;

the danger grades corresponding to less than 0.5 second, 1 second, 3 seconds and 5 seconds are respectively as follows: 1. 2,3 and 5;

the risk levels respectively corresponding to the use of other equipment and the non-use of other equipment are respectively as follows: 4. 1;

the danger levels respectively corresponding to the fixation front and the fixation non-front are respectively as follows: 4. 1.

Preferably, the determining the current dangerous driving index of the road traffic state by using the road traffic state vector and a preset dangerous driving judgment matrix of the road traffic state comprises:

calculating the dangerous driving judgment matrix of the road traffic state

Characteristic weight vector W of₁＝(w₁₁，...，w_1n)^TWherein

n is the number of attributes contained in the road traffic state information; x_ijThe ratio of the influence degree of the attribute j and the attribute i of the road traffic state information on dangerous driving is obtained; i is more than or equal to 1 and less than or equal to n; j is more than or equal to 1 and less than or equal to n;

computing

Obtaining the dangerous driving index P of the current road traffic state₁Wherein v is_1iIs the road traffic state vector v₁＝(v₁₁，...，v_1n) The ith element in (1).

Preferably, the determining the current dangerous driving index of the vehicle state by using the vehicle state vector and a preset dangerous driving judgment matrix of the vehicle state includes:

calculating the vehicle state dangerous driving judgment matrix

Characteristic weight vector W of₂＝(w₂₁，...，w_2k)^TWherein

k is the number of attributes contained in the vehicle state information; y is_ijThe ratio of the influence degree of the attribute j and the attribute i of the vehicle state information on dangerous driving is obtained; i is more than or equal to 1 and less than or equal to k; j is more than or equal to 1 and less than or equal to k;

computing

Obtaining the dangerous driving index P of the current vehicle state₂Wherein v is_2iIs the vehicle state vector v₂＝(v₂₁，...，v_2k) The ith element in (1).

Preferably, the determining the current dangerous driving index of the driver state by using the driver state vector and a preset dangerous driving judgment matrix of the driver state includes:

calculating the driver state dangerous driving judgment matrix

Characteristic weight vector W of₃＝(w₃₁，...，w_3s)^TWherein

s is the number of attributes contained in the driver state information; z_ijThe ratio of the influence degree of the attribute j and the attribute i of the driver state information on dangerous driving respectively is obtained; i is more than or equal to 1 and less than or equal to s; j is more than or equal to 1 and less than or equal to s;

computing

Obtaining the dangerous driving index P of the current driver state₃Wherein v is_3iIs the driver state vector v₃＝(v₃₁，...，v_3s) The ith element inAnd (4) element.

Preferably, the determining the current dangerous driving comprehensive index by using the road traffic state dangerous driving index, the vehicle state dangerous driving index, the driver state dangerous driving index and a preset dangerous driving comprehensive judgment matrix includes:

calculating the dangerous driving comprehensive judgment matrix

The weight vector W ═ W (W) of the features₁，w₂，w₃)^TWherein w is₁+w₂+w₃＝1；

Calculating P ═ w₁P₁+w₂P₂+w₃P₃Obtaining a dangerous driving index P of the current driver state, wherein P₁Is the dangerous driving index, P, of the road traffic state₂The dangerous driving index of the vehicle state is obtained; p₃And the dangerous driving index of the driver state is obtained.

Preferably, the judging whether to alarm at present according to the dangerous driving comprehensive index and a preset early warning threshold value includes:

and if the dangerous driving comprehensive index is larger than a preset early warning threshold value, judging that the alarm is required.

A driving behavior detection apparatus comprising: a processor to:

acquiring current road traffic state information, vehicle state information and driver state information; respectively carrying out danger level standardization processing on the road traffic state information, the vehicle state information and the driver state information to obtain corresponding road traffic state vectors, vehicle state vectors and driver state vectors;

determining a current dangerous driving index of the road traffic state by using the road traffic state vector and a preset dangerous driving judgment matrix of the road traffic state; determining a current dangerous driving index of the vehicle state by using the vehicle state vector and a preset dangerous driving judgment matrix of the vehicle state; determining a current dangerous driving index of the driver state by using the driver state vector and a preset dangerous driving judgment matrix of the driver state;

determining a current dangerous driving comprehensive index by using the road traffic state dangerous driving index, the vehicle state dangerous driving index, the driver state dangerous driving index and a preset dangerous driving comprehensive judgment matrix;

and judging whether the alarm is needed at present or not according to the dangerous driving comprehensive index and a preset early warning threshold value, and executing a corresponding alarm process when needed.

Preferably, the processor is specifically configured to: respectively carrying out danger level standardization processing on the road traffic state information, the vehicle state information and the driver state information comprises the following steps:

determining a danger level corresponding to each attribute in the road traffic state information, the vehicle state information and the driver state information by using a preset dangerous driving level mapping relation; the dangerous driving grade mapping relation comprises a dangerous grade corresponding to each attribute value contained in each attribute of road traffic state information, vehicle state information and driver state information;

for the current road traffic state information, taking the danger level corresponding to each attribute as a vector element, and generating a road traffic state vector corresponding to the road traffic state information;

for the current vehicle state information, generating a vehicle state vector corresponding to the vehicle state information by taking the danger level corresponding to each attribute as a vector element;

and for the current driver state information, generating a driver state vector corresponding to the driver state information by taking the danger level corresponding to each attribute as a vector element.

Preferably, the attributes of the road traffic status information include: road type, road status, and weather status; the attributes of the vehicle state information include: overspeed, distance between front and rear vehicles, whether the vehicles slide in neutral gear, turning turn to turn lights and traffic lights; the attributes of the driver state information include: whether a safety belt is fastened, whether a steering wheel is correctly held, the yawning times, the eye closing time length, whether other equipment is used in the driving process and whether the sight line of a driver keeps watching the front in the driving process of the vehicle.

Preferably, the value range of the road type includes: town roads, urban roads, highways, mountain roads, and other road types;

the value range of the road state comprises: unobstructed, slight congestion, serious congestion and traffic accidents;

the value range of the weather state comprises: normal weather, sand-dust weather, rainy weather, snow weather and heavy fog weather, wherein the normal weather comprises sunny days and cloudy days;

the value range of the vehicle overspeed comprises the following steps: no overspeed, less than 10% overspeed, 10% to 30% overspeed, 30% to 50% overspeed and more than 50% overspeed;

the value range of the front and rear vehicle distances comprises: greater than 100 meters, greater than 80 meters and less than or equal to 100 meters, greater than 60 meters and less than or equal to 80 meters, greater than 50 meters and less than or equal to 60 meters, and less than or equal to 50 meters;

the value range of whether the neutral gear slides comprises the following steps: neutral coasting and no neutral coasting;

the value range of the turning turn signal lamp comprises the following steps: the steering lamp is used correctly, the steering lamp is not turned on, and the steering lamp is turned by mistake;

the value range of the traffic signal lamp comprises: green light running, yellow light running and red light running;

the value range of whether the safety belt is fastened comprises the following steps: belting and unbelting;

the value range of whether the steering wheel is correctly held comprises the following steps: a double-hand-held steering wheel, a single-hand-held steering wheel and a non-held steering wheel;

the value range of the yawning times comprises: the yawning times in 5 minutes are less than or equal to 2, the yawning times in 5 minutes are more than 3 and less than or equal to 5, the yawning times in 5 minutes are more than 5 and less than or equal to 7, and the yawning times in 5 minutes are more than 7;

the value range of the eye closing time comprises the following steps: less than 0.5 seconds, 1 second, 3 seconds, and 5 seconds;

the value range of whether other equipment is used in the driving process comprises the following steps: use and non-use of other devices;

whether the driver's sight keeps watching the value range in the place ahead during the vehicle is gone includes: stay ahead of gaze and not stay ahead of gaze.

Preferably, the danger levels corresponding to the town road, the city road, the expressway, the mountain road and other road types are respectively as follows: 1. 2,3,4, 2;

the corresponding danger levels of the unobstructed road, the slight congestion, the serious congestion and the traffic accident are respectively as follows: 1. 2,3 and 4;

the danger levels corresponding to normal weather, sand and dust weather, rainy weather, snowy weather and heavy fog weather respectively are as follows: 1. 2,3,4, 5;

the corresponding danger grades of the non-overspeed, the overspeed less than 10%, the overspeed 10% to 30%, the overspeed 30% to 50% and the overspeed more than 50% are respectively as follows: 1. 2,3,4, 5;

the danger levels respectively corresponding to more than 100 meters, more than 80 meters and less than or equal to 100 meters, more than 60 meters and less than or equal to 80 meters, more than 50 meters and less than or equal to 60 meters, and less than or equal to 50 meters are: 1. 2,3,4, 5;

the corresponding danger grades of the neutral gear sliding and the non-neutral gear sliding are respectively as follows: 4. 1;

the danger levels corresponding to the correct use of the steering lamp, the no-turn-on of the steering lamp and the wrong turn-on of the steering lamp are respectively as follows: 1. 3, 5;

the danger levels corresponding to the green light running, the yellow light running and the red light running are respectively as follows: 1. 3, 5;

the corresponding danger grades of the belted safety belt and the belted safety belt are respectively as follows: 1. 5;

the danger levels respectively corresponding to the double-hand-held steering wheel, the single-hand-held steering wheel and the steering wheel which is not held are as follows: 1. 3, 5;

the corresponding danger levels of the yawning times within 5 minutes being less than or equal to 2 times, the yawning times within 5 minutes being more than 3 times and less than or equal to 5 times, the yawning times within 5 minutes being more than 5 times and less than or equal to 7 times, and the yawning times within 5 minutes being more than 7 times are respectively as follows: 1. 2,3 and 5;

the danger grades corresponding to less than 0.5 second, 1 second, 3 seconds and 5 seconds are respectively as follows: 1. 2,3 and 5;

the risk levels respectively corresponding to the use of other equipment and the non-use of other equipment are respectively as follows: 4. 1;

the danger levels respectively corresponding to the fixation front and the fixation non-front are respectively as follows: 4. 1.

Preferably, the processor is specifically configured to: determining the current dangerous driving index of the road traffic state by using the road traffic state vector and a preset dangerous driving judgment matrix of the road traffic state comprises the following steps:

calculating the dangerous driving judgment matrix of the road traffic state

Characteristic weight vector W of₁＝(w₁₁，...，w_1n)^TWherein

n is the number of attributes contained in the road traffic state information; x_ijThe ratio of the influence degree of the attribute j and the attribute i of the road traffic state information on dangerous driving is obtained; i is more than or equal to 1 and less than or equal to n; j is more than or equal to 1 and less than or equal to n;

computing

Obtaining the dangerous driving index P of the current road traffic state₁Wherein v is_1iIs the road traffic state vector v₁＝(v₁₁，...，v_1n) The ith element in (1).

Preferably, the processor is specifically configured to: determining the current dangerous driving index of the vehicle state by using the vehicle state vector and a preset dangerous driving judgment matrix of the vehicle state comprises the following steps:

calculating the vehicle state dangerous driving judgment matrix

Characteristic weight vector W of₂＝(w₂₁，...，w_2k)^TWherein

k is the number of attributes contained in the vehicle state information; y is_ijThe ratio of the influence degree of the attribute j and the attribute i of the vehicle state information on dangerous driving is obtained; i is more than or equal to 1 and less than or equal to k; j is more than or equal to 1 and less than or equal to k;

computing

Obtaining the dangerous driving index P of the current vehicle state₂Wherein v is_2iIs the vehicle state vector v₂＝(v₂₁，...，v_2k) The ith element in (1).

Preferably, the processor is specifically configured to: determining a current dangerous driving index of the driver state by using the driver state vector and a preset dangerous driving judgment matrix of the driver state comprises:

calculating the driver state dangerous driving judgment matrix

Characteristic weight vector W of₃＝(w₃₁，...，w_3s)^TWherein

s is the number of attributes contained in the driver state information; z_ijThe ratio of the influence degree of the attribute j and the attribute i of the driver state information on dangerous driving respectively is obtained; i is more than or equal to 1 and less than or equal to s; j is more than or equal to 1 and less than or equal to s;

computing

Obtaining the dangerous driving index P of the current driver state₃Wherein v is_3iIs the driver state vector v₃＝(v₃₁，...，v_3s) The ith element in (1).

Preferably, the processor is specifically configured to: determining the current dangerous driving comprehensive index by using the road traffic state dangerous driving index, the vehicle state dangerous driving index, the driver state dangerous driving index and a preset dangerous driving comprehensive judgment matrix, wherein the step of determining the current dangerous driving comprehensive index comprises the following steps:

calculating the dangerous driving comprehensive judgment matrix

The weight vector W ═ W (W) of the features₁，w₂，w₃)^TWherein w is₁+w₂+w₃＝1；

Calculating P ═ w₁P₁+w₂P₂+w₃P₃Obtaining a dangerous driving index P of the current driver state, wherein P₁Is the dangerous driving index, P, of the road traffic state₂The dangerous driving index of the vehicle state is obtained; p₃And the dangerous driving index of the driver state is obtained.

Preferably, the processor is specifically configured to: judging whether the alarm is required at present according to the dangerous driving comprehensive index and a preset early warning threshold value comprises the following steps:

and if the dangerous driving comprehensive index is larger than a preset early warning threshold value, judging that the alarm is required.

The present application also discloses a non-transitory computer readable storage medium storing instructions that, when executed by a processor, cause the processor to perform the steps of the method of early warning detection of driving behavior as described above.

The application also discloses an electronic device comprising the non-volatile computer-readable storage medium as described above, and the processor having access to the non-volatile computer-readable storage medium.

According to the technical scheme, the driving behavior detection method and the driving behavior detection device provided by the invention firstly carry out the risk level standardization processing on the information uniformly based on the current road traffic state information, the vehicle state information and the driver state information, convert the information into the road traffic state vector, the vehicle state vector and the driver state vector, then respectively carry out the dangerous driving analysis on the state vectors by utilizing the preset road traffic state dangerous driving judgment matrix, the vehicle state dangerous driving judgment matrix and the driver state dangerous driving judgment matrix to obtain three dangerous driving indexes, and finally, judging whether the current dangerous driving needs to be alarmed or not based on the current dangerous driving comprehensive index and a preset early warning threshold value. Therefore, by analyzing the driving behavior data in all directions, the multi-source sensing data is fully utilized to judge the dangerous level of the driving behavior, and the accuracy of recognizing and alarming dangerous driving can be effectively improved.

Drawings

FIG. 1 is a schematic flow chart of a method according to an embodiment of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail with reference to the accompanying drawings and specific embodiments.

Fig. 1 is a schematic flow diagram of an embodiment of the present invention, and as shown in fig. 1, a driving behavior detection method implemented by the embodiment mainly includes:

step 101, acquiring current road traffic state information, vehicle state information and driver state information; and respectively carrying out danger level standardization processing on the road traffic state information, the vehicle state information and the driver state information to obtain corresponding road traffic state vectors, vehicle state vectors and driver state vectors.

In this step, according to a preset danger level standard, the current road traffic state information, vehicle state information and driver state information are subjected to danger level standardization processing, that is, each attribute value in the current information is converted into a danger level in the danger level standard, and then corresponding road traffic state vectors, vehicle state vectors and driver state vectors are constructed according to the danger levels corresponding to each attribute value in the information. Therefore, the information is subjected to the risk grade standardization processing in a unified manner, so that the comprehensive processing is favorably carried out in the subsequent steps based on the road traffic state vector, the vehicle state vector and the driver state vector, and finally a dangerous driving comprehensive index is obtained, so that the dangerous driving can be accurately identified and alarmed.

In practical applications, the current road traffic state information may be acquired from a third-party application, the current vehicle state information may be acquired by a vehicle-mounted sensor, and the current driver state information may be acquired by a vehicle-mounted Driver Monitoring System (DMS), and specific acquisition methods of these information are known by those skilled in the art and will not be described herein again.

Preferably, the road traffic status information, the vehicle status information and the driver status information may be respectively subjected to a risk level standardization process by using the following methods:

and step x1, determining a danger level corresponding to each attribute in the road traffic state information, the vehicle state information and the driver state information by using a preset dangerous driving level mapping relation.

The dangerous driving grade mapping relation comprises a dangerous grade corresponding to each attribute value contained in each attribute in the road traffic state information, the vehicle state information and the driver state information, so that the dangerous grade corresponding to each attribute value in the current road traffic state information, the vehicle state information and the driver state information can be obtained based on the dangerous driving grade mapping relation.

Specifically, those skilled in the art may set the dangerous driving level mapping relationship according to actual conditions.

Step x2, regarding the current road traffic state information, taking the danger level corresponding to each attribute as a vector element, and generating a road traffic state vector corresponding to the road traffic state information;

for the current vehicle state information, generating a vehicle state vector corresponding to the vehicle state information by taking the danger level corresponding to each attribute as a vector element;

and for the current driver state information, generating a driver state vector corresponding to the driver state information by taking the danger level corresponding to each attribute as a vector element.

In practical applications, those skilled in the art can set the attributes included in the road traffic status information, the vehicle status information, and the driver status information according to actual needs. For example, the information may preferably include the following attributes, but is not limited thereto.

The attributes of the road traffic status information include: road type, road status, and weather status.

The attributes of the vehicle state information include: overspeed, distance between front and rear vehicles, whether the vehicles slide in neutral, turning turn to turn lights and traffic lights.

The attributes of the driver state information include: whether a safety belt is fastened, whether a steering wheel is correctly held, the yawning times, the eye closing time length, whether other equipment is used in the driving process and whether the sight line of a driver keeps watching the front in the driving process of the vehicle.

Further, in practical application, a person skilled in the art can set a value range of each attribute contained in the road traffic state information, the vehicle state information, and the driver state information according to an actual driving scene. For example, when the information includes the above attributes, the value range of each attribute may be limited as follows, but is not limited thereto.

Preferably, the value range of the road type includes: town roads, urban roads, highways, mountain roads, and other road types;

the value range of the road state comprises: unobstructed, slight congestion, serious congestion and traffic accidents;

the value range of the weather state comprises: normal weather, sand-dust weather, rainy weather, snow weather and heavy fog weather, wherein the normal weather comprises sunny days and cloudy days;

the value range of the vehicle overspeed comprises the following steps: no overspeed, less than 10% overspeed, 10% to 30% overspeed, 30% to 50% overspeed and more than 50% overspeed;

the value range of the front and rear vehicle distances comprises: greater than 100 meters, greater than 80 meters and less than or equal to 100 meters, greater than 60 meters and less than or equal to 80 meters, greater than 50 meters and less than or equal to 60 meters, and less than or equal to 50 meters;

the value range of whether the neutral gear slides comprises the following steps: neutral coasting and no neutral coasting;

the value range of the turning turn signal lamp comprises the following steps: the steering lamp is used correctly, the steering lamp is not turned on, and the steering lamp is turned by mistake;

the value range of the traffic signal lamp comprises: green light running, yellow light running and red light running;

the value range of whether the safety belt is fastened comprises the following steps: belting and unbelting;

the value range of whether the steering wheel is correctly held comprises the following steps: a double-hand-held steering wheel, a single-hand-held steering wheel and a non-held steering wheel;

the value range of the yawning times comprises: the yawning times in 5 minutes are less than or equal to 2, the yawning times in 5 minutes are more than 3 and less than or equal to 5, the yawning times in 5 minutes are more than 5 and less than or equal to 7, and the yawning times in 5 minutes are more than 7;

the value range of the eye closing time comprises the following steps: less than 0.5 seconds, 1 second, 3 seconds, and 5 seconds;

the value range of whether other equipment is used in the driving process comprises the following steps: use and non-use of other devices;

whether the driver's sight keeps watching the value range in the place ahead during the vehicle is gone includes: stay ahead of gaze and not stay ahead of gaze.

In practical applications, the risk level criteria can be set by those skilled in the art according to practical needs, for example, preferably, the following 5 levels {1,2,3,4,5}, but is not limited thereto.

1: normal road driving conditions;

2: has slight influence on the road traffic safety;

3: has an impact on road traffic safety;

4: the method has great influence on the road traffic safety;

5: has a serious influence on the road traffic safety.

Further, on the basis of the above examples of road traffic status information, vehicle status information and driver status information, and the dangerous level criteria, the corresponding dangerous driving level mapping relationship may be preferably set as follows, but is not limited thereto:

preferably, the danger levels corresponding to the town road, the city road, the expressway, the mountain road and other road types are respectively as follows: 1. 2,3,4, 2;

the corresponding danger levels of the unobstructed road, the slight congestion, the serious congestion and the traffic accident are respectively as follows: 1. 2,3 and 4;

the danger levels corresponding to normal weather, sand and dust weather, rainy weather, snowy weather and heavy fog weather respectively are as follows: 1. 2,3,4, 5;

the corresponding danger grades of the non-overspeed, the overspeed less than 10%, the overspeed 10% to 30%, the overspeed 30% to 50% and the overspeed more than 50% are respectively as follows: 1. 2,3,4, 5;

the danger levels respectively corresponding to more than 100 meters, more than 80 meters and less than or equal to 100 meters, more than 60 meters and less than or equal to 80 meters, more than 50 meters and less than or equal to 60 meters, and less than or equal to 50 meters are: 1. 2,3,4, 5;

the corresponding danger grades of the neutral gear sliding and the non-neutral gear sliding are respectively as follows: 4. 1;

the danger levels corresponding to the correct use of the steering lamp, the no-turn-on of the steering lamp and the wrong turn-on of the steering lamp are respectively as follows: 1. 3, 5;

the danger levels corresponding to the green light running, the yellow light running and the red light running are respectively as follows: 1. 3, 5;

the corresponding danger grades of the belted safety belt and the belted safety belt are respectively as follows: 1. 5;

the danger levels respectively corresponding to the double-hand-held steering wheel, the single-hand-held steering wheel and the steering wheel which is not held are as follows: 1. 3, 5;

the corresponding danger levels of the yawning times within 5 minutes being less than or equal to 2 times, the yawning times within 5 minutes being more than 3 times and less than or equal to 5 times, the yawning times within 5 minutes being more than 5 times and less than or equal to 7 times, and the yawning times within 5 minutes being more than 7 times are respectively as follows: 1. 2,3 and 5;

the danger grades corresponding to less than 0.5 second, 1 second, 3 seconds and 5 seconds are respectively as follows: 1. 2,3 and 5;

the risk levels respectively corresponding to the use of other equipment and the non-use of other equipment are respectively as follows: 4. 1;

the danger levels respectively corresponding to the fixation front and the fixation non-front are respectively as follows: 4. 1.

It should be noted that, in the above example, the larger the risk level data is, the higher the corresponding risk level is, for example, the risk level of neutral coasting is 4, and the risk level of no neutral coasting is 1.

Step 102, determining a current dangerous driving index of the road traffic state by using the road traffic state vector and a preset dangerous driving judgment matrix of the road traffic state; determining a current dangerous driving index of the vehicle state by using the vehicle state vector and a preset dangerous driving judgment matrix of the vehicle state; and determining the current dangerous driving index of the driver state by using the driver state vector and a preset dangerous driving judgment matrix of the driver state.

Preferably, the following method may be adopted to determine the current dangerous driving index of the road traffic state by using the road traffic state vector and a preset dangerous driving judgment matrix of the road traffic state:

calculating the dangerous driving judgment matrix of the road traffic state

Characteristic weight vector W of₁＝(w₁₁，...，w_1n)^TWherein

n is the number of attributes contained in the road traffic state information; x_ijThe ratio of the influence degree of the attribute j and the attribute i of the road traffic state information on dangerous driving is obtained; i is more than or equal to 1 and less than or equal to n; j is more than or equal to 1 and less than or equal to n;

computing

Obtaining the dangerous driving index P of the current road traffic state₁Wherein v is_1iIs the road traffic state vector v₁＝(v₁₁，...，v_1n) The ith element in (1).

In the method, the dangerous driving judgment matrix of the road traffic state

The method can be constructed by a person skilled in the art by utilizing a method in a hierarchical analysis method in combination with an actual traffic application scene.

Preferably, the following method may be adopted to determine the current dangerous driving index of the vehicle state by using the vehicle state vector and a preset dangerous driving judgment matrix of the vehicle state:

calculating the vehicle state dangerous driving judgment matrix

Characteristic weight vector W of₂＝(w₂₁，...，w_2k)^TWherein

k is the number of attributes contained in the vehicle state information; y is_ijThe ratio of the influence degree of the attribute j and the attribute i of the vehicle state information on dangerous driving is obtained; i is more than or equal to 1 and less than or equal to k; j is more than or equal to 1 and less than or equal to k;

computing

Obtaining the dangerous driving index P of the current vehicle state₂Wherein v is_2iIs the vehicle state vector v₂＝(v₂₁，...，v_2k) The ith element in (1).

In the method, the vehicle state dangerous driving judgment matrix

The method can be constructed by a person skilled in the art by utilizing a method in a hierarchical analysis method in combination with an actual traffic application scene.

Preferably, the following method may be adopted to determine the current dangerous driving index of the driver state by using the driver state vector and a preset dangerous driving judgment matrix of the driver state:

calculating the driver state dangerous driving judgment matrix

Characteristic weight vector W of₃＝(w₃₁，...，w_3s)^TWherein

s is the number of attributes contained in the driver state information; z_ijThe ratio of the influence degree of the attribute j and the attribute i of the driver state information on dangerous driving respectively is obtained; i is more than or equal to 1 and less than or equal to s; j is more than or equal to 1 and less than or equal to s;

computing

Obtaining the dangerous driving index P of the current driver state₃Wherein v is_3iIs the driver state vector v₃＝(v₃₁，...，v_3s) The ith element in (1).

The driver state dangerous driving judgment matrix

The method can be constructed by a person skilled in the art by utilizing a method in a hierarchical analysis method in combination with an actual traffic application scene.

And 103, determining the current dangerous driving comprehensive index by using the road traffic state dangerous driving index, the vehicle state dangerous driving index, the driver state dangerous driving index and a preset dangerous driving comprehensive judgment matrix.

In this step, the dangerous driving index of the road traffic state, the dangerous driving index of the vehicle state and the dangerous driving index of the driver state calculated in the step 102 are integrated, and a preset dangerous driving integrated judgment matrix is used to calculate the current dangerous driving integrated index, so that the current dangerous driving condition is judged by fusing the three dangerous driving indexes, the dangerous level of the driving behavior is judged by fully utilizing the multi-source perception data, and the accuracy of identifying and alarming the dangerous driving can be effectively improved.

Preferably, the following method may be adopted to determine the current dangerous driving comprehensive index by using the road traffic state dangerous driving index, the vehicle state dangerous driving index, the driver state dangerous driving index and a preset dangerous driving comprehensive judgment matrix:

calculating the dangerous driving comprehensive judgment matrix

The weight vector W ═ W (W) of the features₁，w₂，w₃)^TWherein w is₁+w₂+w₃＝1；

Calculating P ═ w₁P₁+w₂P₂+w₃P₃Obtaining a dangerous driving index P of the current driver state, wherein P₁Is the dangerous driving index, P, of the road traffic state₂Dangerous driving index for the vehicle state；P₃And the dangerous driving index of the driver state is obtained.

And step 104, judging whether the alarm is needed at present according to the dangerous driving comprehensive index and a preset early warning threshold value, and executing a corresponding alarm process when needed.

Preferably, the following method can be adopted to judge whether to alarm at present according to the dangerous driving comprehensive index and a preset early warning threshold value:

and if the dangerous driving comprehensive index is larger than a preset early warning threshold value, judging that the alarm is required.

In this step, the early warning threshold for determining whether to alarm or not may be set by a person skilled in the art according to actual needs, as long as execution of the alarm process can be triggered in time during dangerous driving.

In practical applications, the alarm process may be implemented by using an existing method, or may be implemented by setting a suitable alarm policy according to actual needs by a person skilled in the art, and details are not described herein.

The present invention is further described below by a driving behavior detection method implemented based on the above method embodiment.

The risk level criteria used in this embodiment include 5 levels {1,2,3,4,5} of:

1: normal road driving conditions;

2: has slight influence on the road traffic safety;

3: has an impact on road traffic safety;

4: the method has great influence on the road traffic safety;

5: has a serious influence on the road traffic safety.

The road traffic state information is mainly provided by third-party software, and comprises data A of three attributes, namely { road type, road state and weather state }, wherein the standardization grade corresponding to each attribute is as follows:

the weather is { normal weather, sand and dust weather, rainy day, snow day, heavy fog weather } {1,2,3,4,5}, wherein the normal weather includes weather conditions which do not have obvious influence on vehicle driving such as sunny days, cloudy days and the like.

Road type ═ town road, urban road, expressway, mountain road, other road type } ═ 1,2,3,4, 2 }; here, it is considered that, in view of,

the road state is { unobstructed, slightly congested, heavily congested and traffic accident } 1,2,3 and 4.

The vehicle state information is mainly acquired by a vehicle-mounted sensor, data B containing 5 attributes is { vehicle overspeed, vehicle distance, whether the vehicle slides in neutral gear, turning turn to a steering lamp and a traffic signal lamp }, and the corresponding standardization of each attribute is classified as follows:

vehicle overspeed is { no overspeed, overspeed is less than 10%, overspeed is 10% -30%, overspeed is 30% -50%, overspeed is greater than 50% } {1,2,3,4,5 };

front and back vehicle distances are { greater than 100 meters, 80 meters to 100 meters, 60 meters to 80 meters, 50 meters to 60 meters, and less than 50 meters } {1,2,3,4,5 };

whether the engine runs in neutral gear is determined as no, yes, 1, 4;

turn signal lamp { correct but using turn signal lamp, not turn signal lamp, wrong turn signal lamp } {1,3,4}

Traffic signal light { green light driving, yellow light driving, red light driving } {1,3,5}

The driver state information is acquired by the vehicle-mounted DMS, and includes C ═ whether to fasten a safety belt, whether to correctly hold a steering wheel, the number of yawning times, the eye closing duration, whether to use other equipment (such as a mobile phone, a flat panel and the like) in the driving process, whether the driver sight line keeps watching the front part in the driving process, and the corresponding grades are as follows:

whether a safety belt is fastened or not is { fastening safety belt, unbuckled safety belt } {1,5 };

whether the steering wheel is correctly held or not is { two hands hold the steering wheel, one hand holds the steering wheel, and no steering wheel is held } {1,3,5 };

the number of yawns in 5 minutes is { 2 or less, 3 to 5, 5 to 7, and more than 7 } {1,2,3,5 };

eye-closing time length { less than 0.5 seconds, 1 second, 3 seconds, 5 seconds } {1,2,3,5 };

whether other equipment is used in the driving process is determined to be { no, yes } - {1,4 };

whether the sight line of the driver keeps looking ahead during the running of the vehicle is determined as { yes, no } - {1,4}

Primary data fusion: and constructing a corresponding decision matrix for each data set, calculating a feature vector corresponding to the maximum feature value of the decision matrix, normalizing the feature vector into a weight of each attribute of the data set, and finally determining the risk index of each data set. The method comprises the following specific steps:

the road traffic state data only comprises 3-dimensional data, and a judgment matrix of the road traffic state with 3 rows multiplied by 3 columns is constructed according to a method in a hierarchical analysis method. The matrix rows and columns correspond to the following table:

	type of road	Road state	Weather conditions
				Type of road	1	3	1/2
Road state	1/3	1	1/5
				Weather conditions	2	5	1

The corresponding dangerous driving judgment matrix is as follows:

calculating the feature weight vector W of X₁＝(w₁₁，w₁₂，w₁₃)^T＝(0.309,0.109,0.582)^TThen there is P₁＝0.309·v₁₁+0.109·v₁₂+0.582·v₁₃Wherein v is_1iAnd obtaining the value of each attribute after the road traffic state data are standardized.

For the vehicle state information data containing 5 attributes, the final P2 is determined by determining the weight of each attribute. A judgment matrix of 5 rows by 5 columns is constructed according to a method in the hierarchical analysis method. The matrix rows and columns correspond to the following table:

the corresponding dangerous driving judgment matrix is as follows:

calculating the weight vector W of the features of Y₂＝(w₂₁，w₂₂，w₂₃，w₂₄，w₂₅)^T＝(0.392,0.102,0.142,0.083,0.281)^TWherein

Then there is P₂＝0.392·v₂₁+0.102·v₂₂+0.142·v₂₃+0.083·v₂₄+0.281·v₂₅Wherein v is_2iAnd obtaining the value of each attribute after the vehicle state information data is standardized.

Similarly, for the driver state information, a driver state information judgment matrix of 6 rows by 6 columns is constructed according to a hierarchical analysis method. The matrix rows and columns correspond to the following table:

the corresponding dangerous driving judgment matrix is as follows:

determining the feature weight W of each attribute₃＝(w₃₁，w₃₂，w₃₃，w₃₄，w₃₅，w₃₆)^T＝(0.408,0.194,0.061,0.145,0.096,0.096)^TWherein

P₃＝0.408·v₃₁+0.194·v₃₂+0.061·v₃₃+0.145·v₃₄+0.096·v₃₅+0.096·v₃₆Wherein v is_3iAnd obtaining values of various attributes after the driver state data are standardized.

(2) And performing secondary fusion, and calculating the final dangerous driving index. The method comprises the following specific steps:

and finishing the primary data fusion, then performing secondary fusion according to the result after the primary fusion, and constructing a judgment matrix with 3 rows and 3 columns according to a hierarchical analysis method. The matrix rows and columns correspond to the following table:

	road traffic state	Vehicle state	Driver state
				Road traffic state	1	1/4	1/3
Vehicle state	4	1	2
				Driver state	3	1/2	1

The correspondence matrix is as follows:

determining the feature weight vector W ═ (W) for each attribute₁，w₂，w₃)^T＝(0.122,0.558,0.320)^TWherein w is₁+w₂+w₃1, the dangerous driving comprehensive index P is obtained as 0.122. P1+ 0.558. P₂+0.320·P₃。

And judging whether the alarm is needed at present or not based on the dangerous driving comprehensive index P and a preset early warning threshold value, and executing a corresponding alarm process when needed.

Corresponding to the above method embodiment, the present invention further provides an embodiment of a driving behavior detection apparatus, including: a processor to:

acquiring current road traffic state information, vehicle state information and driver state information; respectively carrying out danger level standardization processing on the road traffic state information, the vehicle state information and the driver state information to obtain corresponding road traffic state vectors, vehicle state vectors and driver state vectors;

determining a current dangerous driving index of the road traffic state by using the road traffic state vector and a preset dangerous driving judgment matrix of the road traffic state; determining a current dangerous driving index of the vehicle state by using the vehicle state vector and a preset dangerous driving judgment matrix of the vehicle state; determining a current dangerous driving index of the driver state by using the driver state vector and a preset dangerous driving judgment matrix of the driver state;

determining a current dangerous driving comprehensive index by using the road traffic state dangerous driving index, the vehicle state dangerous driving index, the driver state dangerous driving index and a preset dangerous driving comprehensive judgment matrix;

and judging whether the alarm is needed at present or not according to the dangerous driving comprehensive index and a preset early warning threshold value, and executing a corresponding alarm process when needed.

Preferably, the processor is specifically configured to: respectively carrying out danger level standardization processing on the road traffic state information, the vehicle state information and the driver state information comprises the following steps:

determining a danger level corresponding to each attribute in the road traffic state information, the vehicle state information and the driver state information by using a preset dangerous driving level mapping relation; the dangerous driving grade mapping relation comprises a dangerous grade corresponding to each attribute value contained in each attribute of road traffic state information, vehicle state information and driver state information;

for the current road traffic state information, taking the danger level corresponding to each attribute as a vector element, and generating a road traffic state vector corresponding to the road traffic state information;

for the current vehicle state information, generating a vehicle state vector corresponding to the vehicle state information by taking the danger level corresponding to each attribute as a vector element;

and for the current driver state information, generating a driver state vector corresponding to the driver state information by taking the danger level corresponding to each attribute as a vector element.

Preferably, the attributes of the road traffic status information include: road type, road status, and weather status; the attributes of the vehicle state information include: overspeed, distance between front and rear vehicles, whether the vehicles slide in neutral gear, turning turn to turn lights and traffic lights; the attributes of the driver state information include: whether a safety belt is fastened, whether a steering wheel is correctly held, the yawning times, the eye closing time length, whether other equipment is used in the driving process and whether the sight line of a driver keeps watching the front in the driving process of the vehicle.

Preferably, the value range of the road type includes: town roads, urban roads, highways, mountain roads, and other road types;

the value range of the road state comprises: unobstructed, slight congestion, serious congestion and traffic accidents;

the value range of the weather state comprises: normal weather, sand-dust weather, rainy weather, snow weather and heavy fog weather, wherein the normal weather comprises sunny days and cloudy days;

the value range of the vehicle overspeed comprises the following steps: no overspeed, less than 10% overspeed, 10% to 30% overspeed, 30% to 50% overspeed and more than 50% overspeed;

the value range of the front and rear vehicle distances comprises: greater than 100 meters, greater than 80 meters and less than or equal to 100 meters, greater than 60 meters and less than or equal to 80 meters, greater than 50 meters and less than or equal to 60 meters, and less than or equal to 50 meters;

the value range of whether the neutral gear slides comprises the following steps: neutral coasting and no neutral coasting;

the value range of the turning turn signal lamp comprises the following steps: the steering lamp is used correctly, the steering lamp is not turned on, and the steering lamp is turned by mistake;

the value range of the traffic signal lamp comprises: green light running, yellow light running and red light running;

the value range of whether the safety belt is fastened comprises the following steps: belting and unbelting;

the value range of whether the steering wheel is correctly held comprises the following steps: a double-hand-held steering wheel, a single-hand-held steering wheel and a non-held steering wheel;

the value range of the yawning times comprises: the yawning times in 5 minutes are less than or equal to 2, the yawning times in 5 minutes are more than 3 and less than or equal to 5, the yawning times in 5 minutes are more than 5 and less than or equal to 7, and the yawning times in 5 minutes are more than 7;

the value range of the eye closing time comprises the following steps: less than 0.5 seconds, 1 second, 3 seconds, and 5 seconds;

the value range of whether other equipment is used in the driving process comprises the following steps: use and non-use of other devices;

whether the driver's sight keeps watching the value range in the place ahead during the vehicle is gone includes: stay ahead of gaze and not stay ahead of gaze.

Preferably, the danger levels corresponding to the town road, the city road, the expressway, the mountain road and other road types are respectively as follows: 1. 2,3,4, 2;

the corresponding danger levels of the unobstructed road, the slight congestion, the serious congestion and the traffic accident are respectively as follows: 1. 2,3 and 4;

the danger levels corresponding to normal weather, sand and dust weather, rainy weather, snowy weather and heavy fog weather respectively are as follows: 1. 2,3,4, 5;

the corresponding danger grades of the non-overspeed, the overspeed less than 10%, the overspeed 10% to 30%, the overspeed 30% to 50% and the overspeed more than 50% are respectively as follows: 1. 2,3,4, 5;

the danger levels respectively corresponding to more than 100 meters, more than 80 meters and less than or equal to 100 meters, more than 60 meters and less than or equal to 80 meters, more than 50 meters and less than or equal to 60 meters, and less than or equal to 50 meters are: 1. 2,3,4, 5;

the corresponding danger grades of the neutral gear sliding and the non-neutral gear sliding are respectively as follows: 4. 1;

the danger levels corresponding to the correct use of the steering lamp, the no-turn-on of the steering lamp and the wrong turn-on of the steering lamp are respectively as follows: 1. 3, 5;

the danger levels corresponding to the green light running, the yellow light running and the red light running are respectively as follows: 1. 3, 5;

the corresponding danger grades of the belted safety belt and the belted safety belt are respectively as follows: 1. 5;

the danger levels respectively corresponding to the double-hand-held steering wheel, the single-hand-held steering wheel and the steering wheel which is not held are as follows: 1. 3, 5;

the corresponding danger levels of the yawning times within 5 minutes being less than or equal to 2 times, the yawning times within 5 minutes being more than 3 times and less than or equal to 5 times, the yawning times within 5 minutes being more than 5 times and less than or equal to 7 times, and the yawning times within 5 minutes being more than 7 times are respectively as follows: 1. 2,3 and 5;

the danger grades corresponding to less than 0.5 second, 1 second, 3 seconds and 5 seconds are respectively as follows: 1. 2,3 and 5;

the risk levels respectively corresponding to the use of other equipment and the non-use of other equipment are respectively as follows: 4. 1;

the danger levels respectively corresponding to the fixation front and the fixation non-front are respectively as follows: 4. 1.

Preferably, the processor is specifically configured to: determining the current dangerous driving index of the road traffic state by using the road traffic state vector and a preset dangerous driving judgment matrix of the road traffic state comprises the following steps:

calculating the dangerous driving judgment matrix of the road traffic state

Characteristic weight vector W of₁＝(w₁₁，...，w_1n)^TWherein

n is the number of attributes contained in the road traffic state information; x_ijThe ratio of the influence degree of the attribute j and the attribute i of the road traffic state information on dangerous driving is obtained; i is more than or equal to 1 and less than or equal to n; j is more than or equal to 1 and less than or equal to n;

computing

Obtaining the dangerous driving index P of the current road traffic state₁Wherein v is_1iIs the road traffic state vector v₁＝(v₁₁，...，v_1n) The ith element in (1).

Preferably, the processor is specifically configured to: determining the current dangerous driving index of the vehicle state by using the vehicle state vector and a preset dangerous driving judgment matrix of the vehicle state comprises the following steps:

calculating the vehicle state dangerous driving judgment matrix

Characteristic weight vector W of₂＝(w₂₁，...，w_2k)^TWherein

k is the number of attributes contained in the vehicle state information; y is_ijThe ratio of the influence degree of the attribute j and the attribute i of the vehicle state information on dangerous driving is obtained; i is more than or equal to 1 and less than or equal to k; j is more than or equal to 1 and less than or equal to k;

computing

Obtaining the dangerous driving index P of the current vehicle state₂Wherein v is_2iIs the vehicle state vector v₂＝(v₂₁，...，v_2k) The ith element in (1).

Preferably, the processor is specifically configured to: determining a current dangerous driving index of the driver state by using the driver state vector and a preset dangerous driving judgment matrix of the driver state comprises:

calculating the driver state dangerous driving judgment matrix

Characteristic weight vector W of₃＝(w₃₁，...，w_3s)^TWherein

s is the number of attributes contained in the driver state information; z_ijThe ratio of the influence degree of the attribute j and the attribute i of the driver state information on dangerous driving respectively is obtained; i is more than or equal to 1 and less than or equal to s; j is more than or equal to 1 and less than or equal to s;

computing

Obtaining the dangerous driving index P of the current driver state₃Wherein v is_3iIs the driver state vector v₃＝(v₃₁，...，v_3s) The ith element in (1).

Preferably, the processor is specifically configured to: determining the current dangerous driving comprehensive index by using the road traffic state dangerous driving index, the vehicle state dangerous driving index, the driver state dangerous driving index and a preset dangerous driving comprehensive judgment matrix, wherein the step of determining the current dangerous driving comprehensive index comprises the following steps:

calculating the dangerous driving comprehensive judgment matrix

The weight vector W ═ W (W) of the features₁，w₂，w₃)^TWherein w is₁+w₂+w₃＝1；

Calculating P ═ w₁P₁+w₂P₂+w₃P₃Obtaining a dangerous driving index P of the current driver state, wherein P₁As the road traffic state dangerIndex of driving, P₂The dangerous driving index of the vehicle state is obtained; p₃And the dangerous driving index of the driver state is obtained.

Preferably, the processor is specifically configured to: judging whether the alarm is required at present according to the dangerous driving comprehensive index and a preset early warning threshold value comprises the following steps:

and if the dangerous driving comprehensive index is larger than a preset early warning threshold value, judging that the alarm is required.

The present application further provides a non-transitory computer readable storage medium storing instructions that, when executed by a processor, cause the processor to perform the steps of the method of early warning detection of driving behavior as described above.

The present application also provides an electronic device comprising the non-volatile computer-readable storage medium as described above, and the processor having access to the non-volatile computer-readable storage medium.

In summary, the above description is only a preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (22)

1. A driving behavior detection method, characterized by comprising:

acquiring current road traffic state information, vehicle state information and driver state information; respectively carrying out danger level standardization processing on the road traffic state information, the vehicle state information and the driver state information to obtain corresponding road traffic state vectors, vehicle state vectors and driver state vectors;

determining a current dangerous driving index of the road traffic state by using the road traffic state vector and a preset dangerous driving judgment matrix of the road traffic state; determining a current dangerous driving index of the vehicle state by using the vehicle state vector and a preset dangerous driving judgment matrix of the vehicle state; determining a current dangerous driving index of the driver state by using the driver state vector and a preset dangerous driving judgment matrix of the driver state;

determining a current dangerous driving comprehensive index by using the road traffic state dangerous driving index, the vehicle state dangerous driving index, the driver state dangerous driving index and a preset dangerous driving comprehensive judgment matrix;

and judging whether the alarm is needed at present or not according to the dangerous driving comprehensive index and a preset early warning threshold value, and executing a corresponding alarm process when needed.

2. The method of claim 1, wherein the respectively subjecting the road traffic status information, the vehicle status information, and the driver status information to hazard level normalization processing comprises:

determining a danger level corresponding to each attribute in the road traffic state information, the vehicle state information and the driver state information by using a preset dangerous driving level mapping relation; the dangerous driving grade mapping relation comprises a dangerous grade corresponding to each attribute value contained in each attribute of road traffic state information, vehicle state information and driver state information;

for the current road traffic state information, taking the danger level corresponding to each attribute as a vector element, and generating a road traffic state vector corresponding to the road traffic state information;

for the current vehicle state information, generating a vehicle state vector corresponding to the vehicle state information by taking the danger level corresponding to each attribute as a vector element;

and for the current driver state information, generating a driver state vector corresponding to the driver state information by taking the danger level corresponding to each attribute as a vector element.

3. The method of claim 1, wherein the attributes of the road traffic status information comprise: road type, road status, and weather status; the attributes of the vehicle state information include: overspeed, distance between front and rear vehicles, whether the vehicles slide in neutral gear, turning turn to turn lights and traffic lights; the attributes of the driver state information include: whether a safety belt is fastened, whether a steering wheel is correctly held, the yawning times, the eye closing time length, whether other equipment is used in the driving process and whether the sight line of a driver keeps watching the front in the driving process of the vehicle.

4. The method of claim 3, wherein the range of values for the road type comprises: town roads, urban roads, highways, mountain roads, and other road types;

the value range of the road state comprises: unobstructed, slight congestion, serious congestion and traffic accidents;

the value range of the weather state comprises: normal weather, sand-dust weather, rainy weather, snow weather and heavy fog weather, wherein the normal weather comprises sunny days and cloudy days;

the value range of the vehicle overspeed comprises the following steps: no overspeed, less than 10% overspeed, 10% to 30% overspeed, 30% to 50% overspeed and more than 50% overspeed;

the value range of the front and rear vehicle distances comprises: greater than 100 meters, greater than 80 meters and less than or equal to 100 meters, greater than 60 meters and less than or equal to 80 meters, greater than 50 meters and less than or equal to 60 meters, and less than or equal to 50 meters;

the value range of whether the neutral gear slides comprises the following steps: neutral coasting and no neutral coasting;

the value range of the turning turn signal lamp comprises the following steps: the steering lamp is used correctly, the steering lamp is not turned on, and the steering lamp is turned by mistake;

the value range of the traffic signal lamp comprises: green light running, yellow light running and red light running;

the value range of whether the safety belt is fastened comprises the following steps: belting and unbelting;

the value range of whether the steering wheel is correctly held comprises the following steps: a double-hand-held steering wheel, a single-hand-held steering wheel and a non-held steering wheel;

the value range of the yawning times comprises: the yawning times in 5 minutes are less than or equal to 2, the yawning times in 5 minutes are more than 3 and less than or equal to 5, the yawning times in 5 minutes are more than 5 and less than or equal to 7, and the yawning times in 5 minutes are more than 7;

the value range of the eye closing time comprises the following steps: less than 0.5 seconds, 1 second, 3 seconds, and 5 seconds;

the value range of whether other equipment is used in the driving process comprises the following steps: use and non-use of other devices;

whether the driver's sight keeps watching the value range in the place ahead during the vehicle is gone includes: stay ahead of gaze and not stay ahead of gaze.

5. The method of claim 4,

the danger grades corresponding to the town road, the city road, the highway, the mountain road and other road types are respectively as follows: 1. 2,3,4, 2;

the corresponding danger levels of the unobstructed road, the slight congestion, the serious congestion and the traffic accident are respectively as follows: 1. 2,3 and 4;

the danger levels corresponding to normal weather, sand and dust weather, rainy weather, snowy weather and heavy fog weather respectively are as follows: 1. 2,3,4, 5;

the corresponding danger grades of the non-overspeed, the overspeed less than 10%, the overspeed 10% to 30%, the overspeed 30% to 50% and the overspeed more than 50% are respectively as follows: 1. 2,3,4, 5;

the danger levels respectively corresponding to more than 100 meters, more than 80 meters and less than or equal to 100 meters, more than 60 meters and less than or equal to 80 meters, more than 50 meters and less than or equal to 60 meters, and less than or equal to 50 meters are: 1. 2,3,4, 5;

the corresponding danger grades of the neutral gear sliding and the non-neutral gear sliding are respectively as follows: 4. 1;

the danger levels corresponding to the correct use of the steering lamp, the no-turn-on of the steering lamp and the wrong turn-on of the steering lamp are respectively as follows: 1. 3, 5;

the danger levels corresponding to the green light running, the yellow light running and the red light running are respectively as follows: 1. 3, 5;

the corresponding danger grades of the belted safety belt and the belted safety belt are respectively as follows: 1. 5;

the danger levels respectively corresponding to the double-hand-held steering wheel, the single-hand-held steering wheel and the steering wheel which is not held are as follows: 1. 3, 5;

the corresponding danger levels of the yawning times within 5 minutes being less than or equal to 2 times, the yawning times within 5 minutes being more than 3 times and less than or equal to 5 times, the yawning times within 5 minutes being more than 5 times and less than or equal to 7 times, and the yawning times within 5 minutes being more than 7 times are respectively as follows: 1. 2,3 and 5;

the danger grades corresponding to less than 0.5 second, 1 second, 3 seconds and 5 seconds are respectively as follows: 1. 2,3 and 5;

the risk levels respectively corresponding to the use of other equipment and the non-use of other equipment are respectively as follows: 4. 1;

the danger levels respectively corresponding to the fixation front and the fixation non-front are respectively as follows: 4. 1.

6. The method of claim 1, wherein determining the current dangerous driving index of the road traffic state by using the road traffic state vector and a preset dangerous driving judgment matrix of the road traffic state comprises:

calculating the dangerous driving judgment matrix of the road traffic state

Characteristic weight vector W of₁＝(w₁₁，...，w_1n)^TWherein

n is the number of attributes contained in the road traffic state information; x_ijThe ratio of the influence degree of the attribute j and the attribute i of the road traffic state information on dangerous driving is obtained; i is more than or equal to 1 and less than or equal to n; j is more than or equal to 1 and less than or equal to n;

computing

Obtaining the dangerous driving index P of the current road traffic state₁Wherein v is_1iIs the road traffic state vector v₁＝(v₁₁，...，v_1n) The ith element in (1).

7. The method of claim 1, wherein determining the current dangerous driving index of the vehicle state by using the vehicle state vector and a preset dangerous driving judgment matrix of the vehicle state comprises:

calculating the vehicle state dangerous driving judgment matrix

Characteristic weight vector W of₂＝(w₂₁，...，w_2k)^TWherein

k is the number of attributes contained in the vehicle state information; y is_ijThe ratio of the influence degree of the attribute j and the attribute i of the vehicle state information on dangerous driving is obtained; i is more than or equal to 1 and less than or equal to k; j is more than or equal to 1 and less than or equal to k;

computing

Obtaining the dangerous driving index P of the current vehicle state₂Wherein v is_2iIs the vehicle state vector v₂＝(v₂₁，...，v_2k) The ith element in (1).

8. The method of claim 1, wherein determining a current driver state dangerous driving index using the driver state vector and a preset driver state dangerous driving decision matrix comprises:

calculating the driver state dangerous driving judgment matrix

Characteristic weight vector W of₃＝(w₃₁，...，w_3s)^TWherein

s is the number of attributes contained in the driver state information; z_ijThe ratio of the influence degree of the attribute j and the attribute i of the driver state information on dangerous driving respectively is obtained; i is more than or equal to 1 and less than or equal to s; j is more than or equal to 1 and less than or equal to s;

computing

Obtaining the dangerous driving index P of the current driver state₃Wherein v is_3iIs the driver state vector v₃＝(v₃₁，...，v_3s) The ith element in (1).

9. The method of claim 1, wherein the determining a current dangerous driving comprehensive index by using the road traffic state dangerous driving index, the vehicle state dangerous driving index, the driver state dangerous driving index and a preset dangerous driving comprehensive judgment matrix comprises:

calculating the dangerous driving comprehensive judgment matrix

The weight vector W ═ W (W) of the features₁，w₂，w₃)^TWherein w is₁+w₂+w₃＝1；

Calculating P ═ w₁P₁+w₂P₂+w₃P₃Obtaining a dangerous driving index P of the current driver state, wherein P₁Is the dangerous driving index, P, of the road traffic state₂The dangerous driving index of the vehicle state is obtained; p₃And the dangerous driving index of the driver state is obtained.

10. The method of claim 1, wherein the determining whether an alarm is currently required according to the dangerous driving combination index and a preset early warning threshold comprises:

and if the dangerous driving comprehensive index is larger than a preset early warning threshold value, judging that the alarm is required.

11. A driving behavior detection apparatus, characterized by comprising: a processor to:

acquiring current road traffic state information, vehicle state information and driver state information; respectively carrying out danger level standardization processing on the road traffic state information, the vehicle state information and the driver state information to obtain corresponding road traffic state vectors, vehicle state vectors and driver state vectors;

determining a current dangerous driving index of the road traffic state by using the road traffic state vector and a preset dangerous driving judgment matrix of the road traffic state; determining a current dangerous driving index of the vehicle state by using the vehicle state vector and a preset dangerous driving judgment matrix of the vehicle state; determining a current dangerous driving index of the driver state by using the driver state vector and a preset dangerous driving judgment matrix of the driver state;

determining a current dangerous driving comprehensive index by using the road traffic state dangerous driving index, the vehicle state dangerous driving index, the driver state dangerous driving index and a preset dangerous driving comprehensive judgment matrix;

and judging whether the alarm is needed at present or not according to the dangerous driving comprehensive index and a preset early warning threshold value, and executing a corresponding alarm process when needed.

12. The device of claim 11, wherein the processor is specifically configured to: respectively carrying out danger level standardization processing on the road traffic state information, the vehicle state information and the driver state information comprises the following steps:

determining a danger level corresponding to each attribute in the road traffic state information, the vehicle state information and the driver state information by using a preset dangerous driving level mapping relation; the dangerous driving grade mapping relation comprises a dangerous grade corresponding to each attribute value contained in each attribute of road traffic state information, vehicle state information and driver state information;

for the current road traffic state information, taking the danger level corresponding to each attribute as a vector element, and generating a road traffic state vector corresponding to the road traffic state information;

for the current vehicle state information, generating a vehicle state vector corresponding to the vehicle state information by taking the danger level corresponding to each attribute as a vector element;

and for the current driver state information, generating a driver state vector corresponding to the driver state information by taking the danger level corresponding to each attribute as a vector element.

13. The apparatus of claim 11, wherein the attributes of the road traffic status information comprise: road type, road status, and weather status; the attributes of the vehicle state information include: overspeed, distance between front and rear vehicles, whether the vehicles slide in neutral gear, turning turn to turn lights and traffic lights; the attributes of the driver state information include: whether a safety belt is fastened, whether a steering wheel is correctly held, the yawning times, the eye closing time length, whether other equipment is used in the driving process and whether the sight line of a driver keeps watching the front in the driving process of the vehicle.

14. The apparatus of claim 13, wherein the range of values for the road type comprises: town roads, urban roads, highways, mountain roads, and other road types;

the value range of the road state comprises: unobstructed, slight congestion, serious congestion and traffic accidents;

the value range of the weather state comprises: normal weather, sand-dust weather, rainy weather, snow weather and heavy fog weather, wherein the normal weather comprises sunny days and cloudy days;

the value range of the vehicle overspeed comprises the following steps: no overspeed, less than 10% overspeed, 10% to 30% overspeed, 30% to 50% overspeed and more than 50% overspeed;

the value range of the front and rear vehicle distances comprises: greater than 100 meters, greater than 80 meters and less than or equal to 100 meters, greater than 60 meters and less than or equal to 80 meters, greater than 50 meters and less than or equal to 60 meters, and less than or equal to 50 meters;

the value range of whether the neutral gear slides comprises the following steps: neutral coasting and no neutral coasting;

the value range of the turning turn signal lamp comprises the following steps: the steering lamp is used correctly, the steering lamp is not turned on, and the steering lamp is turned by mistake;

the value range of the traffic signal lamp comprises: green light running, yellow light running and red light running;

the value range of whether the safety belt is fastened comprises the following steps: belting and unbelting;

the value range of whether the steering wheel is correctly held comprises the following steps: a double-hand-held steering wheel, a single-hand-held steering wheel and a non-held steering wheel;

the value range of the yawning times comprises: the yawning times in 5 minutes are less than or equal to 2, the yawning times in 5 minutes are more than 3 and less than or equal to 5, the yawning times in 5 minutes are more than 5 and less than or equal to 7, and the yawning times in 5 minutes are more than 7;

the value range of the eye closing time comprises the following steps: less than 0.5 seconds, 1 second, 3 seconds, and 5 seconds;

the value range of whether other equipment is used in the driving process comprises the following steps: use and non-use of other devices;

whether the driver's sight keeps watching the value range in the place ahead during the vehicle is gone includes: stay ahead of gaze and not stay ahead of gaze.

15. The apparatus of claim 14,

the danger grades corresponding to the town road, the city road, the highway, the mountain road and other road types are respectively as follows: 1. 2,3,4, 2;

the corresponding danger levels of the unobstructed road, the slight congestion, the serious congestion and the traffic accident are respectively as follows: 1. 2,3 and 4;

the danger levels corresponding to normal weather, sand and dust weather, rainy weather, snowy weather and heavy fog weather respectively are as follows: 1. 2,3,4, 5;

the corresponding danger grades of the non-overspeed, the overspeed less than 10%, the overspeed 10% to 30%, the overspeed 30% to 50% and the overspeed more than 50% are respectively as follows: 1. 2,3,4, 5;

the danger levels respectively corresponding to more than 100 meters, more than 80 meters and less than or equal to 100 meters, more than 60 meters and less than or equal to 80 meters, more than 50 meters and less than or equal to 60 meters, and less than or equal to 50 meters are: 1. 2,3,4, 5;

the corresponding danger grades of the neutral gear sliding and the non-neutral gear sliding are respectively as follows: 4. 1;

the danger levels corresponding to the correct use of the steering lamp, the no-turn-on of the steering lamp and the wrong turn-on of the steering lamp are respectively as follows: 1. 3, 5;

the danger levels corresponding to the green light running, the yellow light running and the red light running are respectively as follows: 1. 3, 5;

the corresponding danger grades of the belted safety belt and the belted safety belt are respectively as follows: 1. 5;

the danger levels respectively corresponding to the double-hand-held steering wheel, the single-hand-held steering wheel and the steering wheel which is not held are as follows: 1. 3, 5;

the corresponding danger levels of the yawning times within 5 minutes being less than or equal to 2 times, the yawning times within 5 minutes being more than 3 times and less than or equal to 5 times, the yawning times within 5 minutes being more than 5 times and less than or equal to 7 times, and the yawning times within 5 minutes being more than 7 times are respectively as follows: 1. 2,3 and 5;

the danger grades corresponding to less than 0.5 second, 1 second, 3 seconds and 5 seconds are respectively as follows: 1. 2,3 and 5;

the risk levels respectively corresponding to the use of other equipment and the non-use of other equipment are respectively as follows: 4. 1;

the danger levels respectively corresponding to the fixation front and the fixation non-front are respectively as follows: 4. 1.

16. The device of claim 11, wherein the processor is specifically configured to: determining the current dangerous driving index of the road traffic state by using the road traffic state vector and a preset dangerous driving judgment matrix of the road traffic state comprises the following steps:

calculating the dangerous driving judgment matrix of the road traffic state

Characteristic weight vector W of₁＝(w₁₁，...，w_1n)^TWherein

n is the number of attributes contained in the road traffic state information; x_ijThe ratio of the influence degree of the attribute j and the attribute i of the road traffic state information on dangerous driving is obtained; i is more than or equal to 1 and less than or equal to n; j is more than or equal to 1 and less than or equal to n;

computing

Obtaining the dangerous driving index P of the current road traffic state₁Wherein v is_1iIs the road traffic state vector v₁＝(v₁₁，...，v_1n) The ith element in (1).

17. The device of claim 11, wherein the processor is specifically configured to: determining the current dangerous driving index of the vehicle state by using the vehicle state vector and a preset dangerous driving judgment matrix of the vehicle state comprises the following steps:

calculating the vehicle state dangerous driving judgment matrix

Characteristic weight vector W of₂＝(w₂₁，...，w_2k)^TWherein

k is the number of attributes contained in the vehicle state information; y is_ijThe ratio of the influence degree of the attribute j and the attribute i of the vehicle state information on dangerous driving is obtained; i is more than or equal to 1 and less than or equal to k; j is more than or equal to 1 and less than or equal to k;

computing

Obtaining the dangerous driving index P of the current vehicle state₂Wherein v is_2iIs the vehicle state vector v₂＝(v₂₁，...，v_2k) The ith element in (1).

18. The device of claim 11, wherein the processor is specifically configured to: determining a current dangerous driving index of the driver state by using the driver state vector and a preset dangerous driving judgment matrix of the driver state comprises:

calculating the driver state dangerous driving judgment matrix

Characteristic weight vector W of₃＝(w₃₁，...，w_3s)^TWherein

s is the number of attributes contained in the driver state information; z_ijThe ratio of the influence degree of the attribute j and the attribute i of the driver state information on dangerous driving respectively is obtained; i is more than or equal to 1 and less than or equal to s; j is more than or equal to 1 and less than or equal to s;

computing

Obtaining the dangerous driving index P of the current driver state₃Wherein v is_3iIs the driver state vector v₃＝(v₃₁，...，v_3s) The ith element in (1).

19. The device of claim 11, wherein the processor is specifically configured to: determining the current dangerous driving comprehensive index by using the road traffic state dangerous driving index, the vehicle state dangerous driving index, the driver state dangerous driving index and a preset dangerous driving comprehensive judgment matrix, wherein the step of determining the current dangerous driving comprehensive index comprises the following steps:

calculating the dangerous driving comprehensive judgment matrix

The weight vector W ═ W (W) of the features₁，w₂，w₃)^TWherein w is₁+w₂+w₃＝1；

Calculating P ═ w₁P₁+w₂P₂+w₃P₃Obtaining a dangerous driving index P of the current driver state, wherein P₁Is the dangerous driving index, P, of the road traffic state₂The dangerous driving index of the vehicle state is obtained; p₃And the dangerous driving index of the driver state is obtained.

20. The device of claim 5, wherein the processor is specifically configured to: judging whether the alarm is required at present according to the dangerous driving comprehensive index and a preset early warning threshold value comprises the following steps:

and if the dangerous driving comprehensive index is larger than a preset early warning threshold value, judging that the alarm is required.

21. A non-transitory computer readable storage medium storing instructions which, when executed by a processor, cause the processor to perform the steps of the warning detection method of driving behaviour according to any one of claims 1 to 10.

22. An electronic device comprising the non-volatile computer-readable storage medium of claim 21, and the processor having access to the non-volatile computer-readable storage medium.

Images