A method and system for identifying misoperation behaviors of vehicle driver pedals

Abstract

The present invention provides a method and system for identifying misoperation behaviors of an automobile driver. The present invention mainly determines the safety state of the automobile and the driver state based on various parameters of the collected automobile (including the opening degree information of the automobile accelerator pedal, the distance information between the automobile and the obstacle in front, and the steering wheel angle information or the steering wheel angular velocity information) and the driver's line-of-sight fixation information, and can analyze and determine whether the situation of stepping on the accelerator pedal to the bottom is the driver's subjective urgent acceleration requirement or the driver's stepping error. In addition to being able to utilize the determination rules and thresholds suitable for identifying abnormal driving behaviors of the vehicle itself, the present invention can also quantitatively determine the abnormal determination criteria based on the line-of-sight confusion degree and the Gini impurity parameter, can effectively improve the accuracy of identification, is closer to the abnormal driving behavior law of the driver, and can actively control the vehicle in time when encountering the misoperation behavior of the driver, so as to avoid the occurrence of safety accidents caused by accidentally stepping on the accelerator pedal.

Description

Technical Field

[0001] The present invention relates to the technical field of identifying abnormal driving behaviors of drivers, and particularly to a method and system for identifying misoperation behaviors of automotive driver pedals based on multi-source information. Background Art

[0002] In recent years, the automotive industry in China has developed rapidly, the number of motor vehicles in possession has been increasing continuously, various traffic safety risks are intertwined and superimposed, and the task of preventing traffic accidents is extremely arduous. Usually, when a driver is driving a vehicle, if an obstacle with potential safety hazards suddenly appears in front of the vehicle, the driver should directly step on the brake pedal for braking. However, there are frequent occurrences of misoperation behaviors in which the driver mistakes the accelerator pedal for the brake pedal and steps on it all the way in an emergency. Due to the good acceleration ability of the vehicle, there are huge potential safety hazards for traffic accidents.

[0003] Chinese invention patent CN115091948A discloses a device and method for promptly detecting and correcting missteps of the accelerator pedal. When the depression opening of the accelerator pedal exceeds the threshold of the preset depression opening, and the depression speed of the accelerator pedal exceeds the threshold of the preset depression speed, and the driver's attention state is a distracted attention state, and there is a collision risk for the vehicle, it is determined that the driver has misstepped the accelerator pedal; otherwise, it is determined that the driver has not misstepped the accelerator pedal. In this technical solution, the recognition of the driver's attention state only distinguishes between a distracted attention state and a concentrated attention state, and at the same time, the recognition method of the driver's line-of-sight information is not fine enough. In complex driving scenarios where the driver's head needs to turn frequently and significantly, such as in urban commuting and mountain roads, the vehicle cannot make full use of the driver's line-of-sight information to reduce the misjudgment probability. Summary of the Invention

[0004] Aiming at the problems existing in the above-mentioned prior art, the purpose of the present invention is to provide a method and system for identifying misoperation behaviors of automotive driver pedals, which can accurately identify abnormal pedal operation behaviors of the driver based on multi-source information, warn the driver in a flustered state during misoperation, and actively intervene in the driver's control of the vehicle, which will be of great significance to the driving safety of the vehicle.

[0005] The present invention is achieved through the following technical solutions:

[0006] A system for identifying misoperation behaviors of automotive driver pedals includes: an information acquisition module, a central control module, and an execution module;

[0007] The information acquisition module is used to collect information on the opening of the automotive accelerator pedal, information on the driver's gaze area, information on the distance between the vehicle and the obstacle in front, and information on the steering wheel angle or the angular velocity of the steering wheel rotation, and transmit the collected information to the central control module;

[0008] The central control module is used to calculate the line-of-sight confusion degree or Gini impurity based on the driver's gaze area information, and judge whether stepping on the accelerator pedal is an abnormal behavior according to the accelerator pedal opening information, the line-of-sight confusion degree or Gini impurity, the distance information of the vehicle from the obstacle ahead, and the steering wheel angle information or steering wheel rotation angular velocity information; the line-of-sight confusion degree is the weighted value of the driver's gaze area within one sampling period.

[0009] The execution module is used to prompt the driver for misstep of the pedal and / or control the vehicle braking when the judgment result of the central control module is that stepping on the accelerator pedal is an abnormal behavior.

[0010] Preferably, the information acquisition module includes an eye tracker, and the eye tracker is used to acquire the driver's gaze area information.

[0011] And / or, the information acquisition module includes a millimeter-wave radar, and the millimeter-wave radar is used to acquire the distance information of the vehicle from the obstacle ahead.

[0012] Preferably, the information acquisition module includes an accelerator pedal position sensor, and the accelerator pedal position sensor is used to acquire the accelerator pedal opening information.

[0013] And / or, the information acquisition module includes a steering wheel angle sensor, and the steering wheel angle sensor is used to acquire the steering wheel angle information or steering wheel rotation angular velocity information.

[0014] Preferably, the execution module includes a stepper motor driver, a stepper motor, a vibration motor and a speaker installed in the steering wheel; the stepper motor driver and the stepper motor are used to control the vehicle braking; the vibration motor and the speaker are used to prompt the driver for misstep of the pedal.

[0015] A method for identifying misoperation behavior of an automotive driver's pedal includes:

[0016] S1: Acquire the accelerator pedal opening information, the driver's gaze area information, the distance information of the vehicle from the obstacle ahead, and the steering wheel angle information or steering wheel rotation angular velocity information of the vehicle.

[0017] S2: Calculate the line-of-sight confusion degree or Gini impurity based on the driver's gaze area information, and judge whether stepping on the accelerator pedal is an abnormal behavior according to the accelerator pedal opening information, the line-of-sight confusion degree or Gini impurity, the distance information of the vehicle from the obstacle ahead, and the steering wheel angle information or steering wheel rotation angular velocity information. If the judgment result is that stepping on the accelerator pedal is an abnormal behavior, then execute S3; wherein, the line-of-sight confusion degree is the weighted value of the driver's gaze area within one sampling period.

[0018] S3: Prompt the driver for misstep of the pedal and / or control the vehicle braking.

[0019] Preferably, in S2, the specific calculation of the line-of-sight confusion degree based on the driver's gaze area information is as follows: According to the areas gazed at by the driver within one sampling period and the gazing time of different areas, calculate the weighted value of the area weight and the gazing time for each area to obtain the line-of-sight confusion degree;

[0020] The areas are obtained by partitioning the driver's line-of-sight area during driving, and the area weights are preset in advance.

[0021] Preferably, in S2, the specific calculation of the Gini impurity based on the driver's gaze area information is as follows:

[0022] According to the areas gazed at by the driver within one sampling period and the gazing time of different areas, calculate the percentage value of the gazing time of each area in the sampling time as the gazing probability of this area, and calculate the Gini impurity I(f) through the following formula:

[0023]

[0024] where m is the number of areas, and f i is the gazing frequency of each area.

[0025] Preferably, in S2, the specific determination of whether stepping on the accelerator pedal is an abnormal behavior is as follows:

[0026] If the accelerator pedal opening is greater than the preset accelerator pedal opening threshold and the duration is greater than the preset accelerator pedal time threshold; and the distance between the vehicle and the obstacle in front is less than the preset distance threshold; and the line-of-sight confusion degree is greater than the preset line-of-sight confusion degree threshold or the Gini impurity is less than the preset Gini impurity threshold; and the angular velocity of the steering wheel rotation is 0 and the duration is greater than the preset steering wheel time threshold; then stepping on the accelerator pedal is considered an abnormal behavior;

[0027] Or, if the accelerator pedal opening is greater than the preset accelerator pedal opening threshold and the duration is greater than the preset accelerator pedal time threshold; and the distance between the vehicle and the obstacle in front is less than the preset distance threshold; and the line-of-sight confusion degree is greater than the preset line-of-sight confusion degree threshold or the Gini impurity is less than the preset Gini impurity threshold; and the angular velocity of the steering wheel rotation is greater than the preset steering wheel rotation angular velocity threshold; then stepping on the accelerator pedal is considered an abnormal behavior.

[0028] Furthermore, the distance threshold is preset according to the vehicle speed of the vehicle.

[0029] Preferably, S3 is specifically as follows: Prompt the driver that the pedal is misstepped. If the driver does not end the operation of misstepping the pedal when the prompt time reaches the preset warning time threshold, then control the vehicle to brake.

[0030] Compared with the prior art, the present invention has the following beneficial effects:

[0031] The identification method of the present invention mainly determines the safety state of the vehicle and the driver's state based on various parameters of the collected vehicle (including the opening degree information of the vehicle's accelerator pedal, the distance information between the vehicle and the obstacle ahead, and the steering wheel angle information or the angular velocity information of the steering wheel rotation) and the driver's line-of-sight fixation information, and can analyze and determine whether the situation of stepping on the accelerator pedal to the bottom is the driver's subjective urgent acceleration demand or the driver's stepping error. In addition to being able to use the determination rules and thresholds suitable for identifying the driver's abnormal driving behavior of the vehicle itself, this method can also quantitatively determine the abnormal determination criteria based on the line-of-sight confusion degree and the Gini impurity parameter, and at the same time, refine the identification and analysis of the driver's real-time line-of-sight fixation behavior. The abnormal operation behavior identification method based on multi-source information can effectively improve the accuracy of identification, be closer to the law of the driver's abnormal driving behavior, and can actively control the vehicle in time when encountering the driver's misoperation behavior, avoiding the phenomenon of safety accidents caused by accidentally stepping on the accelerator pedal. Using the method of the present invention has practical application value.

[0032] Furthermore, the fact that the opening degree of the accelerator pedal is greater than the accelerator pedal opening degree threshold and the duration exceeds the accelerator pedal time threshold does not directly indicate that the driver accidentally stepped on the accelerator pedal. It is also possible to be a subjective urgent acceleration behavior. When the vehicle is driving normally, if the driver's operations are all normal, even if suddenly stepping on the accelerator pedal hard, the driver's head posture and the line-of-sight fixation area will remain relatively stable, and there will be no large-scale irregular head rotation and the fixation area will not flutter. On the contrary, during abnormal driving, due to the driver's inner panic, the sudden acceleration of the vehicle when accidentally stepping on the accelerator pedal makes the driver's front vision change extremely fast. At this time, the driver's fixation area and the line-of-sight direction will flutter, and the driver's line of sight will subconsciously turn to the driving path that can drive safely to avoid vehicle out of control as much as possible. When the vehicle is driving normally, if the driver's operations are all normal, even if suddenly stepping on the accelerator pedal hard, the steering wheel angle and the angular velocity of rotation will be slightly fluctuating and there will be no large-scale mutations. On the contrary, during misoperation, due to mental panic, the steering wheel angle and the angular velocity of rotation will show extreme situations, either remaining stationary or having large fluctuations. Therefore, the present invention comprehensively judges whether the operation of stepping on the accelerator pedal is abnormal according to the comparison results of these several kinds of information, which can ensure the accuracy of the judgment result. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is a schematic flowchart of an embodiment of the method for identifying the misoperation behavior of the vehicle driver's pedal based on multi-source information of the present invention.

[0034] Figure 2 It is a schematic diagram of the installation position of the in-vehicle sensors in an embodiment of the method for identifying the misoperation behavior of the vehicle driver's pedal based on multi-source information of the present invention.

[0035] Figure 3 This is a schematic diagram of the division of the driver's gaze area and the corresponding weights of each area when calculating the visual confusion degree R and the Gini impurity I(f) in an embodiment of the method for identifying the misoperation behavior of the automotive driver's pedal based on multi-source information according to the present invention. Detailed implementation manners

[0036] To further understand the present invention, the present invention will be described below in conjunction with embodiments. These descriptions are only to further explain the features and advantages of the present invention and are not used to limit the claims of the present invention.

[0037] An automotive driver's pedal misoperation behavior recognition system based on multi-source information according to the present invention includes: an information acquisition module, a central control module, and an execution module;

[0038] The information acquisition module is used to acquire the vehicle speed information, the accelerator pedal opening information, the driver's gaze area information, the distance information between the vehicle and the obstacle ahead, and the steering wheel angle information or the steering wheel rotation angular velocity information of the vehicle, and transmit the acquired information to the central control module;

[0039] The central control module is used to judge whether stepping on the accelerator pedal is an abnormal behavior according to the accelerator pedal opening information, the driver's gaze area information, the distance information between the vehicle and the obstacle ahead, and the steering wheel angle information or the steering wheel rotation angular velocity information;

[0040] The execution module is used to prompt the driver for misstepping the pedal and / or control the vehicle braking when the judgment result of the central control module is that stepping on the accelerator pedal is an abnormal behavior.

[0041] The information acquisition module includes a CAN bus, which is connected to the vehicle body bus. It also includes an eye tracker installed on the vehicle dashboard, a millimeter wave radar installed at the front of the vehicle, an accelerator pedal position sensor installed inside the vehicle, and a steering wheel angle sensor installed inside the vehicle. The eye tracker, the millimeter wave radar, the accelerator pedal sensor, and the steering wheel rotation angular velocity sensor are all signal-connected to the vehicle body bus.

[0042] The eye tracker is used to acquire the driver's gaze area information. By using the eye tracker to obtain the change of the fixation points during the driver's driving process, the viewpoint characteristics of the driver in different sections can be obtained, which can better characterize the driver's visual information processing process.

[0043] The millimeter wave radar is used to acquire the distance information between the vehicle and the obstacle ahead;

[0044] The accelerator pedal position sensor is used to acquire the accelerator pedal opening information.

[0045] The steering wheel angle sensor is used to acquire the steering wheel angle information or the steering wheel rotation angular velocity information.

[0046] The central control module is the ECU of the vehicle, which is a control circuit with an embedded microprocessor as the core and is used for demonstration.

[0047] The execution module includes a stepper motor driver, a high-torque stepper motor, a vibration motor and a speaker installed in the steering wheel. The stepper motor driver and the high-torque stepper motor are used to control vehicle braking. The vibration motor and the speaker are used to prompt the driver for misstepping on the pedal.

[0048] A method for identifying misoperation behavior of vehicle driver pedals based on multi-source information according to the present invention, such as Figure 1 , includes the following steps:

[0049] S1: Collect the vehicle speed information, accelerator pedal opening information, driver's gaze area information, distance information of the vehicle from the obstacle ahead, and steering wheel angle information or steering wheel angular velocity information of the vehicle.

[0050] S2: According to the accelerator pedal opening information, driver's gaze area information, distance information of the vehicle from the obstacle ahead, and steering wheel angle information or steering wheel angular velocity information, judge whether the situation of stepping on the accelerator pedal to the bottom is the driver's subjective urgent acceleration demand or the driver's stepping error, that is, judge whether stepping on the accelerator pedal is an abnormal behavior.

[0051] S3: When the judgment result in S2 is that stepping on the accelerator pedal is an abnormal behavior, prompt the driver for misstepping on the pedal and / or control vehicle braking (actively stop the motor from working).

[0052] An embodiment of the present invention provides a method for identifying misoperation behavior of vehicle driver pedals based on multi-source information. This embodiment is applicable to all vehicles with an accelerator pedal and a brake pedal, and includes the following steps:

[0053] Step 1, collect the opening information of the accelerator pedal and compare it with the set accelerator pedal opening threshold.

[0054] Specifically, collect the accelerator pedal opening information through the CAN bus connected to the vehicle body bus. The ECU compares the real-time pedal opening information with the set accelerator pedal opening threshold x and accelerator pedal time threshold t in each unit time. When the accelerator pedal opening exceeds the accelerator pedal opening threshold x and the duration exceeds the accelerator pedal time threshold t, it is considered that the driver has the behavior of stepping on the accelerator pedal.

[0055] In this embodiment, the accelerator pedal opening threshold is set to 98%, and the accelerator pedal time threshold is set to 3s.

[0056] When the accelerator pedal opening is greater than the accelerator pedal opening threshold x and the duration exceeds the accelerator pedal time threshold t, it does not directly indicate that the driver has accidentally stepped on the accelerator pedal. It may also be a subjective hard acceleration behavior. To determine whether the driver's behavior is an abnormal behavior of accidentally stepping on the accelerator pedal or a subjective hard acceleration behavior, the following steps are continued.

[0057] Step 2: According to the distance between the vehicle and the obstacle in front, the driver's line-of-sight confusion degree, the Gini impurity, and the steering wheel angular velocity information, further determine whether the driver's behavior of stepping on the accelerator pedal is a subjective hard acceleration start or an abnormal start operation, and determine the recognition conditions for the driver's abnormal start operation behavior.

[0058] Step 2 specifically includes the following sub-steps:

[0059] Sub-step 2.1: Preset the distance threshold between the vehicle and the obstacle in front, the driver's line-of-sight confusion degree threshold, and the steering wheel rotation angular velocity threshold;

[0060] Specifically:

[0061] 2.1a: Preset the distance threshold between the vehicle and the obstacle in front.

[0062] The distance between the vehicle and the obstacle in front is the setting of measuring surrounding dangerous objects (pedestrians, vehicles, distance from the roadside) through the millimeter-wave radar in front of the vehicle. According to different vehicle speeds, the distance threshold between the vehicle and the obstacle in front is different. When the distance between the vehicle and the obstacle in front is less than the distance threshold under different working conditions, it is a dangerous distance. The specific setting of the distance threshold is as follows:

[0063] (1) When the vehicle speed is above 100 km / h, the distance threshold is preset to be above 100 m;

[0064] (2) When the vehicle speed is between 60 km / h and 100 km / h, the distance threshold is preset to be not less than 60 m;

[0065] (3) When the vehicle speed is between 50 km / h and 60 km / h, the distance threshold is preset to be not less than 50 m;

[0066] (4) When the vehicle speed is between 40 km / h and 50 km / h, the distance threshold is preset to be not less than 30 m;

[0067] (5) When the vehicle speed is between 20 km / h and 40 km / h, the distance threshold is preset to be not less than 10 m.

[0068] 2.1b: Determine the driver's line-of-sight confusion degree threshold;

[0069] In this embodiment, the adjacent sampling time is greater than or equal to 2 s, that is, the sampling period is at least 2 s at a time.

[0070] Sub-step ①: Within a sampling period of 2 s, an eye tracker installed on the vehicle instrument panel (such as Figure 2 ) captures the facial image information of the driver from the front, monitors the head posture of the driver, and tracks the driver's line of sight;

[0071] Sub-step ②: Obtain the fixation area of the driver from the facial image information of the driver. According to the fixation time of the driver in different fixation areas, determine whether the driver has dangerous driving behaviors such as erratic line of sight and randomly looking around the vehicle.

[0072] Specifically, when the vehicle is driving normally, if the driver's operations are all normal, even when suddenly stepping on the accelerator pedal hard to drive, the head posture and the line-of-sight fixation area of the driver will remain relatively stable, and there will be no large-scale irregular rotation of the head and no erratic fixation area. On the contrary, during abnormal driving, due to the driver's inner panic, when accidentally stepping on the accelerator pedal for 3 s or more and the vehicle accelerates sharply, the driver's forward field of view changes extremely fast. At this time, the driver's fixation area and the direction of the line of sight will be erratic, and the driver's line of sight will subconsciously turn to the driving path that can ensure safe driving to avoid vehicle out of control as much as possible.

[0073] In the embodiment of the present invention, sub-step ② includes two judgment methods: chaos degree and Gini impurity.

[0074] Method 1: Chaos degree method

[0075] This method specifically includes the following sub-steps:

[0076] Sub-step 1) Divide the line-of-sight area of the driver's driving into 5 blocks in advance (see Figure 3 ), specifically including the left rearview mirror as area 1, the area directly in front of the driver's driving vision in the front windshield as area 2, area 2 is set as 1 / 3 of the area of the front windshield, the area in the front windshield on the right front of the driver's driving vision as area 3, area 3 is set as 2 / 3 of the area of the front windshield, the right rearview mirror as area 4, and the in-vehicle central control screen as area 5;

[0077] Sub-step 2) Set the weights of the above 5 fixation areas in advance. The area weight is related to both the area of the fixation area and the fixation time distribution when the driver is driving normally. The larger the area of the fixation area, the higher the fixation frequency when the driver is driving normally, that is, the larger the fixation time distribution, the smaller the influence of this area on the division of the abnormal line-of-sight area when monitoring the driver's dangerous driving, and the smaller the weight of this fixation area. The specific embodiment is: the weight of area 1 is 0.4, the weight of area 2 is 0.3, the weight of area 3 is 0.4, the weight of area 4 is 0.4, and the weight of area 5 is 0.5;

[0078] Sub-step 3) Preset the weighted value of the driver's gaze area within one sampling period as the line-of-sight confusion degree R. In the case of relatively high vehicle speed, the threshold range of the line-of-sight confusion degree R for normal driving is R=(a, b), and the threshold range of the confusion degree R for abnormal driving is R=(b, c). The confusion degree R is calculated by the following formula:

[0079] R = 0.4t1 + 0.3t2 + 0.4t3 + 0.4t4 + 0.5t5

[0080] In the formula, t1 is the time when the driver's gaze area is area 1; t2 is the time when the driver's gaze area is area 2; t3 is the time when the driver's gaze area is area 3; t4 is the time when the driver's gaze area is area 4; t5 is the time when the driver's gaze area is area 5.

[0081] Furthermore, in the embodiment of the present invention, the rule for determining the driving line-of-sight area according to the line-of-sight confusion degree R is specifically as follows:

[0082] (1) Within one sampling period of 2 s, it is set that when the driver is driving normally and the vehicle speed is relatively high, the main gaze areas are area 1, area 2, and area 4, and the driver's gaze time in each area is set to 0.67 s. Then the confusion degree R is:

[0083] R = 0.3×2(t2)=0.6

[0084] R = 0.4×0.67(t1)+0.3×0.67(t2)+0.4×0.67(t4)=0.737

[0085] Then, in the case of relatively high vehicle speed, the threshold range of the line-of-sight confusion degree R for normal driving of the driver is (0.6, 0.737), that is, a is 0.6, and the minimum threshold b of the line-of-sight confusion degree R for abnormal driving is 0.737; within the threshold range R=(a, b) of the line-of-sight confusion degree R for normal driving within one sampling period, it is considered that the driver's behavior of stepping on the accelerator pedal is a normal operation;

[0086] (2) Within one sampling period of 2 s, it is set that when the driver is driving abnormally and the vehicle speed is relatively high, the main gaze areas are area 1, area 2, area 3, and area 4, and the driver's gaze time in each area is set to 0.5 s. Then the confusion degree R is:

[0087] R = 0.4×0.5(t1)+0.3×0.5(t2)+0.4×0.5(t3)+0.4×0.5(t4)=0.75

[0088] When the vehicle speed is relatively fast, the threshold range of the visual confusion degree R of abnormal driving is (0.737, 0.75), that is, c is 0.75; within one sampling period, if the visual confusion degree R is within the threshold range R = (b, c) of the visual confusion degree R of abnormal driving, it is considered that the driver's behavior of stepping on the accelerator pedal is an abnormal operation.

[0089] Method 2: Gini impurity method

[0090] During normal driving, due to the driver's personal habits, there may be a situation where the driver looks back and forth between area 1 and area 2. At this time, the calculated confusion degree R is still very large according to the above method, which is likely to lead to misjudgment. To avoid the above situation, the present invention proposes to define the visual confusion degree by using the Gini impurity method.

[0091] According to the driver's fixation areas within one sampling period and the fixation time of different fixation areas, calculate the percentage value of the fixation time of each area in the sampling time as the fixation probability of this area, and judge the degree of confusion of the obtained data by calculating the Gini impurity. The division of the fixation areas is the same as that in Method 1.

[0092] The calculation formula of the Gini impurity I(f) is:

[0093]

[0094] During normal driving, since this formula includes the fixation frequencies in five areas, and the fixation frequencies in some areas may be 0, which will affect the final judgment, so here it is only calculated for area 1 and area 2. Where m is the number of areas, and f i is the fixation frequency of each area. If the Gini impurity I(f) is smaller and closer to 0, the visual confusion degree is larger; if the Gini impurity I(f) is larger, the visual confusion degree is smaller.

[0095] This method specifically includes the following sub-steps:

[0096] Sub-step 1) Divide the driver's visual area during driving into 5 blocks (see Figure 3 ), specifically including the left rearview mirror as area 1, the area directly in front of the driver's driving vision in the front windshield as area 2, area 2 is set to 1 / 3 of the area of the front windshield, the area in the front windshield on the right front of the driver's driving vision as area 3, area 3 is set to 2 / 3 of the area of the front windshield, the right rearview mirror as area 4, and the in-vehicle central control screen as area 5;

[0097] Sub-step 2) Set the five areas as nominal data. When the driver is driving, only use the percentage of the fixation time of each area in the total time as the probability, and there is no distinction in relative size between the areas;

[0098] Sub-step 3): Preset the Gini impurity of the driver's gaze area within one sampling period as I;

[0099] Ideally, the driver only gazes at area 2, and at this time, the calculated result of the Gini impurity is 1;

[0100] If the driver is used to gazing at area 1 and area 2 simultaneously when driving, the calculated result of the Gini impurity is 0.5;

[0101] Thus, it can be seen that the threshold interval of the Gini impurity I(f) during normal driving is (0.5, 1).

[0102] When the driver is driving abnormally, there are more observation areas within 2 s. For example, when observing five areas simultaneously, theoretically, the minimum value of the Gini impurity can approach 0. However, considering the actual situation, assuming that the gazing time for each area is 0.4 s, the value of the Gini impurity is 0.32. Then, the threshold interval of the Gini impurity I(f) for abnormal driving is (0.32, 0.5).

[0103] Furthermore, in the embodiment of the present invention, the rule for determining the driving line-of-sight area according to the chaos degree R or the Gini impurity I(f) is specifically as follows:

[0104] (1) Within one sampling period of 2 s, it is set that when the driver is driving normally and at a relatively high speed, the main gazing areas are area 1, area 2, and area 4, and the gazing time of the driver in each area is set to 0.67 s. Then, in the case of relatively high speed, the threshold interval of the chaos degree R of the driver's normal driving is (0.6, 0.737), that is, a is 0.6, and the minimum value b of the threshold of the chaos degree R of abnormal driving is 0.737; if the chaos degree R within one sampling period is within the threshold interval R = (a, b) of the chaos degree R of normal driving, it is considered that the driver's behavior of stepping on the accelerator pedal is a normal operation;

[0105] Or:

[0106] Within one sampling period of 2 s, it is set that when the driver is driving normally and at a relatively high speed, the main gazing areas are area 1 and area 2, and the gazing time of the driver in each area is set to 1 s. Then, in the case of relatively high speed, the threshold interval of the Gini impurity I(f) of the driver's normal driving is (0.5, 1); if the Gini impurity I(f) within one sampling period is within the threshold interval I(f) = (0.5, 1) of the Gini impurity I(f) of normal driving, it is considered that the driver's behavior of stepping on the accelerator pedal is a normal operation.

[0107] (2) Within a sampling period of 2 s, when it is set that the driver is driving abnormally and the vehicle speed is relatively fast, the main fixation areas are Area 1, Area 2, Area 3, and Area 4, and the fixation time of the driver in each area is set to 0.5 s. Then, in the case of a relatively fast vehicle speed, the threshold interval of the line-of-sight confusion degree R for abnormal driving is (0.737, 0.75), that is, c is 0.75; within a sampling period, if the line-of-sight confusion degree R is within the threshold interval R = (b, c) of the line-of-sight confusion degree R for abnormal driving, it is considered that the driver's behavior of stepping on the accelerator pedal is an abnormal operation;

[0108] Or:

[0109] Within a sampling period of 2 s, when it is set that the driver is driving abnormally and the vehicle speed is relatively fast, the main fixation areas are Area 1, Area 2, Area 3, Area 4, and Area 5, and the fixation time of the driver in each area is set to 0.4 s. Then, in the case of a relatively fast vehicle speed, the threshold interval of the Gini impurity I(f) for normal driving of the driver is (0.32, 0.5); within a sampling period, if the Gini impurity I(f) is within the threshold interval I(f) = (0.32, 0.5) of the Gini impurity I(f) for normal driving, it is considered that the driver's behavior of stepping on the accelerator pedal is an abnormal operation.

[0110] 2.1c, determine the threshold of the angular velocity of the steering wheel rotation r ;

[0111] The change in the steering wheel angle is the difference in the steering wheel angles within adjacent sampling times, and the angular velocity of the steering wheel rotation refers to the change in the steering wheel angle per unit time. The change situations of these two indicators are essentially the same, and the difference lies in their different times. Assuming that the adjacent sampling time interval is t, there is no distinction in their numerical values and meanings. The present invention mainly uses the angular velocity of the steering wheel rotation as an indicator to determine whether the starting operation is abnormal.

[0112] When the vehicle is driving normally, if the driver's operations are all normal, even if the driver suddenly steps on the accelerator pedal hard, the steering wheel angle and the angular velocity of rotation will only have slight fluctuations and will not show large mutations. On the contrary, during misoperations, due to nervousness, the steering wheel angle and the angular velocity of rotation will show extreme situations, either remaining stationary or having large fluctuations.

[0113] In this embodiment, the angular velocity of the steering wheel rotation is taken as 150° / s, which is the threshold r of the angular velocity of the steering wheel rotation.

[0114] When the accelerator pedal opening exceeds the accelerator pedal opening threshold x, the distance between the vehicle and the obstacle ahead is less than the distance threshold, the driver's line-of-sight confusion degree R is less than or equal to the line-of-sight confusion degree threshold 0.737 and is not 0, or the Gini impurity I(f) is greater than 0.5 and is not 1, and the angular velocity of the steering wheel rotation is less than the angular velocity threshold r of the steering wheel rotation, it is considered that the driver's behavior of stepping on the accelerator pedal is a subjective sudden acceleration driving behavior;

[0115] When the accelerator pedal opening exceeds the accelerator pedal opening threshold x, the distance between the vehicle and the obstacle ahead is less than the distance threshold, the driver's line-of-sight confusion degree R is greater than the line-of-sight confusion degree threshold 0.737 or the Gini impurity I(f) is greater than 0.32 and less than 0.5, and the angular velocity of the steering wheel rotation is greater than or equal to the angular velocity threshold r of the steering wheel rotation or is 0. Since the driver will perform extreme operations in a flurry, resulting in the angular velocity of the steering wheel rotation being greater than or equal to the angular velocity threshold r of the steering wheel rotation or being 0 during the driver's abnormal operation, it is considered that the driver's behavior of stepping on the accelerator pedal is an abnormal operation under flurry.

[0116] Sub-step 2.2, determining the recognition conditions for the driver's misoperation behavior under flurry as:

[0117] 1. The accelerator pedal opening exceeds the accelerator pedal opening threshold x, and the duration is greater than the set accelerator pedal time threshold;

[0118] 2. The distance between the vehicle and the obstacle ahead is less than the distance threshold;

[0119] 3. The driver's line-of-sight confusion degree R is greater than the set line-of-sight confusion degree threshold 0.737 or the Gini impurity I(f) is greater than 0.32 and less than 0.5;

[0120] 4. The angular velocity of the steering wheel rotation is 0, and the duration is greater than the set steering wheel time threshold.

[0121] Or:

[0122] 1. The accelerator pedal opening exceeds the accelerator pedal opening threshold x, and the duration is greater than the set accelerator pedal time threshold;

[0123] 2. The distance between the vehicle and the obstacle ahead is less than the distance threshold;

[0124] 3. The driver's line-of-sight confusion degree R is greater than the set line-of-sight confusion degree threshold or the Gini impurity I(f) is greater than 0.32 and less than 0.5;

[0125] 4. The angular velocity of the steering wheel rotation is greater than the angular velocity threshold r of the steering wheel rotation.

[0126] Obviously, the minimum value of the angular velocity of the steering wheel rotation is 0. However, the state where the angular velocity of the steering wheel rotation is 0 cannot be completely determined as an abnormal starting operation. This is because the driver may perform extreme operations in a panic, and the duration needs to be added as a determination condition for abnormal starting operations together. Therefore, one of the determination conditions for abnormal starting operations in this embodiment is: the opening degree of the accelerator pedal is greater than the accelerator pedal opening degree threshold x, the duration exceeds the accelerator pedal time threshold, the distance between the vehicle and the obstacle in front is less than the distance threshold, the driver's line-of-sight confusion degree R is greater than the set line-of-sight confusion degree threshold 0.737 or the Gini impurity I(f) is greater than 0.32 and less than 0.5, and the angular velocity of the steering wheel is 0, and the duration exceeds the steering wheel time threshold.

[0127] Step 3, the system prompts the driver in a state of misoperation panic.

[0128] Specifically, if it is determined that the driver is in a state of misoperation panic, the ECU issues commands to the vibration motor and the speaker located inside the steering wheel, and prompts the driver of the misoperation behavior through the vibration touch and warning voice of the steering wheel.

[0129] Step 4, the system actively intervenes in the driver's control of the vehicle.

[0130] Specifically, the execution module is linked with the vehicle speed. If the driver still shows no tendency to end the misoperation when the prompt time reaches the warning time threshold, the ECU issues commands to the stepper motor driver and the stepper motor, ordering them to take over the driving right of the vehicle, help the vehicle brake in a timely and safe manner, and avoid collisions.

[0131] According to different vehicle speed ranges within one sampling period, different warning time thresholds are preset. If the prompt time is greater than the warning time threshold preset for this vehicle speed range, it is determined that the driver has no tendency to end the misstep on the pedal, and at this time, the vehicle is controlled to brake. Specifically, the warning time threshold is set as:

[0132] (1) When the vehicle speed is greater than 160 km / h, the warning time threshold is 3 s;

[0133] (2) When the vehicle speed is in the range of 120 km / h to 160 km / h, the warning time threshold is 4 s;

[0134] (3) When the vehicle speed is in the range of 80 km / h to 120 km / h, the warning time threshold is 5 s;

[0135] (4) When the vehicle speed is less than 80 km / h, the warning time threshold is 6 s.

[0136] As described above, it is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims described above.

Claims

1. A method for identifying misoperation behaviors of a vehicle driver's pedal, characterized in that, Including: S1: Collect the opening degree information of the vehicle's accelerator pedal, the driver's gaze area information, the distance information of the vehicle from the obstacle ahead, and the steering wheel angle information or the angular velocity information of the steering wheel rotation; S2: Calculate the visual confusion degree or Gini impurity according to the driver's gaze area information. According to the opening degree information of the accelerator pedal, the visual confusion degree or Gini impurity, the distance information of the vehicle from the obstacle ahead, and the steering wheel angle information or the angular velocity information of the steering wheel rotation, determine whether stepping on the accelerator pedal is an abnormal behavior. If the judgment result is that stepping on the accelerator pedal is an abnormal behavior, then execute S3; wherein, the visual confusion degree is the weighted value of the driver's gaze area within one sampling period; The specific calculation of the Gini impurity according to the driver's gaze area information is as follows: According to the area where the driver gazes and the gazing time in different areas within one sampling period, calculate the percentage of the gazing time in each area accounting for the sampling time as the gazing probability of this area, and calculate the Gini impurity through the following formula :[[]]END]] where m is the number of regions, is the video frequency for each region; The specific method for determining whether stepping on the accelerator pedal is an abnormal behavior is: If the opening degree of the accelerator pedal is greater than the preset accelerator pedal opening degree threshold and the duration is greater than the preset accelerator pedal time threshold; and, the distance of the vehicle from the obstacle ahead is less than the preset distance threshold; and, the visual confusion degree is greater than the preset visual confusion degree threshold or the Gini impurity is less than the preset Gini impurity threshold; and, the angular velocity of the steering wheel rotation is 0 and the duration is greater than the preset steering wheel time threshold; then it is considered that stepping on the accelerator pedal is an abnormal behavior; Or, if the opening degree of the accelerator pedal is greater than the preset accelerator pedal opening degree threshold and the duration is greater than the preset accelerator pedal time threshold; and, the distance of the vehicle from the obstacle ahead is less than the preset distance threshold; and, the visual confusion degree is greater than the preset visual confusion degree threshold or the Gini impurity is less than the preset Gini impurity threshold; and, the angular velocity of the steering wheel rotation is greater than the preset angular velocity threshold of the steering wheel rotation; then it is considered that stepping on the accelerator pedal is an abnormal behavior; S3: Prompt the driver that the pedal is misstepped and / or control the vehicle to brake.

2. The method for identifying the misoperation behavior of an automotive driver's pedal according to claim 1, wherein In S2, the specific calculation of the visual confusion degree according to the driver's gaze area information is: According to the area where the driver gazes within one sampling period and the gaze time of different areas, calculate the weighted value of the area weight and the gaze time to obtain the visual confusion degree; The area is obtained by partitioning the driver's driving visual area, and the area weight is preset.

3. The method for identifying misoperation behavior of an automotive driver pedal according to claim 1, wherein The preset distance threshold is obtained according to the vehicle speed of the vehicle.

4. The method for identifying misoperation behavior of an automotive driver pedal according to claim 1, characterized in that, S3 is specifically: Prompt the driver that the pedal is misstepped. If the driver does not end the operation of misstepping the pedal when the prompt time reaches the preset warning time threshold, then control the vehicle to brake.

5. An automotive driver pedal misoperation behavior recognition system, characterized in that, Including: An information collection module, a central control module, and an execution module; The information collection module is used to collect the opening degree information of the vehicle's accelerator pedal, the driver's gaze area information, the distance information of the vehicle from the obstacle ahead, and the steering wheel angle information or the angular velocity information of the steering wheel rotation, and transmit the collected information to the central control module; The central control module is used to calculate the line-of-sight confusion degree or Gini impurity according to the driver's gaze area information, and judge whether stepping on the accelerator pedal is an abnormal behavior according to the accelerator pedal opening information, the line-of-sight confusion degree or Gini impurity, the distance information of the vehicle from the obstacle ahead, and the steering wheel angle information or steering wheel angular velocity information; the line-of-sight confusion degree is the weighted value of the driver's gaze area within one sampling period; calculating the Gini impurity according to the driver's gaze area information specifically is: According to the area where the driver gazes within one sampling period and the gaze time in different areas, calculate the percentage of the gaze time in each area occupying the sampling time as the gaze probability of this area, and calculate the Gini impurity through the following formula : where m is the number of regions, is the video frequency for each region; Judging whether stepping on the accelerator pedal is an abnormal behavior specifically is: If the accelerator pedal opening is greater than the preset accelerator pedal opening threshold and the duration is greater than the preset accelerator pedal time threshold; and, the distance of the vehicle from the obstacle ahead is less than the preset distance threshold; and, the line-of-sight confusion degree is greater than the preset line-of-sight confusion degree threshold or the Gini impurity is less than the preset Gini impurity threshold; and, the steering wheel angular velocity is 0 and the duration is greater than the preset steering wheel time threshold; then it is considered that stepping on the accelerator pedal is an abnormal behavior; Or, if the accelerator pedal opening is greater than the preset accelerator pedal opening threshold and the duration is greater than the preset accelerator pedal time threshold; and, the distance of the vehicle from the obstacle ahead is less than the preset distance threshold; and, the line-of-sight confusion degree is greater than the preset line-of-sight confusion degree threshold or the Gini impurity is less than the preset Gini impurity threshold; and, the steering wheel angular velocity is greater than the preset steering wheel angular velocity threshold; then it is considered that stepping on the accelerator pedal is an abnormal behavior; The execution module is used to, when the judgment result of the central control module is that stepping on the accelerator pedal is an abnormal behavior, prompt the driver for misstepping on the pedal and / or control the vehicle to brake.

6. The vehicle driver pedal misoperation behavior recognition system according to claim 5, wherein The information acquisition module includes an eye tracker, and the eye tracker is used to acquire the driver's gaze area information; And / or, the information acquisition module includes a millimeter-wave radar, and the millimeter-wave radar is used to acquire the distance information of the vehicle from the obstacle ahead.

7. The automotive driver pedal misoperation behavior recognition system according to claim 5, characterized in that The information acquisition module includes an accelerator pedal position sensor, and the accelerator pedal position sensor is used to acquire the accelerator pedal opening information; And / or, the information acquisition module includes a steering wheel angle sensor, and the steering wheel angle sensor is used to acquire the steering wheel angle information or steering wheel angular velocity information.

8. The vehicle driver pedal misoperation behavior recognition system according to claim 5, characterized in that The execution module includes a stepper motor driver, a stepper motor, a vibration motor and a speaker installed in the steering wheel; the stepper motor driver and the stepper motor are used to control the vehicle to brake; the vibration motor and the speaker are used to prompt the driver for misstepping on the pedal.

Images