CN114492654A - Method and system for detecting mistaken stepping of accelerator pedal - Google Patents

Method and system for detecting mistaken stepping of accelerator pedal Download PDF

Info

Publication number
CN114492654A
CN114492654A CN202210118184.6A CN202210118184A CN114492654A CN 114492654 A CN114492654 A CN 114492654A CN 202210118184 A CN202210118184 A CN 202210118184A CN 114492654 A CN114492654 A CN 114492654A
Authority
CN
China
Prior art keywords
accelerator pedal
data
vehicle
stepping
prediction model
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202210118184.6A
Other languages
Chinese (zh)
Inventor
淦健
刘泽
张洪超
肖柏宏
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Weilai Automobile Technology Anhui Co Ltd
Original Assignee
Weilai Automobile Technology Anhui Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Weilai Automobile Technology Anhui Co Ltd filed Critical Weilai Automobile Technology Anhui Co Ltd
Priority to CN202210118184.6A priority Critical patent/CN114492654A/en
Publication of CN114492654A publication Critical patent/CN114492654A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/10—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration
    • G01C21/12—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration executed aboard the object being navigated; Dead reckoning
    • G01C21/16—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration executed aboard the object being navigated; Dead reckoning by integrating acceleration or speed, i.e. inertial navigation

Abstract

The invention relates to a method, a system, a computer storage medium, a computer device and a vehicle for detecting accelerator pedal missteps. A method for detecting accelerator pedal misstep according to one aspect of the present invention includes: acquiring vehicle driving data; processing the vehicle driving data by using a wrong-stepping prediction model to predict whether the accelerator pedal is stepped by mistake; judging whether the vehicle is in a preset acceleration scene or not in response to the fact that the accelerator pedal is predicted to be mistakenly stepped; and determining that the accelerator pedal is mistakenly stepped in response to the fact that the vehicle is not in the preset acceleration scene.

Description

Method and system for detecting mistaken stepping of accelerator pedal
Technical Field
The present invention relates to the field of vehicle safety, and more particularly to a method, system, computer storage medium, computer device and vehicle for detecting an accelerator pedal misstep.
Background
When a driver needs emergency braking in the face of an emergency in the driving process of a vehicle, a severe traffic accident is sometimes caused by mistakenly touching an accelerator pedal or mistakenly using the accelerator pedal as a brake pedal due to fatigue driving, inattention, over-strain, insufficient experience and the like.
At present, in order to prevent accelerator pedal from being stepped on by mistake, pedal reaction moment can be obviously increased in a short time when the distance of following the car is too close, thereby remind the driver to avoid following the car distance too close through the sense of touch, or through installing multiunit camera discernment vehicle surrounding environment on the automobile body, and have the operation of barrier but the mistake is stepped on accelerator pedal in the place ahead of discernment vehicle and carry out automatic braking intervention in order to avoid or slow down the collision.
However, the method for preventing the accelerator pedal from being stepped on by mistake only utilizes limited environmental information to identify the behavior of stepping on the accelerator pedal by mistake, so that the normal driving behavior may be mistaken as the behavior of stepping on the accelerator pedal by mistake in an actual application scene for intervention, and the driving experience is seriously influenced.
Disclosure of Invention
To solve or at least alleviate one or more of the above problems, the following technical solutions are provided.
According to a first aspect of the present invention, there is provided a method for detecting accelerator pedal misstep, comprising: acquiring vehicle driving data; processing the vehicle driving data by using a wrong-stepping prediction model to predict whether the accelerator pedal is stepped by mistake; judging whether the vehicle is in a preset acceleration scene or not in response to the fact that the accelerator pedal is predicted to be mistakenly stepped; and determining that the accelerator pedal is mistakenly stepped in response to the fact that the vehicle is not in the preset acceleration scene.
According to an embodiment of the invention, the method for detecting the mistaken stepping of the accelerator pedal comprises one or more of the following steps of: vehicle speed, steering wheel angle, longitudinal acceleration, accelerator pedal opening, rate of change of accelerator pedal opening.
The method for detecting a mis-tip-in of an accelerator pedal according to an embodiment of the invention or any one of the above embodiments, wherein processing the vehicle driving data using a mis-tip-in prediction model to predict whether a mis-tip-in of an accelerator pedal occurs comprises: inputting the vehicle driving data into the false tread prediction model to generate a false tread probability value; and predicting that the accelerator pedal is mistakenly stepped when the generated mistaken stepping probability value is larger than a preset probability value.
The method for detecting accelerator pedal mis-stepping according to an embodiment of the invention or any of the above embodiments, wherein the predetermined acceleration scenario includes one or more of: the method comprises the steps of slope ascending acceleration, curve bending exit acceleration, acceleration during forward and backward gear switching and brake release post-addition.
The method for detecting accelerator pedal misstep according to one embodiment of the invention or any of the above embodiments, wherein the method further comprises: determining a road scene in which the vehicle is currently positioned based on vehicle GPS data before processing the vehicle driving data by using a false stepping prediction model to predict whether a false stepping of an accelerator pedal occurs; and selecting a false stepping prediction model associated with the determined road scene where the vehicle is currently located.
The method for detecting a mis-tip-in of an accelerator pedal according to an embodiment of the invention or any of the above embodiments, wherein the mis-tip-in prediction model is trained by: performing data characteristic extraction on data samples of normal running of the vehicle and sample data of mistakenly stepping on an accelerator pedal to obtain a first training data sample set; carrying out negative sample enhancement on the sample data of the mistakenly stepped accelerator pedal to obtain a second training data sample set; and inputting the first training data sample set and the second training data sample set into the false tread prediction model to train the false tread prediction model.
The method for detecting an accelerator pedal mis-pressing according to an embodiment of the invention or any embodiment of the above, wherein the performing data feature extraction on the data sample of normal driving of the vehicle and the sample data of the accelerator pedal mis-pressing to obtain the first training data sample set includes: and selecting data characteristics of moments before and after the change of the opening degree variation of the accelerator pedal in a preset time period is larger than a preset variation as a first training data sample set, wherein the data characteristics are associated with the vehicle running data.
The method for detecting accelerator pedal mis-stepping according to an embodiment of the invention or any of the above embodiments, wherein the mis-stepping prediction model is trained based on an XGBoost algorithm.
The method for detecting accelerator pedal mis-actuation according to one embodiment of the invention or any of the above embodiments, wherein model parameters of the mis-actuation prediction model are dynamically adjusted based on driver behavior data.
According to a second aspect of the present invention, there is provided a system for detecting a false accelerator pedal depression, comprising: a collection unit configured to acquire vehicle travel data; a processing unit configured to process the vehicle travel data using a mis-tip prediction model to predict whether a mis-tip of an accelerator pedal occurs; and a judging unit configured to: judging whether the vehicle is in a preset acceleration scene or not in response to the fact that the accelerator pedal is predicted to be mistakenly stepped; and determining that the accelerator pedal is mistakenly stepped in response to the fact that the vehicle is not in the preset acceleration scene.
According to an embodiment of the invention, the control system for detecting accelerator pedal misstep is provided, wherein the vehicle driving data comprises one or more of the following: vehicle speed, steering wheel angle, longitudinal acceleration, accelerator pedal opening, rate of change of accelerator pedal opening.
The control system for detecting accelerator pedal misstep according to one embodiment of the invention or any one of the above embodiments, wherein the processing unit is further configured to: inputting the vehicle driving data into the false tread prediction model to generate a false tread probability value; and predicting that the accelerator pedal is mistakenly stepped when the generated mistaken stepping probability value is larger than a preset probability value.
The control system for detecting accelerator pedal mis-actuation according to one embodiment of the invention or any of the above embodiments, wherein the predetermined acceleration scenario includes one or more of: the method comprises the steps of ascending acceleration, curve exit acceleration, acceleration during forward and backward gear switching and acceleration after brake release.
The control system for detecting accelerator pedal misstep according to one embodiment of the invention or any one of the above embodiments, wherein the system further includes a selection unit configured to: determining a road scene in which the vehicle is currently positioned based on vehicle GPS data before processing the vehicle driving data by using a false stepping prediction model to predict whether a false stepping of an accelerator pedal occurs; and selecting a false stepping prediction model associated with the determined road scene where the vehicle is currently located.
The control system for detecting a mis-tip-in of an accelerator pedal according to an embodiment of the invention or any one of the above embodiments, wherein the mis-tip-in prediction model is trained by: performing data characteristic extraction on data samples of normal running of the vehicle and sample data of mistakenly stepping on an accelerator pedal to obtain a first training data sample set; carrying out negative sample enhancement on the sample data of the mistakenly stepped accelerator pedal to obtain a second training data sample set; and inputting the first training data sample set and the second training data sample set into the false tread prediction model to train the false tread prediction model.
The control system for detecting an accelerator pedal mis-pressing according to an embodiment of the invention or any one of the above embodiments, wherein the data feature extraction of the data samples of the normal running of the vehicle and the sample data of the accelerator pedal mis-pressing to obtain the first training data sample set includes: and selecting data characteristics of moments before and after the change of the opening degree variation of the accelerator pedal in a preset time period is larger than a preset variation as a first training data sample set, wherein the data characteristics are associated with the vehicle running data.
The control system for detecting accelerator pedal mis-stepping according to an embodiment of the invention or any of the above embodiments, wherein the mis-stepping prediction model is trained based on an XGBoost algorithm.
The control system for detecting accelerator pedal mis-actuation according to one embodiment of the invention or any of the above embodiments, wherein model parameters of the mis-actuation prediction model are dynamically adjusted based on driver behavior data.
According to a third aspect of the present invention, a computer storage medium is provided, comprising instructions which, when executed, perform the steps of the method for detecting a false tip-in of an accelerator pedal according to the first aspect of the present invention.
According to a fourth aspect of the present invention, there is provided a computer device comprising a memory, a processor and a computer program stored on the memory and running on the processor, the processor implementing the steps of the method for detecting a false tip-in of an accelerator pedal according to the first aspect of the present invention when executing the computer program.
According to a fifth aspect of the present invention, there is provided a vehicle comprising a system for detecting accelerator pedal missteps according to the second aspect of the present invention.
According to the scheme for detecting the mistaken stepping of the accelerator pedal, the mistaken stepping prediction model can be established by combining sample data of the actual mistaken stepping of the accelerator pedal, and the accuracy and the reliability of the model prediction result are improved by improving the selection process of the model training sample and the adjustment process of the model parameters. In addition, whether the vehicle is in a special acceleration scene or not is further judged under the condition that the accelerator pedal is predicted to be mistakenly stepped by the mistaken stepping prediction model, so that the detection robustness of the mistaken stepping of the accelerator pedal is further improved, the interference caused by the fact that the normal driving behavior in the special acceleration scene is mistaken for the mistaken stepping of the accelerator pedal is avoided, and the driving experience is improved.
Drawings
The above and/or other aspects and advantages of the present invention will become more apparent and more readily appreciated from the following description of the various aspects taken in conjunction with the accompanying drawings, in which like or similar elements are designated with like reference numerals. In the drawings:
fig. 1 shows a flow chart of a method for detecting a false tip-in of an accelerator pedal according to an embodiment of the invention.
FIG. 2 shows a schematic diagram of a system for detecting accelerator pedal missteps according to one embodiment of the present invention.
FIG. 3 illustrates a process diagram for building a false step prediction model, according to one embodiment of the present invention.
FIG. 4 shows a block diagram of a computer device, in accordance with one embodiment of the present invention.
Detailed Description
The following description of the specific embodiments is merely exemplary in nature and is in no way intended to limit the disclosed technology or the application and uses of the disclosed technology. Furthermore, there is no intention to be bound by any expressed or implied theory presented in the preceding technical field, background, or the following detailed description.
In the following detailed description of embodiments, numerous specific details are set forth in order to provide a more thorough understanding of the disclosed technology. It will be apparent, however, to one of ordinary skill in the art that the disclosed techniques may be practiced without these specific details. In other instances, well-known features have not been described in detail to avoid unnecessarily complicating the description.
Words such as "comprise" and "comprise" mean that in addition to having elements and steps which are directly and explicitly stated in the description, the present solution does not exclude the presence of other elements and steps which are not directly or explicitly stated. Terms such as "first" and "second" do not denote an order of elements in time, space, size, etc., but rather are used to distinguish one element from another.
Hereinafter, exemplary embodiments according to the present invention will be described in detail with reference to the accompanying drawings.
Fig. 1 shows a flow chart of a method for detecting a false tip-in of an accelerator pedal according to an embodiment of the invention.
As shown in fig. 1, in step 110, vehicle travel data is acquired. Alternatively, the vehicle travel data may include, but is not limited to, vehicle speed, steering wheel angle, longitudinal acceleration, accelerator pedal opening, rate of change of accelerator pedal opening, and the like.
For example, the vehicle speed may be acquired by an electronic stability controller system, the steering wheel angle may be acquired by a steering wheel angle sensor, the longitudinal acceleration may be acquired by an inertial measurement unit, and the accelerator opening may be directly acquired by collecting the accelerator pedal sensor voltage.
In step 120, the acquired vehicle travel data is processed using a false tip-in prediction model to predict whether a false tip-in of the accelerator pedal has occurred. The training process of the false stepping prediction model will be described in detail below with reference to fig. 3. In the context of the invention, a wrong-stepping prediction model is established at least partially based on sample data of wrong stepping of an actual accelerator pedal, and the selection process of other model training samples and the adjustment process of model parameters are further improved, so that the accuracy and the reliability of the model prediction result are improved.
Alternatively, in step 120, the vehicle driving data may be used as an input of the mis-pressing prediction model to generate a mis-pressing probability value, and the mis-pressing of the accelerator pedal is predicted when the generated mis-pressing probability value is greater than a preset probability value. As an example, the preset probability value may be given by a mis-stepping prediction model obtained from initial training, and is updated synchronously in an update iteration process of a subsequent mis-stepping prediction model.
If it is predicted that the accelerator pedal is erroneously stepped on, the routine proceeds to step 130. In step 130, it is determined whether the vehicle is in a predetermined acceleration scenario. By way of example, the predetermined acceleration scenario may include, but is not limited to, uphill acceleration, curve out-of-bend acceleration, acceleration during a forward-reverse gear shift, and post-brake release acceleration, among others. It is to be understood that the predetermined acceleration scenario herein refers to a vehicle travel scenario in which the vehicle travel data has a high similarity to the vehicle travel data in the case where the accelerator pedal is mistakenly stepped, and the above-listed predetermined acceleration scenarios are merely exemplary. Therefore, when the accelerator pedal is erroneously depressed as a result of the prediction by the erroneous depression prediction model, it is necessary to further determine whether the vehicle is in a predetermined acceleration scene. And when the vehicle is judged not to be in the preset acceleration scene, determining that the accelerator pedal is mistakenly stepped. Therefore, robustness and effectiveness of detecting mistaken stepping of the accelerator pedal are improved, and driving experience is prevented from being seriously influenced by judging normal driving behaviors in a special acceleration scene as mistaken stepping behaviors. Intervention may be performed using a response strategy specified within the vehicle controller when it is determined that a mis-tip in of the accelerator pedal has occurred. Illustratively, response strategies may include, but are not limited to, audible alert alerts, tactile alerts through pedal reaction forces, disabling torque output, hard braking, and the like.
For example, in step 130, it may be determined whether the vehicle is in an uphill acceleration scene through a longitudinal acceleration obtained by an inertia measurement unit of the vehicle, assisted by a turn signal to determine whether the vehicle is in a curve-out acceleration scene, assisted by a gear signal to determine whether the vehicle is in a forward-backward gear shift-time acceleration scene, and determined whether the vehicle is in a brake-released post-acceleration scene through a brake pedal state signal.
In one embodiment, prior to step 120, a road scenario in which the vehicle is currently located may be determined based on the vehicle GPS data, and a false tread prediction model associated with the determined road scenario in which the vehicle is currently located may be selected. Illustratively, road scenes may include, but are not limited to, urban roads, expressways, national roads, provincial roads, county roads, rural roads, parking lot roads, and the like. In one embodiment, the road scenario in which the vehicle is currently located may be determined based on vehicle GPS data, and a response strategy associated with the determined road scenario in which the vehicle is currently located may be selected for intervening in driver mis-tip-in accelerator pedal behavior if it is determined that a mis-tip-in of the accelerator pedal occurs. For example, when the vehicle is determined to be on a parking lot road, the driver can be reminded of mistakenly stepping on the accelerator pedal through a sound warning signal, and when the vehicle is determined to be on an urban road, the driver can be reminded of mistakenly stepping on the accelerator pedal through the touch of the reaction force of the pedal.
The mistaken-stepping prediction model associated with the road scene where the vehicle is located at present is selected to process the vehicle driving data to predict whether the accelerator pedal is mistakenly stepped, so that the model can be better suitable for different road scenes, and the accuracy and the reliability of the model prediction result are improved. In addition, the response strategy associated with the determined road scene where the vehicle is currently located is selected to intervene the driver's mistaken accelerator pedal stepping behavior under the condition that the driver is determined to mistakenly step on the accelerator pedal, so that the intervention on the driver's driving behavior can be reduced as much as possible, and the driving experience is improved.
According to the method for detecting the accelerator pedal misstep provided by one aspect of the invention, whether the vehicle is in a special acceleration scene can be further judged under the condition that the misstep prediction model predicts the misstep of the accelerator pedal, so that the detection robustness of the misstep of the accelerator pedal is improved, and the interference caused by misjudging the normal driving behavior in the special acceleration scene as the misstep behavior of the accelerator pedal is avoided, thereby improving the driving experience.
FIG. 2 shows a schematic diagram of a system for detecting accelerator pedal missteps according to one embodiment of the present invention.
As shown in fig. 2, the system 200 for detecting an accelerator pedal misstep includes an acquisition unit 210, a processing unit 220, and a determination unit 230.
The acquisition unit 210 may be configured to acquire vehicle travel data. Alternatively, the vehicle travel data may include, but is not limited to, vehicle speed, steering wheel angle, longitudinal acceleration, accelerator pedal opening, rate of change of accelerator pedal opening, and the like.
For example, the acquisition unit 210 may acquire a vehicle speed through an electronic stability controller system, a steering wheel angle through a steering wheel angle sensor, a longitudinal acceleration through an inertia measurement unit, and an accelerator pedal opening directly through acquisition of an accelerator pedal sensor voltage.
The processing unit 220 may be configured to process the acquired vehicle travel data using a false tip-in prediction model to predict whether a false tip-in of the accelerator pedal has occurred. The training process of the false stepping prediction model will be described in detail below with reference to fig. 3.
Alternatively, the processing unit 220 may be configured to use the vehicle driving data as an input of the mis-pressing prediction model to generate a mis-pressing probability value, and predict that the accelerator pedal is mis-pressed when the generated mis-pressing probability value is greater than a preset probability value. As an example, the preset probability value may be given by a mis-stepping prediction model obtained from initial training, and is updated synchronously in an update iteration process of a subsequent mis-stepping prediction model.
The determination unit 230 may be configured to determine whether the vehicle is in a predetermined acceleration scenario in response to the processing unit 220 predicting that the accelerator pedal is misstepped, and determine that the accelerator pedal is misstepped in response to determining that the vehicle is not in the predetermined acceleration scenario.
By way of example, the predetermined acceleration scenario may include, but is not limited to, an uphill acceleration, a curve out-of-curve acceleration, a forward-reverse gear shift acceleration, and a brake release post-acceleration, among others. It is to be understood that the predetermined acceleration scenario herein refers to a vehicle travel scenario in which the vehicle travel data has a high similarity to the vehicle travel data in the case where the accelerator pedal is mistakenly stepped, and the above-listed predetermined acceleration scenarios are merely exemplary. Therefore, when the accelerator pedal is erroneously depressed as a result of the prediction by the erroneous depression prediction model, it is necessary to further determine whether the vehicle is in a predetermined acceleration scene. And when the vehicle is judged not to be in the preset acceleration scene, determining that the accelerator pedal is mistakenly stepped. Therefore, robustness and effectiveness of detecting mistaken stepping of the accelerator pedal are improved, and driving experience is prevented from being seriously influenced by judging normal driving behaviors in a special acceleration scene as mistaken stepping behaviors.
For example, the determining unit 230 may be configured to determine whether the vehicle is in a scene of uphill acceleration through the longitudinal acceleration obtained by the inertia measuring unit of the vehicle, assist to determine whether the vehicle is in a scene of curve-out acceleration in combination with the turn signal, determine whether the vehicle is in a scene of acceleration at the time of forward-backward gear shift in combination with the gear signal, and determine whether the vehicle is in a scene of acceleration after brake release through the brake pedal state signal.
In one embodiment, the system 200 for detecting accelerator pedal mis-actuation may further include a selection unit (not shown in fig. 2), which may be configured to determine a road scene in which the vehicle is currently located based on the vehicle GPS data and select a mis-actuation prediction model associated with the determined road scene in which the vehicle is currently located, before processing the vehicle travel data using the mis-actuation prediction model to predict whether the accelerator pedal is mis-actuated. Illustratively, road scenes may include, but are not limited to, urban roads, expressways, national roads, provincial roads, county roads, rural roads, parking lot roads, and the like. In one embodiment, the selection unit may be further configured to determine a road scenario in which the vehicle is currently located based on the vehicle GPS data, and to select a response strategy associated with the determined road scenario in which the vehicle is currently located for intervening in a driver's accelerator pedal mis-actuation behavior in case it is determined that the accelerator pedal is mis-actuated. For example, when the vehicle is determined to be on a parking lot road, the driver can be reminded of mistakenly stepping on the accelerator pedal through a sound warning signal, and when the vehicle is determined to be on an urban road, the driver can be reminded of mistakenly stepping on the accelerator pedal through the touch of the reaction force of the pedal.
The mistaken-stepping prediction model associated with the road scene where the vehicle is located at present is selected to process the vehicle driving data to predict whether the accelerator pedal is mistakenly stepped, so that the model can be better suitable for different road scenes, and the accuracy and the reliability of the model prediction result are improved. In addition, the response strategy associated with the determined road scene where the vehicle is currently located is selected to intervene the driver's mistaken accelerator pedal stepping behavior under the condition that the driver is determined to mistakenly step on the accelerator pedal, so that the intervention on the driver's driving behavior can be reduced as much as possible, and the driving experience is improved.
The system for detecting the mistaken stepping of the accelerator pedal provided by one aspect of the invention can further judge whether the vehicle is in a special acceleration scene or not under the condition that the mistaken stepping of the accelerator pedal is predicted by the mistaken stepping prediction model, so that the detection robustness of the mistaken stepping of the accelerator pedal is improved, the interference caused by the mistaken judgment of the normal driving behavior in the special acceleration scene as the mistaken stepping of the accelerator pedal is avoided, and the driving experience is improved.
FIG. 3 is a schematic diagram of a process for building a false step prediction model, according to one embodiment of the present invention. As shown in fig. 3, the process for building the mis-stepping prediction model may include three stages, namely a data sample acquisition and feature extraction stage, a model training stage, and a model output stage.
Block 310 illustrates the data sample collection and feature extraction phase for building a false tread prediction model. Optionally, in this stage, data feature extraction may be performed on data samples of normal running of the vehicle and sample data of mistakenly stepping on the accelerator pedal to obtain a first training data sample set, negative sample enhancement may be performed on the sample data of mistakenly stepping on the accelerator pedal to obtain a second training data sample set, and the first training data sample set and the second training data sample set may be used as training data of the mistakenly stepping prediction model to train the mistakenly stepping prediction model. By taking sample data of mistakenly stepping on the accelerator pedal as a part of the training data sample, the training model can be better adapted to the data characteristics of the condition that the user actually mistakenly steps on the accelerator pedal, and the effectiveness of the model prediction result is ensured.
Optionally, the data feature extraction of the data samples of normal driving of the vehicle and the sample data of mistaken stepping on the accelerator pedal to obtain the first training data sample set may include: and selecting data characteristics of the moment before and after the change that the opening variation of the accelerator pedal is larger than the preset variation in a preset time period as a first training data sample set. For example, the predetermined period of time may be set to 1 second, the preset variation may be set to 70%, and the data feature at the time before and after the change in which the variation in the opening degree of the accelerator pedal is greater than 70% in 1 second may be selected as the first training data sample set. For example, the predetermined period of time may be adjusted in a calibrated manner. For example, the extracted data features may include, but are not limited to, vehicle speed, steering wheel angle, longitudinal acceleration, accelerator pedal opening, rate of change of accelerator pedal opening, and the like.
In one embodiment, the data sample collection and feature extraction can comprise collecting multi-vehicle driving sample data, collecting data of time sequences before and after the single-vehicle accelerator pedal is stepped on by mistake, extracting data features and enhancing negative sample data.
For example, vehicle operation data meeting a screening condition may be collected from an operation vehicle database as multi-vehicle driving sample data, wherein the screening condition may be determined based on big data statistics. For example, in most cases before and after the accelerator pedal is erroneously stepped on, the accelerator pedal opening is larger than 90% and the rate of change of the accelerator pedal opening in 1 second is larger than 20%. Therefore, the data of the normal running of the vehicle satisfying the condition in the running vehicle database can be screened as the multi-vehicle driving sample data by taking the condition as the screening condition.
For example, time sequence data of corresponding time slices can be collected from the running vehicle database according to the occurrence time of the feedback accelerator pedal event and used as the time sequence data before and after the single vehicle accelerator pedal is mistakenly stepped.
Illustratively, the data feature extraction comprises data segmentation and feature extraction operations on multi-vehicle driving sample data and accelerator pedal mistaken-stepping sample data. For example, a data feature may be extracted at a time when the accelerator pedal opening changes from less than 20% of the initial value to more than 90% of the opening in 1 second.
Illustratively, the negative sample data enhancement comprises negative sample enhancement on the data of time series before and after the accelerator pedal mistaken stepping occurs so as to increase the number of mistakenly stepping data samples. For example, negative sample enhancement may be performed by changing the steering wheel angle to the right for which false stepping data occurs to the left and by changing the steering wheel angle to the left for which false stepping data occurs to the right. Similarly, other data features may be processed similarly.
By carrying out negative sample enhancement on the accelerator pedal mistakenly-stepping data sample, the problem that the model training result is influenced because the mistakenly-stepping data training sample is limited due to the fact that the accelerator pedal is mistakenly stepped on the extremely-small-probability event can be solved.
Block 320 illustrates a model training phase. Optionally, in this stage, a model suitable for the data features may be selected for training, while the prediction accuracy is taken as the model optimization direction. Illustratively, the XGboost algorithm may be selected for model training based on factors such as data size and data distribution characteristics. By utilizing the XGboost algorithm to train the model, the precision, the efficiency and the flexibility of the model training can be improved.
It is noted that the training of the model using the XGBoost algorithm is merely exemplary, and other model training algorithms may be selected to train the model without departing from the spirit and scope of the present invention.
Optionally, the model parameters of the misstep prediction model may be given by the misstep prediction model obtained by the initial training, and are updated synchronously in the subsequent update iteration process of the misstep prediction model. For example, model parameters of the false tread prediction model may be dynamically adjusted based on driver behavior data. For example, the driving behavior of the driver can be monitored at a buried point in the vehicle, and the driving behavior model of the driver can be established based on the driving behavior, so that different model parameters can be set for different drivers based on the driving behavior model result to calibrate the model, and the calibrated model can accurately judge whether the operation of the driver is the operation of mistakenly stepping on the accelerator pedal according to the driving habits of the driver. In one embodiment, a model parameter interface for dynamically adjusting model parameters is reserved for the false tread prediction model, so that the model parameters can be dynamically adjusted according to driving behavior habits of different drivers.
Block 330 shows the model output results. Alternatively, the model output may indicate a probability value of the occurrence of an accelerator pedal misstep.
According to the scheme for training the mistaken stepping prediction model provided by one or more embodiments of the invention, the sample data of the actual mistaken stepping of the accelerator pedal can be combined, and the accuracy and the reliability of the model prediction result are improved by improving the selection process of the model training sample and the adjustment process of the model parameters.
FIG. 4 shows a block diagram of a computer device, in accordance with one embodiment of the present invention. As shown in fig. 4, the computer device 400 includes a memory 410, a processor 420, and a computer program 430 stored on the memory 410 and executable on the processor 420. The processor 420, when executing the computer program 430, performs the steps of a method for detecting an accelerator pedal misstep according to an aspect of the present invention, such as that shown in fig. 1.
In addition, as described above, the present invention may also be embodied as a computer storage medium in which a program for causing a computer to execute the method for detecting an accelerator pedal misstep according to an aspect of the present invention is stored.
Here, as the computer storage medium, various types of computer storage media such as a disk (e.g., a magnetic disk, an optical disk, etc.), a card (e.g., a memory card, an optical card, etc.), a semiconductor memory (e.g., a ROM, a nonvolatile memory, etc.), a tape (e.g., a magnetic tape, a cassette tape, etc.), and the like can be used.
Where applicable, the various embodiments provided by the present disclosure may be implemented using hardware, software, or a combination of hardware and software. Also, where applicable, the various hardware components and/or software components set forth herein may be combined into composite components comprising software, hardware, and/or both without departing from the scope of the present disclosure. Where applicable, the various hardware components and/or software components set forth herein may be separated into sub-components comprising software, hardware, or both without departing from the scope of the present disclosure. Further, where applicable, it is contemplated that software components may be implemented as hardware components, and vice versa.
Software in accordance with the present disclosure (such as program code and/or data) can be stored on one or more computer storage media. It is also contemplated that the software identified herein may be implemented using one or more general purpose or special purpose computers and/or computer systems that are networked and/or otherwise. Where applicable, the order of various steps described herein may be changed, combined into composite steps, and/or separated into sub-steps to provide features described herein.
The embodiments and examples set forth herein are presented to best explain embodiments in accordance with the invention and its particular application and to thereby enable those skilled in the art to make and utilize the invention. However, those skilled in the art will recognize that the foregoing description and examples have been presented for the purpose of illustration and example only. The description as set forth is not intended to cover all aspects of the invention or to limit the invention to the precise form disclosed.

Claims (9)

1. A method for detecting a mis-depression of an accelerator pedal, the method comprising:
obtaining vehicle travel data, wherein the vehicle travel data includes one or more of: vehicle speed, steering wheel angle, longitudinal acceleration, accelerator pedal opening, rate of change of accelerator pedal opening;
processing the vehicle driving data by using a wrong-stepping prediction model to predict whether the accelerator pedal is mistakenly stepped, wherein the step of processing the vehicle driving data by using the wrong-stepping prediction model to predict whether the accelerator pedal is mistakenly stepped comprises the steps of inputting the vehicle driving data into the wrong-stepping prediction model to generate a wrong-stepping probability value, and predicting that the accelerator pedal is mistakenly stepped when the generated wrong-stepping probability value is larger than a preset probability value;
determining whether a vehicle is in a predetermined acceleration scenario in response to predicting a mis-tip of the accelerator pedal, wherein the predetermined acceleration scenario includes one or more of: ascending acceleration, curve exit acceleration, acceleration during forward and backward gear switching and acceleration after brake release; and
determining that the accelerator pedal is mistakenly stepped in response to judging that the vehicle is not in the preset acceleration scene;
wherein the mis-stepping prediction model is trained by the following steps:
performing data characteristic extraction on data samples of normal running of the vehicle and sample data of mistakenly stepping on an accelerator pedal to obtain a first training data sample set;
carrying out negative sample enhancement on the sample data of the mistakenly stepped accelerator pedal to obtain a second training data sample set; and
inputting the first training data sample set and the second training data sample set into the false stepping prediction model to train the false stepping prediction model, wherein the false stepping prediction model is trained based on an XGboost algorithm, and model parameters of the false stepping prediction model are dynamically adjusted based on driver behavior data.
2. The method of claim 1, wherein the method further comprises:
determining a road scene in which the vehicle is currently positioned based on vehicle GPS data before processing the vehicle driving data by using a false stepping prediction model to predict whether a false stepping of an accelerator pedal occurs; and
selecting a false step prediction model associated with the determined road scenario in which the vehicle is currently located.
3. The method of claim 1, wherein performing data feature extraction on data samples of normal vehicle driving and sample data of accelerator pedal misstep to obtain a first training data sample set comprises:
and selecting data characteristics of moments before and after the change of the opening degree variation of the accelerator pedal in a preset time period is larger than a preset variation as a first training data sample set, wherein the data characteristics are associated with the vehicle running data.
4. A system for detecting accelerator pedal missteps, the system comprising:
a collection unit configured to acquire vehicle travel data, wherein the vehicle travel data comprises one or more of: vehicle speed, steering wheel angle, longitudinal acceleration, accelerator pedal opening, rate of change of accelerator pedal opening;
a processing unit configured to process the vehicle driving data using a mis-pressing prediction model to predict whether a mis-pressing of an accelerator pedal occurs, wherein the processing unit is further configured to input the vehicle driving data to the mis-pressing prediction model to generate a mis-pressing probability value, and predict that the accelerator pedal is mis-pressed when the generated mis-pressing probability value is greater than a preset probability value; and
a determination unit configured to:
determining whether a vehicle is in a predetermined acceleration scenario in response to predicting a mis-tip of the accelerator pedal, wherein the predetermined acceleration scenario includes one or more of: ascending acceleration, curve exit acceleration, acceleration during forward and backward gear switching and acceleration after brake release; and
determining that the accelerator pedal is mistakenly stepped in response to the fact that the vehicle is not in the preset acceleration scene;
wherein the mis-stepping prediction model is trained by the following steps:
performing data characteristic extraction on data samples of normal running of the vehicle and sample data of mistakenly stepping on an accelerator pedal to obtain a first training data sample set;
carrying out negative sample enhancement on the sample data of the mistakenly stepped accelerator pedal to obtain a second training data sample set; and
inputting the first training data sample set and the second training data sample set into the false stepping prediction model to train the false stepping prediction model, wherein the false stepping prediction model is trained based on an XGboost algorithm, and model parameters of the false stepping prediction model are dynamically adjusted based on driver behavior data.
5. The system of claim 4, wherein the system further comprises a selection unit configured to:
determining a road scene in which the vehicle is currently positioned based on vehicle GPS data before processing the vehicle driving data by using a false stepping prediction model to predict whether a false stepping of an accelerator pedal occurs; and
selecting a false step prediction model associated with the determined road scenario in which the vehicle is currently located.
6. The system of claim 4, wherein the data feature extraction of data samples of normal vehicle driving and sample data of accelerator pedal misstep to obtain the first training data sample set comprises:
and selecting data characteristics of moments before and after the change of the opening degree variation of the accelerator pedal in a preset time period is larger than a preset variation as a first training data sample set, wherein the data characteristics are associated with the vehicle running data.
7. A computer storage medium comprising instructions that when executed perform the method of any one of claims 1 to 3.
8. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor when executing the program implements the method of any one of claims 1 to 3.
9. A vehicle, characterized by comprising a system for detecting accelerator pedal missteps according to any one of claims 4 to 6.
CN202210118184.6A 2022-02-08 2022-02-08 Method and system for detecting mistaken stepping of accelerator pedal Pending CN114492654A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210118184.6A CN114492654A (en) 2022-02-08 2022-02-08 Method and system for detecting mistaken stepping of accelerator pedal

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210118184.6A CN114492654A (en) 2022-02-08 2022-02-08 Method and system for detecting mistaken stepping of accelerator pedal

Publications (1)

Publication Number Publication Date
CN114492654A true CN114492654A (en) 2022-05-13

Family

ID=81478320

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210118184.6A Pending CN114492654A (en) 2022-02-08 2022-02-08 Method and system for detecting mistaken stepping of accelerator pedal

Country Status (1)

Country Link
CN (1) CN114492654A (en)

Similar Documents

Publication Publication Date Title
CN108248609B (en) Hybrid vehicle and method of predicting driving pattern in hybrid vehicle
JP6003349B2 (en) Vehicle behavior prediction device
EP2899087B1 (en) Vehicle driving behavior predicting device
CN110466351B (en) Detection control method for preventing mistaken stepping on accelerator of passenger car
JP2009001066A (en) Accelerator pedal erroneous operation detector for vehicle
CN112937520B (en) Emergency braking method and device for vehicle, commercial vehicle and storage medium
CN110871802A (en) Vehicle control method and device and vehicle
US8666563B2 (en) Determination apparatus for determining erroneous application of accelerator rather than brake
CN113327458A (en) Vehicle collision prediction method, vehicle collision prediction system, and electronic device
US11618447B2 (en) Sudden acceleration prevention method and vehicle using the same
CN112542041A (en) Driving behavior evaluation device, method, and computer-readable storage medium
CN111881808A (en) Method for identifying abnormal operation behaviors of driver in starting stage of electric bus based on continuous tracking data
JP2013002619A (en) Shift control device, shift control method, and shift control program for automobile
CN114103917A (en) Vehicle driving support device, vehicle driving support method, and program
CN115320407B (en) Vehicle control method and vehicle control device
CN111483458B (en) Power system control method and device
CN114492654A (en) Method and system for detecting mistaken stepping of accelerator pedal
CN114228719B (en) Auxiliary braking method for vehicle, electronic device and storage medium
US20220172089A1 (en) Behavior prediction device
CN113442723B (en) Control method and device for preventing mistaken stepping on accelerator of vehicle and vehicle
CN112078573B (en) Vehicle and anti-collision method and device of vehicle
CN114834452A (en) Driving mode switching method, relation determining method, device and system and vehicle
JP4083552B2 (en) Vehicle alarm device
US20230249679A1 (en) Driver assistance device, method, non-transitory storage medium, and vehicle
JP2006178674A (en) Traveling supporting device

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination