CN111731312A - Experimental system for extracting driving style characteristic parameters and driving style identification method - Google Patents
Experimental system for extracting driving style characteristic parameters and driving style identification method Download PDFInfo
- Publication number
- CN111731312A CN111731312A CN202010540775.3A CN202010540775A CN111731312A CN 111731312 A CN111731312 A CN 111731312A CN 202010540775 A CN202010540775 A CN 202010540775A CN 111731312 A CN111731312 A CN 111731312A
- Authority
- CN
- China
- Prior art keywords
- driving
- driving style
- characteristic parameters
- experimental system
- average value
- 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
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W40/00—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models
- B60W40/08—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models related to drivers or passengers
- B60W40/09—Driving style or behaviour
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
- G01D21/00—Measuring or testing not otherwise provided for
- G01D21/02—Measuring two or more variables by means not covered by a single other subclass
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
- G05B13/042—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators in which a parameter or coefficient is automatically adjusted to optimise the performance
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Automation & Control Theory (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Transportation (AREA)
- General Engineering & Computer Science (AREA)
- Probability & Statistics with Applications (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Mathematical Physics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mechanical Engineering (AREA)
- Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Traffic Control Systems (AREA)
- Control Of Driving Devices And Active Controlling Of Vehicle (AREA)
Abstract
The invention discloses an experimental system for extracting driving style characteristic parameters and a driving style identification method, wherein the experimental system for extracting the driving style characteristic parameters comprises a driver operation platform, a muscle electric signal sensor, a lower computer and an upper computer, wherein the driver operation platform is used for simulating driving scenes with different traffic density and can obtain opening information of a brake pedal, opening information of an accelerator pedal and corner information of a steering wheel in real time; identifying and storing a speed value of the vehicle in the driving process, and identifying and storing coordinates of the vehicle in a map; the muscle electric signal sensor is used for acquiring gastrocnemius signals and tibialis anterior muscle signals of a driver, driving style characteristic parameters can be comprehensively and accurately acquired through an experimental system, the driving style classification and identification considering the influence of traffic density are further realized, and the device has breakthrough significance in characteristic parameter acquisition and application.
Description
Technical Field
The invention relates to the technical field of automobile energy management strategies, in particular to an experimental system for extracting driving style characteristic parameters and a driving style identification method.
Background
The driving style is the behavior characteristic of a driver in the process of driving a vehicle, and is reflected in human-to-vehicle input and whole-vehicle response in the driving process. In the development process of the energy management strategy of the whole automobile, the effective identification of the driving style can enhance the self-adaptive capacity of the whole automobile to different driving styles, which has important significance for improving the fuel economy and emission performance of the hybrid electric vehicle.
The classification and identification of the driving style can be greatly influenced by the difference of traffic flow densities on the road, so that the extracted characteristic parameters of the same group of driving style may show different driving style types under different traffic flow densities, such as: the same set of driving style characteristic parameters show an aggressive style under the condition of large traffic density, but show a common style under the condition of small traffic density. For this reason, the driving style under different traffic densities needs to be redefined and corrected. Therefore, it is necessary to design an experimental system and a driving style recognition method for extracting driving style characteristic parameters, and a driving style classification and recognition method considering the influence of traffic density based on the experimental system.
Disclosure of Invention
The invention aims to design an experimental system for extracting driving style characteristic parameters and a driving style classification and identification method based on the experimental system and considering the influence of traffic flow density. The driving style characteristic parameter table under different traffic flow densities is established by adopting a multilevel hybrid algorithm based on the driving style characteristic parameters obtained by extraction, and the driving styles of drivers under different traffic flow densities are accurately identified through an algorithm for identifying the driving styles, namely a random forest algorithm, so that the fuel economy of the automobile is improved through the formulation and optimization of a driving style self-adaptive control strategy.
In order to achieve the purpose, the invention provides the following technical scheme: the utility model provides an extract experimental system of driving style characteristic parameter, includes driver's operation panel, muscle electric signal sensor, next machine and host computer, wherein:
the driver operation console is used for simulating driving scenes with different traffic flow densities and can obtain the opening information of a brake pedal, the opening information of an accelerator pedal and the corner information of a steering wheel in real time; identifying and storing a speed value of the vehicle in the driving process, and identifying and storing coordinates of the vehicle in a map;
the muscle electric signal sensor is used for acquiring gastrocnemius signals and tibialis anterior signals of a driver;
the lower computer can transmit the information obtained by the driver console and the information detected by the muscle electric signal sensor to the upper computer;
the upper computer extracts driving characteristic parameters under different traffic densities through information transmitted by the lower computer, and performs multi-level hybrid algorithm processing on the driving characteristic parameters, so as to establish a driving style recognition model and train the driving style recognition model.
Furthermore, a brake pedal, an accelerator pedal, a steering wheel, and driving simulation software and visual recognition software which are arranged on the driver operating platform, the brake pedal, the accelerator pedal and the steering wheel detect the opening information of the brake pedal, the opening information of the accelerator pedal and the turning angle information of the steering wheel through angle sensors, the driving simulation software can simulate four driving scenes with traffic density, and the visual recognition software comprises instrument panel recognition software and GPS coordinate recognition software.
Furthermore, the driving scenes of the four traffic densities are 10%, 40%, 70% and 100% of four urban work conditions respectively.
Furthermore, the lower computer is an STM32 singlechip.
Further, the multilevel hybrid algorithm comprises a principal component analysis method, a subtraction clustering algorithm and a K-means clustering algorithm, and a driving style characteristic parameter table under different traffic flow densities is established, so that a driving style identification model is formed.
Further, the angle sensor is an incremental photoelectric rotary encoder.
Further, the driving characteristic parameters include a vehicle speed average value, a vehicle speed standard deviation, an acceleration standard deviation, a positive acceleration average value, a positive acceleration standard deviation, a negative acceleration average value, a negative acceleration standard deviation, a required power coefficient standard deviation, an acceleration required power coefficient average value, a deceleration required power coefficient average value, a gastrocnemius electromyogram time domain characteristic root mean square, an electromyogram frequency domain characteristic average power frequency of the gastrocnemius, an electromyogram time domain characteristic root mean of the tibialis anterior muscle, an electromyogram frequency domain characteristic average power frequency of the tibialis anterior muscle, a sharp turning frequency, a lane changing frequency, a steering wheel stroke standard deviation, a steering wheel angular velocity average value, a steering wheel angular velocity standard deviation, an accelerator pedal stroke average value, an accelerator pedal stroke standard deviation, an accelerator pedal change rate average value, a vehicle speed average value, the average value of the change rate of the released accelerator pedal, the average value of the travel of the brake pedal, the standard deviation of the change rate of the brake pedal, the average value of the change rate of the treaded brake pedal and the average value of the change rate of the released brake pedal.
A driving style recognition method is realized by using the experimental system for extracting the characteristic parameters of the driving style, and comprises the following steps:
s01: acquiring driving characteristic parameters of drivers of different styles and types in a set period under simulated road working conditions of different traffic densities by using an experimental system;
s02: performing comprehensive and dimension reduction processing on all driving characteristic parameters in S01 by using a principal component analysis method to obtain comprehensive characteristic parameters;
s03: performing cluster analysis on the comprehensive characteristic parameters in the step S02 by using a subtractive clustering algorithm method, so as to obtain the clustering centers and the clustering numbers of the comprehensive characteristic parameters of different driving styles under different traffic flow densities;
s04: correcting the comprehensive characteristic parameter clustering center in the K-means clustering algorithm S03 to obtain corrected comprehensive characteristic parameter clustering center results of various driving styles under different traffic densities;
s05: calculating various corrected driving style characteristic parameters, and establishing characteristic parameter tables of different driving styles under different traffic densities;
s06: and establishing a driving style recognition model based on a random forest algorithm, and training the driving style recognition model under each traffic density by using driving style data under different traffic densities.
Further, in the step S03, the same parameters are used for the cluster analysis process of the comprehensive characteristic parameters by using a subtractive clustering algorithm under different traffic densities.
Further, the training method of the driving style recognition model comprises the following steps:
and 3, repeating the step 1 and the step 2 until the number of the decision trees reaches a preset number, and forming a random forest model for driving style recognition.
Compared with the prior art, the invention has the following beneficial effects:
the characteristic parameters for classifying and identifying the driving style can be comprehensively and accurately acquired through the experiment system, and the characteristic parameter acquisition has breakthrough significance.
The collection of the simulated driving data is independent of the simulated driving software by using the visual recognition software and the angle sensor, so that third-party simulated driving software which is not open source can be selected for collecting the driving style characteristic parameters.
By using a clustering method based on a multi-level hybrid algorithm and a driving style identification method based on a random forest algorithm, the proper driving style clustering number and clustering center can be calculated according to the traffic flow density value of the road where the vehicle is located, and the driving style identification model under the corresponding traffic flow density is adopted for identifying the driving style, so that different driving style identification models are provided for vehicles under the road working conditions with different traffic flow densities, the driving style identification precision is improved, a more effective whole-vehicle energy management strategy can be adopted, a foundation is laid for the development of driving style self-adaptive control strategies under different traffic flow densities, and the method has important significance for improving the fuel economy and emission performance of a hybrid electric vehicle.
Drawings
FIG. 1 is a schematic diagram of a driver pedal signal acquisition device in an experimental system for extracting driving style characteristic parameters;
FIG. 2 is a schematic diagram of a steering wheel signal acquisition device in an experimental system for extracting driving style characteristic parameters;
FIG. 3 is a flowchart of the operation of extracting driving style characteristic parameters by the experimental system;
FIG. 4 is a schematic diagram of the operation of the dashboard recognition software;
FIG. 5 is a schematic diagram of GPS coordinate recognition software;
FIG. 6 is a driving style classification and identification method based on traffic density;
FIG. 7 shows the result of clustering the driving styles when the traffic density is 10%;
FIG. 8 shows the result of clustering the driving styles when the traffic density is 40%;
FIG. 9 shows a result of clustering the driving style with a traffic density of 70%;
fig. 10 shows the driving style clustering result when the traffic density is 100%.
In the figure, 1, a driver operating platform, 2, driving simulation software, 3, visual recognition software, 4, an angle sensor, 5, a muscle electric signal sensor, 6, a lower computer, 7, an upper computer, 8, a brake pedal, 9, an accelerator pedal, 10, a steering wheel, 11, instrument panel recognition software, 12, GPS coordinate recognition software, 13, a driving gear, 14, a driven gear and 15 output shafts.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In the description of the present invention, it should be noted that the terms "upper", "lower", "inner", "outer", "front", "rear", "both ends", "one end", "the other end", and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the drawings, and are only for convenience of description of the present invention, and do not indicate or imply that the referred devices or elements must have a specific orientation, be configured in a specific orientation, and operate, and thus, should not be construed as limiting the present invention. Furthermore, the terms "first" and "second" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
In the description of the present invention, it is to be noted that, unless otherwise explicitly specified or limited, the terms "mounted," "disposed," "connected," and the like are to be construed broadly, such as "connected," which may be fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood in specific cases to those skilled in the art.
Referring to the attached drawings, an experimental system for extracting driving style characteristic parameters comprises a driver operating platform 1, driving simulation software 2, visual recognition software 3, an angle sensor 4, a muscle electric signal sensor 5, a lower computer 6 and an upper computer 7; as shown in fig. 1 and 2, the driver operating console is provided with a brake pedal 8, an accelerator pedal 9 and a steering wheel 10, and transmits a brake pedal opening signal, an accelerator pedal opening signal and a steering wheel angle signal to the simulated driving software 2 in real time; the rotor of the angle sensor 4 is respectively connected to a brake pedal 8, an accelerator pedal 9 and a steering wheel 10 and is respectively used for acquiring a brake pedal opening signal, an accelerator pedal opening signal and a steering wheel angle signal; as shown in fig. 4 and 5, the visual recognition software includes instrument panel recognition software 11, GPS coordinate recognition software 12; the muscle electric signal sensor is respectively used for acquiring gastrocnemius and tibialis anterior muscle signals of a driver;
the data acquisition workflow of the whole experimental system is shown in fig. 3, which is specifically explained as follows:
a: the driver operates a brake pedal 8, an accelerator pedal 9, and a steering wheel 10 on the driver's console 1.
b: the driver operating platform 1 transmits the collected brake pedal opening degree signal, the collected accelerator pedal opening degree signal and the collected steering wheel angle signal to the simulation driving software 2 in real time.
c: the information displayed on the screen by the simulated driving software 2 is identified and stored by the visual identification software 3, wherein the instrument panel identification software identifies and stores the speed value of the vehicle in the driving process, and the GPS coordinate identification software identifies and stores the coordinate of the vehicle in the map.
d: the muscle electric signal sensor 5 is used for acquiring gastrocnemius and tibialis anterior muscle signals of the driver respectively;
and e, the angle sensor 4 is respectively used for acquiring a brake pedal opening signal, an accelerator pedal opening signal and a steering wheel rotation angle signal on the driver operating platform 1 in real time.
And f, the muscle electric signal sensor 5 transmits the acquired gastrocnemius and tibialis anterior muscle signals to the lower computer 6 through an ADC (analog-to-digital converter) port of the lower computer 6.
And g, transmitting the opening degree signal of the brake pedal, the opening degree signal of the accelerator pedal and the corner signal of the steering wheel, which are acquired by the angle sensor 4, to the lower computer 6 through a PWM (pulse width modulator) port of the lower computer 6.
And h, the lower computer 6 transmits the opening degree signal of the brake pedal, the opening degree signal of the accelerator pedal, the steering wheel angle signal and the muscle electric signal to the upper computer 7 in real time through an RS232 interface.
The characteristic parameters extracted by the comprehensive brake pedal opening signal, the accelerator pedal opening signal, the steering wheel corner signal, the muscle electric signal, the speed signal and the GPS coordinate signal comprise a vehicle speed average value, a vehicle speed standard deviation, an acceleration standard deviation, a positive acceleration average value, a positive acceleration standard deviation, a negative acceleration average value, a negative acceleration standard deviation, a required power coefficient standard deviation, an acceleration required power coefficient average value, a deceleration required power coefficient average value, a gastrocnemius electromyogram time domain characteristic root mean square, an electromyogram frequency domain characteristic average power frequency of gastrocnemius, an electromyogram time domain characteristic root mean square of tibialis anterior, an electromyogram frequency domain characteristic average power frequency of tibialis anterior, a sharp turning frequency, a lane changing frequency, a steering wheel stroke standard deviation, a steering wheel angular velocity average value, a steering wheel angular velocity standard deviation, an accelerator pedal stroke average value, The standard deviation of the travel of the accelerator pedal, the standard deviation of the rate of change of the accelerator pedal, the average value of the rate of change of the stepping on the accelerator pedal, the average value of the rate of change of the stepping off the accelerator pedal, the average value of the travel of the brake pedal, the standard deviation of the rate of change of the brake pedal, the average value of the rate of change of the stepping on the brake pedal and the average value of the.
In this embodiment, the lower computer adopts the STM32 singlechip, and angle sensor adopts incremental formula photoelectricity rotary encoder, and rotation on the steering wheel is transmitted to output shaft 15 through driving gear 13, driven gear 14 on.
The invention provides a driving style classification and identification method based on traffic flow density on the basis of extracting driving style characteristic parameters by using the experimental system, wherein the types of the traffic flow density are 4, the calculation process is shown in fig. 6, and the method specifically comprises the following steps:
s01: and acquiring characteristic parameters of drivers of different styles and types in a set period under the simulated road working conditions of different traffic densities by using an experimental system.
S02: and (5) performing synthesis and dimension reduction processing on all the characteristic parameters in the S01 by using a principal component analysis method to obtain comprehensive characteristic parameters.
S03: and (4) performing cluster analysis on the comprehensive characteristic parameters in the step (S02) by using a subtractive clustering algorithm method, so as to obtain the clustering centers and the clustering numbers of the comprehensive characteristic parameters of different driving styles under different traffic flow densities.
S04: and (5) correcting the comprehensive characteristic parameter clustering center in the K-means clustering algorithm S03 to obtain corrected comprehensive characteristic parameter clustering center results of various driving styles under different traffic densities.
S05: and calculating the characteristic parameters of various driving styles after correction, and establishing a characteristic parameter table of different driving styles under different traffic densities.
S06: and establishing a driving style recognition model based on a random forest algorithm, and training the driving style recognition model under each traffic density by using driving style data under different traffic densities.
In the step S03, the same parameters are used for the cluster analysis process of the comprehensive characteristic parameters by using a subtractive clustering algorithm under different traffic densities.
The following describes a specific embodiment of the driving style classification and recognition based on the traffic density.
(1) Based on the experimental system, a 3D Instructor 2 is used for simulating traffic flow density setting options of driving software, four urban working conditions with traffic flow densities of 10%, 40%, 70% and 100% are set, data such as vehicle speed, muscle electric signals, steering wheel angle, accelerator pedal opening degree and brake pedal opening degree are collected under the four traffic flow densities respectively through a driver in-loop experiment for 44 drivers, and 176 groups of effective test data samples are collected together (44 groups are collected under each traffic flow density working condition). According to the acquired experimental data, proper characteristic parameters are selected to represent the driving style, and the specific characteristic parameters are shown in table 1.
TABLE 1 Driving Style characteristic parameters and test data thereof
(2) According to the driving style characteristic parameters extracted from the table 1 and the test data thereof, the main components are comprehensively processed to obtain 29 main components, namely 29 combination modes of the original characteristic parameters, and the main component analysis is carried out on the basis,
and taking the first 4 principal components to comprehensively represent the characteristic parameters of the driving style. The scores of the first 4 principal components are shown in table 2, and a comprehensive characteristic parameter sample matrix is established according to the principal component score matrix.
TABLE 2 principal Components score
(3) Based on the comprehensive characteristic parameter sample matrix, extracting a comprehensive characteristic parameter clustering center by using a subtractive clustering algorithm, and setting X to { X ═ XiI | ═ 1,2,3, …, n } is a set of sample points in m-dimensional space, n is the number of sample points, and the specific algorithm steps are as follows:
1) for each sample point X in XiThe density function value is calculated according to the formula (1).
In the formula, raIs the neighborhood radius of the point.
The data point x with the highest density function valuec1As the first cluster center point, the corresponding density function value is Dc1。
2) Let x beckFor the k-th selected cluster center point, the corresponding density function value is DckThe density function value of each of the remaining sample points is modified according to equation (2).
In the formula: sfThe method is used for multiplying the radius value of the neighborhood of the center of the determined cluster, so that the possibility that the peripheral points are regarded as one part of the cluster is eliminated, and the problem of overlapped clustering caused by too dense cluster centers or the problem of underslassification caused by insufficient classification can be avoided by setting proper parameter values.
Selecting the sample point x with the highest density function valueck+1As a new cluster center point.
3) The value of expression (3) is determined by setting <1 to a predetermined parameter.
If the formula is not satisfied, the step 2) is carried out, and if the formula is satisfied, the subtractive clustering process is ended.
Under the condition of sample set determination, the number and the position of the clustering centers obtained by subtractive clustering are determined by parameters and SfAnd (4) determining. In the driving style classification algorithm provided herein, the driving style is divided into more appropriate classes through repeated comparison experiments, and a comprehensive characteristic parameter clustering center under four traffic densities is obtained, as shown in table 3.
TABLE 3 Integrated characteristic parameter clustering center under four traffic flow densities
As can be seen from table 3, the data of the comprehensive characteristic parameters at the four traffic densities are subjected to subtractive clustering using the same parameters, and then the data are properly classified into 3 types. Therefore, under various working conditions, the driving styles can be divided into 3 types, and the driving styles are divided into a cautious type, a steady type and an aggressive type according to the intensity of the aggressive degree of the driver driving the vehicle.
(4) And correcting the comprehensive characteristic parameter clustering center in the sample data by using a K-means clustering algorithm to obtain the corrected comprehensive characteristic parameter clustering center results of various driving styles under different traffic flow densities, wherein the specific algorithm comprises the following steps:
1) selecting k samples as initial clustering centers (z) according to a certain principle aiming at n samples1， z2，…，zk)。
2) Applying Euclidean distance will leave any sample xiTo the cluster center closest to them. Euclidean distance is the square root of the sum of the squares of the differences between all n variable values of two samples, i.e.
In the formula, xiIs the variable value of the ith variable of sample x; y isiIs the variable value of the ith variable of sample y.
3) The average of the objects in each cluster is recalculated and used as the new cluster center.
4) And repeating the steps until the cluster center is not changed any more.
The corrected comprehensive characteristic parameter clustering centers obtained after iteration is stopped are shown in table 4, and the obtained clustering results are shown in fig. 8 to fig. 10, wherein the clustering results 1,2 and 3 respectively represent the driving styles of the 1 st class, the 2 nd class and the 3 rd class. And calculating and analyzing the clustered samples, correcting the driving style characteristic parameters under different traffic flow densities, and establishing characteristic parameter tables of different driving styles under different traffic flow densities, as shown in tables 5 to 8.
TABLE 4 corrected comprehensive characteristic parameter clustering center
TABLE 5 Driving Style characteristic parameter value (traffic density 10%)
Table 6 driving style characteristic parameter value (traffic density 40%)
TABLE 7 Driving Style characteristic parameter value (traffic density 70%)
TABLE 8 Driving Style characteristic parameter value (traffic density of 100%)
(5) The method comprises the steps of establishing a driving style recognition model based on a random forest algorithm, training the driving style recognition model under the corresponding traffic density by using driving style data under various traffic densities, and taking a road working condition with the traffic density of 10% as an example, and explaining specific algorithm steps in detail.
For the working condition that the traffic density is 10%, the category of each sample in the 44 samples is obtained through clustering analysis, data processing is carried out on the basis, a matrix meeting the requirements of a random forest algorithm is constructed, and the data is shown in table 5.
TABLE 9 Driving Style characteristic parameters for identification (traffic density 10%)
The driving style recognition model training process is as follows:
and step 1, in N driving style characteristic parameter data samples, randomly selecting N samples (namely allowing repeated samples to exist) in a replacement mode to form 1 sampling set, and training 1 decision tree by using the sampling set.
And 2, setting that each sample has M characteristics, randomly selecting M characteristics (M < < M) from the M characteristics when each node of the decision tree needs to be split, traversing all possible splitting methods for each selected characteristic, respectively calculating Gini indexes of the characteristics, and finally selecting the characteristic corresponding to the minimum Gini index as the node splitting characteristic. Determining each node of the decision tree according to the method until the node can not be split or a threshold (such as the number of leaf nodes or the depth of the tree) set by us is reached, and establishing a decision tree at the moment;
and 3, repeating the step 1 and the step 2 until the number of the decision trees reaches a preset number, and forming a random forest model for driving style recognition.
The random forest model obtained by training by the method is used as a driving style recognition model, and the driving style of the driver under the road working conditions of the four traffic densities can be recognized.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned.
Claims (10)
1. An experimental system for extracting driving style characteristic parameters is characterized in that: including driver's operation panel, muscle signal of telecommunication sensor, next machine and host computer, wherein:
the driver operation console is used for simulating driving scenes with different traffic flow densities and can obtain opening information of a brake pedal, opening information of an accelerator pedal and steering wheel angle information in real time; identifying and storing a speed value of the vehicle in the driving process, and identifying and storing coordinates of the vehicle in a map;
the muscle electric signal sensor is used for acquiring gastrocnemius signals and tibialis anterior muscle signals of a driver;
the lower computer can transmit the information obtained by the driver console and the information detected by the muscle electric signal sensor to the upper computer;
the upper computer extracts driving characteristic parameters under different traffic densities through information transmitted by the lower computer, and performs multi-level hybrid algorithm processing on the driving characteristic parameters, so as to establish a driving style recognition model and train the driving style recognition model.
2. The experimental system for extracting the driving style characteristic parameters of claim 1, wherein: the driver operating platform is provided with a brake pedal, an accelerator pedal, a steering wheel, driving simulation software and visual recognition software, wherein the driving simulation software and the visual recognition software are arranged on the driver operating platform, the brake pedal, the accelerator pedal and the steering wheel detect opening information of the brake pedal, the opening information of the accelerator pedal and corner information of the steering wheel through an angle sensor, the driving simulation software can simulate four driving scenes with traffic density, and the visual recognition software comprises instrument panel recognition software and GPS coordinate recognition software.
3. The experimental system for extracting the driving style characteristic parameters as claimed in claim 2, wherein: the driving scenes of the four traffic densities are 10%, 40%, 70% and 100% of four urban work conditions respectively.
4. The experimental system for extracting the driving style characteristic parameters as claimed in claim 3, wherein: the lower computer is an STM32 single chip microcomputer.
5. The experimental system for extracting the driving style characteristic parameters as claimed in claim 4, wherein: the multilevel hybrid algorithm comprises a principal component analysis method, a subtraction clustering algorithm and a K-means clustering algorithm, and a driving style characteristic parameter table under different traffic flow densities is established, so that a driving style recognition model is formed.
6. The experimental system for extracting the driving style characteristic parameters as claimed in claim 5, wherein: the angle sensor is an incremental photoelectric rotary encoder.
7. The experimental system for extracting the driving style characteristic parameters as claimed in claim 6, wherein: the driving characteristic parameters comprise a vehicle speed average value, a vehicle speed standard deviation, an acceleration standard deviation, a positive acceleration average value, a positive acceleration standard deviation, a negative acceleration average value, a negative acceleration standard deviation, a demand power coefficient standard deviation, an acceleration demand power coefficient average value, a deceleration demand power coefficient average value, a gastrocnemius electromyography time domain characteristic root mean square, an electromyography frequency domain characteristic average power frequency of gastrocnemius, an electromyography time domain characteristic root mean square of tibialis anterior muscle, an electromyography frequency domain characteristic average power frequency of tibialis anterior muscle, a sharp turning frequency, a lane changing frequency, a steering wheel stroke standard deviation, a steering wheel angular velocity average value, a steering wheel angular velocity standard deviation, an accelerator pedal stroke average value, an accelerator pedal stroke standard deviation, an accelerator pedal change rate standard deviation, an accelerator pedal stepping change rate average value, an accelerator pedal releasing change rate average value, The average value of the travel of the brake pedal, the standard deviation of the change rate of the brake pedal, the average value of the change rate of stepping on the brake pedal and the average value of the change rate of loosening the brake pedal.
8. A driving style identification method implemented using the experimental system for extracting driving style characteristic parameters according to any one of claims 1 to 7, characterized in that: the method comprises the following steps:
s01: acquiring driving characteristic parameters of drivers of different styles and types in a set period under simulated road conditions of different traffic flow densities by using an experimental system;
s02: performing comprehensive and dimension reduction processing on all driving characteristic parameters in S01 by using a principal component analysis method to obtain comprehensive characteristic parameters;
s03: performing cluster analysis on the comprehensive characteristic parameters in the step S02 by using a subtractive clustering algorithm method, thereby obtaining the clustering centers and the clustering numbers of the comprehensive characteristic parameters of different driving styles under different traffic flow densities;
s04: correcting the comprehensive characteristic parameter clustering center in the K-means clustering algorithm S03 to obtain corrected comprehensive characteristic parameter clustering center results of various driving styles under different traffic densities;
s05: calculating various corrected driving style characteristic parameters, and establishing characteristic parameter tables of different driving styles under different traffic densities;
s06: and establishing a driving style recognition model based on a random forest algorithm, and training the driving style recognition model under each traffic density by using driving style data under different traffic densities.
9. The driving style recognition method according to claim 8, wherein: in the step S03, the same parameters are used for the cluster analysis process of the comprehensive characteristic parameters by using a subtractive clustering algorithm under different traffic densities.
10. The driving style recognition method according to claim 9, wherein: the training method of the driving style recognition model comprises the following steps:
step 1, in N driving style characteristic parameter samples, randomly selecting N samples in a back-to-back mode to form 1 sampling set, and training 1 decision tree by using the sampling set;
step 2, setting each sample to have M characteristics, randomly selecting M characteristics from the M characteristics when each node of the decision tree needs to be split, wherein M < M, traversing all possible splitting methods aiming at each selected characteristic, respectively calculating Gini indexes of the characteristics, finally selecting the characteristic corresponding to the minimum Gini index as the splitting characteristic of the node, determining each node of the decision tree according to the method until the node cannot be split or reaches a set threshold value, and establishing a decision tree at this time;
and 3, repeating the step 1 and the step 2 until the number of the decision trees reaches a preset number, and forming a random forest model for driving style recognition.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010540775.3A CN111731312A (en) | 2020-06-15 | 2020-06-15 | Experimental system for extracting driving style characteristic parameters and driving style identification method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010540775.3A CN111731312A (en) | 2020-06-15 | 2020-06-15 | Experimental system for extracting driving style characteristic parameters and driving style identification method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111731312A true CN111731312A (en) | 2020-10-02 |
Family
ID=72649090
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010540775.3A Pending CN111731312A (en) | 2020-06-15 | 2020-06-15 | Experimental system for extracting driving style characteristic parameters and driving style identification method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111731312A (en) |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112277845A (en) * | 2020-10-28 | 2021-01-29 | 中国第一汽车股份有限公司 | Vehicle accelerator pedal opening obtaining method, active sound production system and vehicle |
| CN112319488A (en) * | 2020-10-20 | 2021-02-05 | 易显智能科技有限责任公司 | Method and system for identifying driving style of motor vehicle driver |
| CN112373482A (en) * | 2020-11-23 | 2021-02-19 | 浙江天行健智能科技有限公司 | Driving habit modeling method based on driving simulator |
| CN112396118A (en) * | 2020-11-25 | 2021-02-23 | 浙江天行健智能科技有限公司 | GM-HMM-based driver acceleration intention modeling method |
| CN112466118A (en) * | 2020-11-25 | 2021-03-09 | 武汉光庭信息技术股份有限公司 | Vehicle driving behavior recognition method, system, electronic device and storage medium |
| CN112622920A (en) * | 2020-12-28 | 2021-04-09 | 华北科技学院 | Auxiliary driving system |
| CN112660140A (en) * | 2020-12-28 | 2021-04-16 | 华北科技学院 | Driving style classification method and device based on machine learning and electronic equipment |
| CN113096379A (en) * | 2021-03-03 | 2021-07-09 | 东南大学 | Driving style identification method based on traffic conflict |
| CN113246982A (en) * | 2021-04-01 | 2021-08-13 | 东风汽车集团股份有限公司 | Torque control method and device adaptive to driving style |
| CN113297685A (en) * | 2021-07-27 | 2021-08-24 | 中汽研（天津）汽车工程研究院有限公司 | Vehicle operation condition mode identification method |
| CN113392892A (en) * | 2021-06-08 | 2021-09-14 | 重庆大学 | Method and device for identifying driving habits of driver based on data fusion |
| CN113386779A (en) * | 2021-06-23 | 2021-09-14 | 华人运通(江苏)动力电池系统有限公司 | Driving style recognition method, device and storage medium |
| CN113581188A (en) * | 2021-06-30 | 2021-11-02 | 桂林电子科技大学 | Commercial vehicle driver driving style identification method based on Internet of vehicles data |
| CN113771863A (en) * | 2021-09-30 | 2021-12-10 | 长安大学 | Driver driving style identification method, medium and equipment based on real vehicle data |
| CN114030476A (en) * | 2021-10-27 | 2022-02-11 | 东风汽车集团股份有限公司 | Vehicle acceleration control method and related equipment |
| CN114403904A (en) * | 2021-12-31 | 2022-04-29 | 北京津发科技股份有限公司 | Device for determining muscle state based on electromyographic signals and muscle blood oxygen saturation |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110243433A1 (en) * | 2010-03-31 | 2011-10-06 | Microsoft Corporation | Classification and encoder selection based on content |
| CN107526906A (en) * | 2017-10-11 | 2017-12-29 | 吉林大学 | A kind of driving style device for identifying and method based on data acquisition |
| CN108629372A (en) * | 2018-05-07 | 2018-10-09 | 福州大学 | Obtain experimental system and the driving style recognition methods of driving style characteristic parameter |
| CN110378397A (en) * | 2019-06-27 | 2019-10-25 | 深圳大学 | A kind of driving style recognition methods and device |
| CN110858312A (en) * | 2018-08-23 | 2020-03-03 | 重庆大学 | Driver driving style classification method based on fuzzy C-means clustering algorithm |
-
2020
- 2020-06-15 CN CN202010540775.3A patent/CN111731312A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110243433A1 (en) * | 2010-03-31 | 2011-10-06 | Microsoft Corporation | Classification and encoder selection based on content |
| CN107526906A (en) * | 2017-10-11 | 2017-12-29 | 吉林大学 | A kind of driving style device for identifying and method based on data acquisition |
| CN108629372A (en) * | 2018-05-07 | 2018-10-09 | 福州大学 | Obtain experimental system and the driving style recognition methods of driving style characteristic parameter |
| CN110858312A (en) * | 2018-08-23 | 2020-03-03 | 重庆大学 | Driver driving style classification method based on fuzzy C-means clustering algorithm |
| CN110378397A (en) * | 2019-06-27 | 2019-10-25 | 深圳大学 | A kind of driving style recognition methods and device |
Cited By (22)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112319488A (en) * | 2020-10-20 | 2021-02-05 | 易显智能科技有限责任公司 | Method and system for identifying driving style of motor vehicle driver |
| CN112319488B (en) * | 2020-10-20 | 2022-06-03 | 易显智能科技有限责任公司 | Method and system for identifying driving style of motor vehicle driver |
| CN112277845A (en) * | 2020-10-28 | 2021-01-29 | 中国第一汽车股份有限公司 | Vehicle accelerator pedal opening obtaining method, active sound production system and vehicle |
| CN112373482B (en) * | 2020-11-23 | 2021-11-05 | 浙江天行健智能科技有限公司 | Driving habit modeling method based on driving simulator |
| CN112373482A (en) * | 2020-11-23 | 2021-02-19 | 浙江天行健智能科技有限公司 | Driving habit modeling method based on driving simulator |
| CN112396118A (en) * | 2020-11-25 | 2021-02-23 | 浙江天行健智能科技有限公司 | GM-HMM-based driver acceleration intention modeling method |
| CN112466118A (en) * | 2020-11-25 | 2021-03-09 | 武汉光庭信息技术股份有限公司 | Vehicle driving behavior recognition method, system, electronic device and storage medium |
| CN112622920A (en) * | 2020-12-28 | 2021-04-09 | 华北科技学院 | Auxiliary driving system |
| CN112660140A (en) * | 2020-12-28 | 2021-04-16 | 华北科技学院 | Driving style classification method and device based on machine learning and electronic equipment |
| CN113096379A (en) * | 2021-03-03 | 2021-07-09 | 东南大学 | Driving style identification method based on traffic conflict |
| CN113246982A (en) * | 2021-04-01 | 2021-08-13 | 东风汽车集团股份有限公司 | Torque control method and device adaptive to driving style |
| CN113246982B (en) * | 2021-04-01 | 2022-07-29 | 东风汽车集团股份有限公司 | Torque control method and device adaptive to driving style |
| CN113392892A (en) * | 2021-06-08 | 2021-09-14 | 重庆大学 | Method and device for identifying driving habits of driver based on data fusion |
| CN113386779A (en) * | 2021-06-23 | 2021-09-14 | 华人运通(江苏)动力电池系统有限公司 | Driving style recognition method, device and storage medium |
| CN113386779B (en) * | 2021-06-23 | 2022-10-18 | 华人运通(江苏)动力电池系统有限公司 | Driving style recognition method, device and storage medium |
| CN113581188A (en) * | 2021-06-30 | 2021-11-02 | 桂林电子科技大学 | Commercial vehicle driver driving style identification method based on Internet of vehicles data |
| CN113297685B (en) * | 2021-07-27 | 2021-11-02 | 中汽研（天津）汽车工程研究院有限公司 | Vehicle operation condition mode identification method |
| CN113297685A (en) * | 2021-07-27 | 2021-08-24 | 中汽研（天津）汽车工程研究院有限公司 | Vehicle operation condition mode identification method |
| CN113771863A (en) * | 2021-09-30 | 2021-12-10 | 长安大学 | Driver driving style identification method, medium and equipment based on real vehicle data |
| CN114030476A (en) * | 2021-10-27 | 2022-02-11 | 东风汽车集团股份有限公司 | Vehicle acceleration control method and related equipment |
| CN114030476B (en) * | 2021-10-27 | 2023-06-20 | 东风汽车集团股份有限公司 | Vehicle acceleration control method and related equipment |
| CN114403904A (en) * | 2021-12-31 | 2022-04-29 | 北京津发科技股份有限公司 | Device for determining muscle state based on electromyographic signals and muscle blood oxygen saturation |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111731312A (en) | Experimental system for extracting driving style characteristic parameters and driving style identification method | |
| CN107492251B (en) | Driver identity recognition and driving state monitoring method based on machine learning and deep learning | |
| CN111038485B (en) | Hybrid electric vehicle control method and system based on driving style recognition | |
| CN111267863B (en) | Driver driving type identification method and device, storage medium and terminal equipment | |
| CN111428960B (en) | Intelligent vehicle driving automatic evaluation method fusing multi-source vehicle-mounted sensor information | |
| CN107526906A (en) | A kind of driving style device for identifying and method based on data acquisition | |
| CN107657279A (en) | A kind of remote sensing target detection method based on a small amount of sample | |
| CN112801143B (en) | Steering road feel simulation method based on K-Means and Gaussian process regression | |
| CN110516746A (en) | A kind of driver's follow the bus behavior genre classification method based on no label data | |
| US10913455B2 (en) | Method for the improved detection of objects by a driver assistance system | |
| CN114140903B (en) | Road type recognition vehicle-mounted device based on decision tree generation rule | |
| CN114169444B (en) | Driving style classification method considering risk potential field distribution under vehicle following working condition | |
| CN112699793A (en) | Fatigue driving detection optimization identification method based on random forest | |
| CN109635830A (en) | For estimating the screening technique of the valid data of car mass | |
| CN110778714B (en) | Fuel vehicle gear identification method and system | |
| CN111474934A (en) | Personalized cruise control system based on high-precision map | |
| CN107953888A (en) | A kind of road type recognition methods and system | |
| CN108407814A (en) | A kind of driver characteristics discrimination method | |
| CN112466118A (en) | Vehicle driving behavior recognition method, system, electronic device and storage medium | |
| CN114162132B (en) | Driving mode identification method based on subjective and objective evaluation | |
| CN113743466B (en) | Road type identification method and system based on decision tree | |
| CN113642114B (en) | Personified random following driving behavior modeling method capable of making mistakes | |
| CN112550450B (en) | Road feel simulation method based on K-Means and CART regression tree | |
| CN117653121A (en) | Vehicle driving comfort evaluation method, system and device | |
| CN107862342A (en) | Lift the visual analysis system and method for tree-model |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20201002 |
|
| RJ01 | Rejection of invention patent application after publication |