CN112287461A - Automobile driving simulator braking system modeling method based on Gaussian process regression - Google Patents
Automobile driving simulator braking system modeling method based on Gaussian process regression Download PDFInfo
- Publication number
- CN112287461A CN112287461A CN202011206710.1A CN202011206710A CN112287461A CN 112287461 A CN112287461 A CN 112287461A CN 202011206710 A CN202011206710 A CN 202011206710A CN 112287461 A CN112287461 A CN 112287461A
- Authority
- CN
- China
- Prior art keywords
- brake system
- data
- brake
- modeling
- gaussian process
- 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.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/15—Vehicle, aircraft or watercraft design
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Geometry (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Computer Hardware Design (AREA)
- Automation & Control Theory (AREA)
- Aviation & Aerospace Engineering (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Regulating Braking Force (AREA)
Abstract
The invention discloses a method for modeling a brake system of an automobile driving simulator based on Gaussian process regression, which comprises the following steps: designing a mixed working condition; carrying out a brake system real vehicle test, and collecting original brake system data; screening original brake system data to obtain a brake system model training database and a model quality detection database; randomly selecting a brake system model training database, and modeling a brake system by using a Gaussian process regression modeling method after obtaining modeling point data; performing brake system model quality detection by using a model quality detection database; and judging whether to redesign the mixed working condition and carrying out the whole modeling according to the quality detection result of the braking system model. The method is based on the actual vehicle test data of the braking system, and a Gaussian process regression modeling method is used to obtain the braking system model of the automobile driving simulator with high confidence level, so that the dependence of the automobile driving system on the software and hardware of the real braking system is reduced, and the development and use cost is reduced.
Description
Technical Field
The invention relates to the technical field of automobiles, in particular to a modeling method of an automobile driving simulator braking system based on Gaussian process regression.
Background
Assisted driving and autonomous driving are developing at an unprecedented rate today with the rapid convergence of computer technology and traditional automotive technology. The automobile driving simulator is a development method which can greatly reduce the workload of drive tests and greatly reduce the test cost of real automobiles, and is used as an important development tool by intelligent driving developers. The brake system, which is an important component of a vehicle system, is directly related to the safety and dynamics of the vehicle, and is an important subsystem that must be taken into consideration in a driving system of the vehicle. However, since a brake system model with high confidence cannot be obtained, a large number of automobile driving systems choose to embed a real brake system into an automobile driving simulation hardware system, so as to obtain accurate brake data. The method not only increases the cost of the automobile driving system, but also increases the space occupied by the hardware of the automobile driving system to a certain extent, so that the automobile driving simulation system is large in size and expensive. Therefore, the accurate modeling method of the braking system can not only reduce the hardware volume of the automobile driving simulator, but also further reduce the development cost of the automobile driving system, and provide better automobile driving service for intelligent driving developers.
Patent CN201811029833.5 discloses a method and system for driving a car, which provides a method for recognizing a driver gesture signal and generating a car driving signal, but does not relate to a simulation method of a brake system. Patent CN201610892651.5 discloses an automobile driving system, which provides an automobile driving system using VR technology for human-computer interaction, but does not relate to a high-reliability brake system modeling method.
Disclosure of Invention
In order to solve the technical problems, the invention aims to provide a brake system modeling method of an automobile driving simulator based on Gaussian process regression, which is used for obtaining a brake system model with high confidence for the automobile driving system by using the Gaussian process regression modeling method on the basis of actual automobile test data of the brake system.
In order to achieve the purpose, the invention provides the following technical scheme:
the modeling method of the automobile driving simulator braking system based on Gaussian process regression comprises the following steps:
designing a mixed working condition;
carrying out a brake system real vehicle test, and collecting original brake system data;
screening original brake system data to obtain a brake system model training database and a model quality detection database;
randomly selecting a brake system model training database, and modeling a brake system by using a Gaussian process regression modeling method after obtaining modeling point data;
performing brake system model quality detection by using a model quality detection database;
and judging whether to redesign the mixed working condition and carrying out the whole modeling method according to the quality detection result of the braking system model.
Further, the designed mixed working condition comprises various pedal stroke change conditions in the working condition needing to be simulated by the automobile driving system, including different brake pedal strokes, different brake pedal stroke derivatives, the vehicle running speed, the road friction coefficient, the vehicle quality and the brake wheel cylinder pressure, so that the representative mixed working condition is obtained.
Furthermore, when the original brake system data are acquired in a brake system real-vehicle test, the change condition of the brake pedal under the designed mixed working condition must be strictly reproduced, two groups of original brake system data are acquired, and the acquisition of the two groups of original brake system data must be separated by more than one day; each set of raw brake system data contains six types of signal data, namely brake pedal travel, brake pedal travel derivative, vehicle travel speed, road friction coefficient, vehicle mass, and brake cylinder pressure.
Furthermore, when two groups of original brake system data obtained by the brake system real-vehicle test are screened, the same data except the pressure of a brake wheel cylinder are regarded as repeated data, and only one repeated data is reserved; and randomly selecting one of the two groups of screened original brake system data as a brake system model training database, and the other one of the two groups of screened original brake system data as a brake system model quality detection database.
Furthermore, when a vehicle brake system model is established by using a Gaussian process regression modeling method, a plurality of data points are randomly selected from a brake system model training database to obtain modeling point data, and then the modeling point data is used for carrying out Gaussian process regression modeling; the input variables of the obtained Gaussian process regression model are brake pedal travel, brake pedal travel derivative, vehicle running speed, road friction coefficient and vehicle mass; the output variable is the brake wheel cylinder pressure.
Furthermore, when the quality of the brake system model is detected, the quality detection database of the brake system model is used for detecting the confidence of the model, namely the travel of a brake pedal and the differential travel of the brake pedal in the quality detection database of the brake system model are used as model input data to obtain brake pressure prediction data; calculating the mean square error between the prediction data and the real brake pressure data in the brake system model quality detection database; if the mean square error is smaller than a certain threshold value, the brake system model has higher confidence, otherwise, the mixed working condition is redesigned, and the steps of the whole modeling method are carried out.
In a preferred embodiment, the threshold is 0.1.
Due to the adoption of the technical scheme, the invention achieves the following technical effects: the method is based on the actual vehicle test data of the braking system, uses the Gaussian process regression modeling method to obtain the high-confidence automobile driving system braking system model, and can give braking pressure with high confidence according to the position signal of the brake pedal, thereby reducing the dependence of the automobile driving system on the software and hardware of the actual braking system, reducing the development and use cost of the automobile driving system, and making up the defects of the prior art.
Drawings
FIG. 1 is a flow chart of the steps of the method for modeling a brake system of a driving simulator of an automobile based on Gaussian process regression according to the present invention.
Detailed Description
In order to make the technical solution of the embodiments of the present invention better understood, the technical solution of the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments, 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 obtained by equivalent changes and modifications by one skilled in the art based on the embodiments of the present invention, shall fall within the scope of the present invention.
As shown in FIG. 1, the invention provides a method for modeling a brake system of an automobile driving simulator based on Gaussian process regression, which comprises the following steps:
s1, designing a mixed working condition:
the designed mixed working condition comprises various pedal stroke change conditions in the working condition needing to be simulated by the automobile driving system, including different brake pedal strokes, different brake pedal stroke derivatives, the vehicle running speed, the road friction coefficient, the vehicle quality and the brake wheel cylinder pressure, so as to obtain a representative mixed working condition.
S2, carrying out a brake system real vehicle test:
when the original brake system data are acquired in the brake system real vehicle test, the change condition of the brake pedal under the designed mixed working condition must be strictly reproduced, two groups of original test data are acquired, and the acquisition of the two groups of original test data must be separated by more than one day. Each set of data includes six types of signal data, namely brake pedal travel, brake pedal travel derivative, vehicle travel speed, road friction coefficient, vehicle mass, and brake cylinder pressure.
S3, screening original brake system data to obtain a brake system model training database and a brake system model quality detection database:
when two groups of original test data obtained by a brake system real vehicle test are screened, the same data except the pressure of a brake wheel cylinder are regarded as repeated data. Only one of all data is retained for which duplicates occur.
S4, randomly selecting a brake system model training database to obtain modeling point data to perform brake system modeling;
and randomly selecting one of the two groups of screened original test data as a brake system model training database, and the other one of the two groups of screened original test data as a brake system model quality detection database. And randomly selecting one of the two groups of screened original test data as a brake system model training database, and the other one of the two groups of screened original test data as a brake system model quality detection database.
S5, performing brake system model quality detection by using a model quality monitoring database:
when a vehicle brake system model is established by using a Gaussian process regression modeling method, a plurality of data points are randomly selected from a brake system model training database to obtain modeling point data, and then the modeling point data is used for carrying out Gaussian process regression modeling. The input variables of the obtained Gaussian process regression model are brake pedal travel, brake pedal travel derivative, vehicle running speed, road friction coefficient and vehicle mass; the output variable is the brake wheel cylinder pressure.
The steps for performing gaussian process regression modeling using these modeling point data are as follows:
1. defining a training set
D＝(X,y)
Wherein: x ═ Xi}，y＝{yi}，xiRepresents the ith input sample (including brake pedal travel, brake pedal travel derivative, vehicle travel speed, road coefficient of friction, vehicle mass), yiIndicates the ith output value (brake wheel cylinder pressure).
Then there are:
y＝f(xn)+ξn
y mean u, kernel function k (x)i,xj). Noise matrix
2. Determining kernel functions
The choice is to use square exponential covariance function (squared expo)nearest covariance function, SE), solving the hyperparameters of the sum function by maximum likelihood estimation, including σn、σfAnd l. SE and function are defined as follows:
3. predicting new sample values
Given a new sample input x*Then the corresponding output is y*. According to Bayes principle, the value y is output*The joint distribution with the training samples is
Thus, the corresponding posterior distribution y may be calculated. The predicted output y may be expressed as:
y*|X,y,x*～N(μ,∑)
wherein the content of the first and second substances,
the mean value μ of the predicted distribution in the equation is actually an estimate of the test output.
S6, judging whether to perform test data acquisition again according to the quality detection result of the brake system model:
when the quality of the brake system model is detected, the brake system model detection database is used for detecting the confidence coefficient of the model, namely the brake pedal stroke and the differential of the brake pedal stroke in the brake system model detection database are used as model input data to obtain brake pressure prediction data; calculating the mean square error between the predicted data and the actual brake pressure data in the brake system model detection database; if the mean square error is smaller than a certain threshold, it indicates that the braking system model has higher confidence, otherwise, the process returns to S1 again. In the present embodiment, the threshold value is preferably 0.1.
The automobile driving simulator braking system model obtained by the modeling method can provide braking pressure with high certainty degree according to the position signal of the brake pedal.
The above description is only for the preferred embodiment of the present invention, and is not intended to limit the scope of the present invention; also, the above description should be understood as being readily apparent to those skilled in the relevant art and can be implemented, and therefore, other equivalent changes and modifications without departing from the concept disclosed herein are intended to be included within the scope of the present invention.
Claims (7)
1. A modeling method for a brake system of an automobile driving simulator based on Gaussian process regression is characterized by comprising the following steps:
designing a mixed working condition;
carrying out a brake system real vehicle test, and collecting original brake system data;
screening original brake system data to obtain a brake system model training database and a model quality detection database;
randomly selecting a brake system model training database, and modeling a brake system by using a Gaussian process regression modeling method after obtaining modeling point data;
performing brake system model quality detection by using a model quality detection database;
and judging whether to redesign the mixed working condition and carrying out the whole modeling method according to the quality detection result of the braking system model.
2. The method for modeling the brake system of the automobile driving simulator based on the Gaussian process regression as claimed in claim 1, wherein the designed mixed working condition comprises a plurality of pedal travel changes in the working condition to be simulated by the automobile driving system, including different brake pedal travel, different brake pedal travel derivatives, vehicle running speed, road friction coefficient, vehicle mass and brake cylinder pressure, so as to obtain a representative mixed working condition.
3. The method for modeling the brake system of the automobile driving simulator based on the gaussian process regression as claimed in claim 1, wherein when the actual vehicle test of the brake system is performed to collect the original brake system data, the change condition of the brake pedal of the designed mixed working condition must be strictly reproduced, and two sets of original brake system data are collected, wherein the collection of the two sets of original brake system data must be separated by more than one day; each set of raw brake system data contains six types of signal data, namely brake pedal travel, brake pedal travel derivative, vehicle travel speed, road friction coefficient, vehicle mass, and brake cylinder pressure.
4. The gaussian process regression-based modeling method for a brake system of an automotive driving simulator according to claim 3, wherein:
when two groups of original brake system data obtained by a brake system real-time test are screened, the data which are the same except the pressure of a brake wheel cylinder are regarded as repeated data, and only one repeated data is reserved;
and randomly selecting one of the two groups of screened original brake system data as a brake system model training database, and the other one of the two groups of screened original brake system data as a brake system model quality detection database.
5. The gaussian process regression-based modeling method for a brake system of an automotive driving simulator according to claim 3, wherein:
when a vehicle brake system model is established by using a Gaussian process regression modeling method, a plurality of data points are randomly selected from a brake system model training database to obtain modeling point data, and then the modeling point data is used for carrying out Gaussian process regression modeling; the input variables of the obtained Gaussian process regression model are brake pedal travel, brake pedal travel derivative, vehicle running speed, road friction coefficient and vehicle mass; the output variable is the brake wheel cylinder pressure.
6. The gaussian process regression-based modeling method for a brake system of an automotive driving simulator according to claim 5, wherein:
when the quality of the brake system model is detected, the quality detection database of the brake system model is used for detecting the confidence coefficient of the model, namely the brake pedal travel and the differential of the brake pedal travel in the quality detection database of the brake system model are used as model input data to obtain brake pressure prediction data; calculating the mean square error between the prediction data and the real brake pressure data in the brake system model quality detection database; if the mean square error is smaller than a certain threshold value, the brake system model has higher confidence, otherwise, the mixed working condition is redesigned, and the steps of the whole modeling method are carried out.
7. The gaussian process regression-based modeling method for a brake system of an automotive driving simulator according to claim 6, wherein: the threshold is 0.1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011206710.1A CN112287461B (en) | 2020-11-03 | 2020-11-03 | Automobile driving simulator braking system modeling method based on Gaussian process regression |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011206710.1A CN112287461B (en) | 2020-11-03 | 2020-11-03 | Automobile driving simulator braking system modeling method based on Gaussian process regression |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112287461A true CN112287461A (en) | 2021-01-29 |
| CN112287461B CN112287461B (en) | 2023-06-23 |
Family
ID=74352909
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011206710.1A Active CN112287461B (en) | 2020-11-03 | 2020-11-03 | Automobile driving simulator braking system modeling method based on Gaussian process regression |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112287461B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112949173A (en) * | 2021-02-26 | 2021-06-11 | 南京经纬达汽车科技有限公司 | Road feel simulation method based on K-Medoids and Gaussian process regression |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105825241A (en) * | 2016-04-15 | 2016-08-03 | 长春工业大学 | Driver braking intention identification method based on fuzzy neural network |
| US20180229723A1 (en) * | 2017-02-15 | 2018-08-16 | Ford Global Technologies, Llc | Feedback-Based Control Model Generation For An Autonomous Vehicle |
| CN110610009A (en) * | 2018-06-14 | 2019-12-24 | 复旦大学 | SRAM circuit yield analysis method based on Bayesian model |
| US20200089981A1 (en) * | 2018-09-17 | 2020-03-19 | Goodrich Corporation | Simulator to simulate target detection and recognition |
| CN111402577A (en) * | 2020-02-27 | 2020-07-10 | 北京交通大学 | Method for predicting parking time of driver at signalized intersection |
-
2020
- 2020-11-03 CN CN202011206710.1A patent/CN112287461B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105825241A (en) * | 2016-04-15 | 2016-08-03 | 长春工业大学 | Driver braking intention identification method based on fuzzy neural network |
| US20180229723A1 (en) * | 2017-02-15 | 2018-08-16 | Ford Global Technologies, Llc | Feedback-Based Control Model Generation For An Autonomous Vehicle |
| CN110610009A (en) * | 2018-06-14 | 2019-12-24 | 复旦大学 | SRAM circuit yield analysis method based on Bayesian model |
| US20200089981A1 (en) * | 2018-09-17 | 2020-03-19 | Goodrich Corporation | Simulator to simulate target detection and recognition |
| CN111402577A (en) * | 2020-02-27 | 2020-07-10 | 北京交通大学 | Method for predicting parking time of driver at signalized intersection |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112949173A (en) * | 2021-02-26 | 2021-06-11 | 南京经纬达汽车科技有限公司 | Road feel simulation method based on K-Medoids and Gaussian process regression |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112287461B (en) | 2023-06-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111238825B (en) | Intelligent driving automatic emergency braking performance testing method for combined test pavement | |
| JP6918137B2 (en) | Driving behavior evaluation method, device and computer-readable storage medium | |
| CN108769104B (en) | Road condition analysis and early warning method based on vehicle-mounted diagnosis system data | |
| CN113188807B (en) | Automatic abs result judging algorithm | |
| CN111126438B (en) | Driving behavior recognition method and system | |
| CN112163285B (en) | Modeling method of road surface type prediction model for simulating driving system | |
| CN113688558A (en) | Automobile driving condition construction method and system based on large database samples | |
| CN112287461B (en) | Automobile driving simulator braking system modeling method based on Gaussian process regression | |
| CN112985830A (en) | Abs result automatic judging algorithm | |
| CN113642114B (en) | Personified random following driving behavior modeling method capable of making mistakes | |
| CN106021639B (en) | CAE simulation analysis result-based damaged part damage judgment and classification method and maintenance man-hour estimation method | |
| CN111832442B (en) | Method for automatically separating temperature strain components from massive bridge dynamic strain data | |
| CN117033891A (en) | Traffic accident severity assessment method based on single vehicle traffic accident database | |
| Zhang et al. | Cycle condition identification of loader based on optimized KNN algorithm | |
| CN116975671A (en) | L3-level automatic driving vehicle driver trust evaluation method and system | |
| CN113156913A (en) | ABS fault diagnosis system and method | |
| CN108459996B (en) | User travel mining method and device based on supervised learning | |
| CN111881988B (en) | Heterogeneous unbalanced data fault detection method based on minority class oversampling method | |
| CN115221599A (en) | Chassis abnormal sound diagnosis method and system and automobile | |
| CN114169247A (en) | Method, device and equipment for generating simulated traffic flow and computer readable storage medium | |
| CN113267344A (en) | Automobile abs detection device and method | |
| CN114648114A (en) | Driver activity degree identification method and system based on limited Boltzmann machine | |
| CN117367775B (en) | Hydraulic braking system simulation test device and method | |
| CN115017014B (en) | Highway electromechanical monitoring system and method | |
| CN110553849A (en) | Driving condition evaluation system and method |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |