CN110346767A - A kind of test method and device for automobile lane change miscellaneous function - Google Patents
A kind of test method and device for automobile lane change miscellaneous function Download PDFInfo
- Publication number
- CN110346767A CN110346767A CN201910469119.6A CN201910469119A CN110346767A CN 110346767 A CN110346767 A CN 110346767A CN 201910469119 A CN201910469119 A CN 201910469119A CN 110346767 A CN110346767 A CN 110346767A
- Authority
- CN
- China
- Prior art keywords
- data
- bus
- lca
- neural network
- lane change
- 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
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/40—Means for monitoring or calibrating
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
- G06V20/58—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
- G06V20/584—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads of vehicle lights or traffic lights
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/08—Detecting or categorising vehicles
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Computer Networks & Wireless Communication (AREA)
- Multimedia (AREA)
- Traffic Control Systems (AREA)
- Image Analysis (AREA)
Abstract
The present invention relates to a kind of test method and device for automobile lane change miscellaneous function, test method is specifically includes the following steps: onboard installation camera captures the target vehicle at rear and side when this vehicle is run, acquisition video data；Obtain the bus data in vehicle CAN bus；Video data and bus data are sent to trained LCA test neural network model；LCA tests that neural network model alarms to correct in alarm scene and false alarm is classified, and records problem scenes.Compared with prior art, the present invention is conducive to save the cost of human and material resources, can have better accuracy and lower error rate, incidental when avoiding manual testing to fail to report, and integrally improves the quality of production and efficiency of product.
Description
Technical field
The present invention relates to a kind of automobile function detection fields, more particularly, to a kind of survey for automobile lane change miscellaneous function
Method for testing and device.
Background technique
Lane change auxiliary system (LCA) is to be detected by radar to the adjacent two sides lane of vehicle and rear；Obtain vehicle
The motion information of side and rear object, and the state of current vehicle is combined to be judged；It is finally reminded and is driven in a manner of sound, light etc.
The person of sailing；Driver is allowed to grasp best lane change opportunity, the traffic accident for preventing lane change from causing；Also have to rearward collision simultaneously relatively good
Prevention effect.
Existing test method for LCA is: two sides fill a camera respectively and are used to record video behind the side of vehicle, use
All signal datas in CANoe recording CAN bus, the signal of later period playback acquisition simultaneously analyze the relevant signal with LCA alarm, then
To be analyzed after camera data and CANoe data fusion with radar target testing tool, judge radar target whether and
Truth is consistent.This method needs manually to carry out data readback and analysis from collected video information and CAN bus information,
Failing to report for LCA function, the mistakes such as wrong report are searched, therefore expend a large amount of manpower and time, low efficiency, is easy to happen and misses error
The case where.
Summary of the invention
It is an object of the present invention to overcome the above-mentioned drawbacks of the prior art and provide one kind to be directed to automobile lane change
The test method and device of miscellaneous function.
The purpose of the present invention can be achieved through the following technical solutions:
A kind of test method for automobile lane change miscellaneous function, specifically includes the following steps:
S1. the target vehicle at rear and side, obtains video data when onboard installation camera captures the operation of this vehicle；
S2. the bus data in vehicle CAN bus is obtained；
S3. video data and bus data are sent to trained LCA and test neural network model；
S4.LCA tests that neural network model alarms to correct in alarm scene and false alarm is classified, and records
Problem scenes.
Further, in the step S1, after a camera is installed respectively at the two corners of roof rear end to capture
The target vehicle of side；Target vehicle of one camera to capture side is installed respectively at sided mirror unit.
Further, in the step S2, bus data includes the motion state number of the operation data of driver, vehicle
According to and radar issue doubling auxiliary alarm signal data.
Further, the time synchronization of the video data and bus data is corresponding.
Further, LCA test neural network model training process is as follows:
A1. using the video data of history and bus data as training sample, feature vector group is obtained according to training sample
Data, and feature vector group data is normalized, bus data is classified, according to the opportunity of doubling additional alarm and
Whether there is or not different label vector cbit is obtained, thus obtaining training sample is Y=(cbit, data), and wherein cbit is that target is defeated
Out, data is the input of neural network；
A2. LCA test neural network model is established, which includes input layer, hidden layer and output layer；
A3. LCA test neural network model is trained: model is initialized first, then by feature vector
Group data input model, is calculated in conjunction with weight matrix and bias matrix, pre- when being unsatisfactory for by backpropagation control algolithm
If when accuracy requirement, adjusting the number and network weight of hidden layer, until meeting accuracy requirement, LCA test nerve is saved
Network model parameter.
Further, the LCA test neural network model parameter includes Recognition with Recurrent Neural Network model, input neuron
Several and network weight.
A kind of test device for automobile lane change miscellaneous function, comprising:
Video acquiring module, the target vehicle at rear and side, obtains video data when for capturing the operation of this vehicle；
Bus acquisition module, for obtaining the bus data in vehicle CAN bus；
Data transmission module, for video data and bus data to be sent to testing analysis module；
Testing analysis module, by trained LCA test neural network model to alarm scene in it is correct alarm with
False alarm is classified, and records problem scenes.
Further, the video acquiring module includes four cameras, installs one respectively at the two corners of roof rear end
Target vehicle of a camera to capture rear；One camera is installed respectively at sided mirror unit to capture side
Target vehicle.
Further, in the bus acquisition module, the bus data of acquisition includes the operation data of driver, vehicle
The alarm signal data for the doubling auxiliary that motion state data and radar issue.
Further, the time synchronization of the video data and bus data is corresponding.
Compared with prior art, the invention has the following advantages that
1, the present invention tests neural network model by LCA and replaces traditional manual testing, nothing with the method for machine debugging
By there is very big advantage from manpower, time, cost, can have better accuracy and lower error rate, avoid people
Work is incidental when testing to be failed to report, and the quality of production and efficiency of product are integrally improved.
2, the present invention can allow the view of acquisition by installing video camera at the two corners of roof rear end and at sided mirror unit
Frequency is conducive to the reliability for improving functional test according to more accurate.
3, video data and the time of bus data should synchronous corresponding record can be convenient for the data correlation in later period, reduce work
It measures, improves data analysis efficiency.
4, LCA tests neural network model and has the advantage that a. can handle noise using neural network model: LCA is surveyed
After the completion of trying neural network model training, even if having part loss in the data of input, it still has the ability to recognize sample.B. not
It is easy to damage: because LCA test neural network model method in a distributed manner indicates data, when the damage of certain units,
It still can normally work.It c. can be with parallel processing.D. it can learn new alarm scene.
Detailed description of the invention
Fig. 1 is the structural diagram of the present invention.
Specific embodiment
The present invention is described in detail with specific embodiment below in conjunction with the accompanying drawings.The present embodiment is with technical solution of the present invention
Premised on implemented, the detailed implementation method and specific operation process are given, but protection scope of the present invention is not limited to
Following embodiments.
As described in Figure 1, present embodiments provide a kind of test method for automobile lane change miscellaneous function, specifically include with
Lower step:
Step S1 onboard installs the target vehicle at rear and side when camera captures the operation of this vehicle, obtains video counts
According to.Since the region of LCA concern is at the side rear of this vehicle, used so respectively installing a camera at two relief angles of this vehicle roof
To capture the target vehicle at rear, while a camera being also installed respectively to capture side at this vehicle sided mirror unit
Target vehicle.
It installs and its USB port is inserted into industrial personal computer after camera, using dedicated recording software recorded video, obtain video counts
According to.
Step S2 grabs bus data using CANoe (bus development environment) connection vehicle CAN bus, the number of buses
According to the alarm signal number for the doubling auxiliary that the operation data, the motion state data of vehicle and radar that include driver issue
According to.Video data and the time of bus data will synchronous corresponding record, be associated with when to analyze.
Step S3, video data and bus data are saved on the hard disk of industrial personal computer, then are sent out data by 5G technology
To cloud server, cloud server tests neural network model equipped with trained LCA.
After LCA that step S4, bus data and video data are transmitted to test neural network model, LCA test nerve net
Network model starts analysis alarm scene, classifies to correct alarm and false alarm, which the scene of false alarm classified again
A little is wrong report, which is failed to report, and corresponding problem scenes are recorded.Finally the result of verifying is sent by 5G technology again
To vehicle-mounted industrial personal computer, technical staff is fed back to, where facilitating technical staff quickly to navigate to LCA function problem, is greatly mentioned
Working efficiency is risen.
In the above-mentioned methods, it is as follows to test neural network model training process by LCA:
Step 1, using the video data of history and bus data as training sample, according to training sample obtain feature to
Amount group data, and feature vector group data is normalized, bus data is classified, according to doubling additional alarm when
Machine and whether there is or not different label vector cbit is obtained, thus obtaining training sample is Y=(cbit, data), and wherein cbit is mesh
Mark output, data are the input of neural network.
Step 2 establishes LCA test neural network model, which includes input layer, hidden layer and output
Layer.
Step 3 is trained LCA test neural network model: initializing first to model, then by feature
Vector Groups data input model, is calculated in conjunction with weight matrix and bias matrix, by backpropagation control algolithm, ought not
When meeting default accuracy requirement, the number and network weight of hidden layer are adjusted, until meeting accuracy requirement, LCA is saved and surveys
Try neural network model parameter.Neural network parameter includes Recognition with Recurrent Neural Network model, inputs neuron number and network weight,
Training terminates.
LCA test neural network model after training can identify that LCA is failed to report and reported by mistake well, and triggering one
The opportunity of secondary alarm.
The preferred embodiment of the present invention has been described in detail above.It should be appreciated that those skilled in the art without
It needs creative work according to the present invention can conceive and makes many modifications and variations.Therefore, all technologies in the art
Personnel are available by logical analysis, reasoning, or a limited experiment on the basis of existing technology under this invention's idea
Technical solution, all should be within the scope of protection determined by the claims.
Claims (10)
1. a kind of test method for automobile lane change miscellaneous function, which is characterized in that specifically includes the following steps:
S1. the target vehicle at rear and side, obtains video data when onboard installation camera captures the operation of this vehicle；
S2. the bus data in vehicle CAN bus is obtained；
S3. video data and bus data are sent to trained LCA and test neural network model；
S4.LCA tests that neural network model alarms to correct in alarm scene and false alarm is classified, and records problem
Scene.
2. the test method according to claim 1 for automobile lane change miscellaneous function, which is characterized in that the step S1
In, target vehicle of the camera to capture rear is installed respectively at the two corners of roof rear end；At sided mirror unit
Target vehicle of one camera to capture side is installed respectively.
3. the test method according to claim 1 for automobile lane change miscellaneous function, which is characterized in that the step S2
In, bus data includes the report for the doubling auxiliary that the operation data of driver, the motion state data of vehicle and radar issue
Alert signal data.
4. the test method according to claim 1 for automobile lane change miscellaneous function, which is characterized in that the video
Data and the time synchronization of bus data are corresponding.
5. the test method according to claim 1 for automobile lane change miscellaneous function, which is characterized in that the LCA
It is as follows to test neural network model training process:
A1. using the video data of history and bus data as training sample, feature vector group data is obtained according to training sample,
And feature vector group data is normalized, bus data is classified, according to the opportunity of doubling additional alarm and whether there is or not
Different label vector cbit is obtained, thus obtaining training sample is Y=(cbit, data), and wherein cbit is target output,
Data is the input of neural network；
A2. LCA test neural network model is established, which includes input layer, hidden layer and output layer；
A3. LCA test neural network model is trained: model is initialized first, then by feature vector group
Data input model, is calculated in conjunction with weight matrix and bias matrix, by backpropagation control algolithm, when being unsatisfactory for presetting
When accuracy requirement, the number and network weight of hidden layer are adjusted, until meeting accuracy requirement, LCA is saved and tests nerve net
Network model parameter.
6. the test method according to claim 5 for automobile lane change miscellaneous function, which is characterized in that the LCA is surveyed
Examination neural network model parameter includes Recognition with Recurrent Neural Network model, input neuron number and network weight.
7. a kind of test device for automobile lane change miscellaneous function characterized by comprising
Video acquiring module, the target vehicle at rear and side, obtains video data when for capturing the operation of this vehicle；
Bus acquisition module, for obtaining the bus data in vehicle CAN bus；
Data transmission module, for video data and bus data to be sent to testing analysis module；
Testing analysis module tests neural network model to the correct alarm and mistake in alarm scene by trained LCA
Alarm is classified, and records problem scenes.
8. the test device according to claim 7 for automobile lane change miscellaneous function, which is characterized in that the video obtains
Modulus block includes four cameras, installs target carriage of the camera to capture rear respectively at the two corners of roof rear end
?；Target vehicle of one camera to capture side is installed respectively at sided mirror unit.
9. the test method according to claim 7 for automobile lane change miscellaneous function, which is characterized in that the bus is adopted
Collect in module, the bus data of acquisition includes that the operation data of driver, the motion state data of vehicle and radar issue
The alarm signal data of doubling auxiliary.
10. the test method according to claim 7 for automobile lane change miscellaneous function, which is characterized in that the view
Frequency evidence and the time synchronization of bus data are corresponding.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910469119.6A CN110346767A (en) | 2019-05-31 | 2019-05-31 | A kind of test method and device for automobile lane change miscellaneous function |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910469119.6A CN110346767A (en) | 2019-05-31 | 2019-05-31 | A kind of test method and device for automobile lane change miscellaneous function |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110346767A true CN110346767A (en) | 2019-10-18 |
Family
ID=68174519
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910469119.6A Pending CN110346767A (en) | 2019-05-31 | 2019-05-31 | A kind of test method and device for automobile lane change miscellaneous function |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110346767A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111428317A (en) * | 2020-04-06 | 2020-07-17 | 宁波智诚祥科技发展有限公司 | Joint friction torque compensation method based on 5G and recurrent neural network |
| CN116935350A (en) * | 2023-09-18 | 2023-10-24 | 常州星宇车灯股份有限公司 | Method for identifying lane change of target vehicle based on machine learning |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104590122A (en) * | 2014-12-11 | 2015-05-06 | 重庆长安汽车股份有限公司 | Testing device for driveway deviation alarm system and method |
| CN106447045A (en) * | 2016-10-14 | 2017-02-22 | 重庆交通大学 | Evaluation method for ADAS (Advanced Driver Assistant System) based on machine learning |
| CN107479532A (en) * | 2017-08-04 | 2017-12-15 | 安徽江淮汽车集团股份有限公司 | The domain controller test system and method for a kind of intelligent automobile |
| CN108303271A (en) * | 2018-03-01 | 2018-07-20 | 北京理工华创电动车技术有限公司 | Auxiliary drives product testing system and test method |
| CN108445866A (en) * | 2018-03-13 | 2018-08-24 | 重庆大学 | LDW based on convolutional neural networks accidentally fails to report test method and test system |
| CN108829087A (en) * | 2018-07-19 | 2018-11-16 | 山东省科学院自动化研究所 | A kind of intelligent test system and test method of autonomous driving vehicle |
| JP2019021201A (en) * | 2017-07-20 | 2019-02-07 | 株式会社デンソー | Learning server, and assist system |
| CN109598066A (en) * | 2018-12-05 | 2019-04-09 | 百度在线网络技术（北京）有限公司 | Effect evaluation method, device, equipment and the storage medium of prediction module |
-
2019
- 2019-05-31 CN CN201910469119.6A patent/CN110346767A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104590122A (en) * | 2014-12-11 | 2015-05-06 | 重庆长安汽车股份有限公司 | Testing device for driveway deviation alarm system and method |
| CN106447045A (en) * | 2016-10-14 | 2017-02-22 | 重庆交通大学 | Evaluation method for ADAS (Advanced Driver Assistant System) based on machine learning |
| JP2019021201A (en) * | 2017-07-20 | 2019-02-07 | 株式会社デンソー | Learning server, and assist system |
| CN107479532A (en) * | 2017-08-04 | 2017-12-15 | 安徽江淮汽车集团股份有限公司 | The domain controller test system and method for a kind of intelligent automobile |
| CN108303271A (en) * | 2018-03-01 | 2018-07-20 | 北京理工华创电动车技术有限公司 | Auxiliary drives product testing system and test method |
| CN108445866A (en) * | 2018-03-13 | 2018-08-24 | 重庆大学 | LDW based on convolutional neural networks accidentally fails to report test method and test system |
| CN108829087A (en) * | 2018-07-19 | 2018-11-16 | 山东省科学院自动化研究所 | A kind of intelligent test system and test method of autonomous driving vehicle |
| CN109598066A (en) * | 2018-12-05 | 2019-04-09 | 百度在线网络技术（北京）有限公司 | Effect evaluation method, device, equipment and the storage medium of prediction module |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111428317A (en) * | 2020-04-06 | 2020-07-17 | 宁波智诚祥科技发展有限公司 | Joint friction torque compensation method based on 5G and recurrent neural network |
| CN111428317B (en) * | 2020-04-06 | 2023-06-09 | 宁波智诚祥科技发展有限公司 | Joint friction torque compensation method based on 5G and cyclic neural network |
| CN116935350A (en) * | 2023-09-18 | 2023-10-24 | 常州星宇车灯股份有限公司 | Method for identifying lane change of target vehicle based on machine learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109887281B (en) | Method and system for monitoring traffic incident | |
| CN107195024B (en) | Universal vehicle operation data record system and processing method | |
| CN102765365B (en) | Pedestrian detection method based on machine vision and pedestrian anti-collision warning system based on machine vision | |
| CN109345829B (en) | Unmanned vehicle monitoring method, device, equipment and storage medium | |
| CN102164270A (en) | Intelligent video monitoring method and system capable of exploring abnormal events | |
| CN109767626A (en) | A kind of curb parking method for inspecting, system, cruiser and server | |
| CN111856417B (en) | Performance analysis method, device, terminal and storage medium of vehicle millimeter wave radar | |
| CN103500518A (en) | False alarm detection method and device for forward collision alarm | |
| CN110346767A (en) | A kind of test method and device for automobile lane change miscellaneous function | |
| CN100440972C (en) | Mobile device for measuring speed and taking snap pictures | |
| CN114049624A (en) | Intelligent detection method and system for ship cabin based on machine vision | |
| CN115527364A (en) | Traffic accident tracing method and system based on radar vision data fusion | |
| CN112560546A (en) | Throwing behavior detection method and device and storage medium | |
| CN110111018B (en) | Method, device, electronic equipment and storage medium for evaluating vehicle sensing capability | |
| CN110443221A (en) | A kind of licence plate recognition method and system | |
| CN118247359A (en) | Automatic calibration method and device for fish-eye camera, computer equipment and storage medium | |
| CN113640013A (en) | Road test data processing method for driving assistance | |
| CN212724329U (en) | Illegal whistle and illegal refitted vehicle snapshot system covering bidirectional lane | |
| CN106448161A (en) | Road monitoring method and road monitoring device | |
| CN101667323A (en) | Real-time monitoring and alarming system for distraction of driver | |
| CN112991769A (en) | Traffic volume investigation method and device based on video | |
| CN210760742U (en) | Intelligent vehicle auxiliary driving system | |
| KR101736431B1 (en) | System and Method for Smart Traffic Monitoring Using Multiple Image | |
| CN110647884A (en) | Whistling snapshot method and related device | |
| CN114202817B (en) | ETC radio environment monitoring and guaranteeing method, system, equipment and medium |
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 |