CN108674413A - Traffic crash protection method and system - Google Patents
Traffic crash protection method and system Download PDFInfo
- Publication number
- CN108674413A CN108674413A CN201810483121.4A CN201810483121A CN108674413A CN 108674413 A CN108674413 A CN 108674413A CN 201810483121 A CN201810483121 A CN 201810483121A CN 108674413 A CN108674413 A CN 108674413A
- Authority
- CN
- China
- Prior art keywords
- pedestrian
- vehicle
- prediction
- data
- real
- 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
- 238000000034 method Methods 0.000 title claims abstract description 32
- 238000012549 training Methods 0.000 claims description 24
- 238000013528 artificial neural network Methods 0.000 claims description 14
- 230000008447 perception Effects 0.000 claims description 6
- 238000004140 cleaning Methods 0.000 claims description 3
- 238000001914 filtration Methods 0.000 claims description 3
- 230000001360 synchronised effect Effects 0.000 claims description 3
- 210000004218 nerve net Anatomy 0.000 claims 1
- 238000005516 engineering process Methods 0.000 abstract description 8
- 206010039203 Road traffic accident Diseases 0.000 abstract description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Classifications
-
- 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
- B60W30/00—Purposes of road vehicle drive control systems not related to the control of a particular sub-unit, e.g. of systems using conjoint control of vehicle sub-units
- B60W30/08—Active safety systems predicting or avoiding probable or impending collision or attempting to minimise its consequences
- B60W30/095—Predicting travel path or likelihood of collision
- B60W30/0953—Predicting travel path or likelihood of collision the prediction being responsive to vehicle dynamic parameters
-
- 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
- B60W30/00—Purposes of road vehicle drive control systems not related to the control of a particular sub-unit, e.g. of systems using conjoint control of vehicle sub-units
- B60W30/08—Active safety systems predicting or avoiding probable or impending collision or attempting to minimise its consequences
- B60W30/095—Predicting travel path or likelihood of collision
- B60W30/0956—Predicting travel path or likelihood of collision the prediction being responsive to traffic or environmental parameters
-
- 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
- B60W50/00—Details of control systems for road vehicle drive control not related to the control of a particular sub-unit, e.g. process diagnostic or vehicle driver interfaces
- B60W50/08—Interaction between the driver and the control system
- B60W50/14—Means for informing the driver, warning the driver or prompting a driver intervention
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/16—Anti-collision systems
- G08G1/166—Anti-collision systems for active traffic, e.g. moving vehicles, pedestrians, bikes
-
- 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
- B60W50/00—Details of control systems for road vehicle drive control not related to the control of a particular sub-unit, e.g. process diagnostic or vehicle driver interfaces
- B60W50/08—Interaction between the driver and the control system
- B60W50/14—Means for informing the driver, warning the driver or prompting a driver intervention
- B60W2050/143—Alarm means
-
- 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
- B60W2520/00—Input parameters relating to overall vehicle dynamics
- B60W2520/10—Longitudinal speed
-
- 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
- B60W2554/00—Input parameters relating to objects
- B60W2554/80—Spatial relation or speed relative to objects
Landscapes
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Transportation (AREA)
- Mechanical Engineering (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses a kind of traffic crash protection method and system, after predicting first predicted position and vehicle second predicted position setting time interval after of the pedestrian behind setting time interval, to combine the real time kinematics speed of pedestrian, it corresponds to after depicting the first predicted motion region of the pedestrian behind setting time interval, according to the coincidence situation in the first predicted motion region of pedestrian and the second predicted position of vehicle, it can predict to obtain the collision that vehicle and pedestrian may occur, and carry out anti-collision warning.Specific location of the present invention by dynamic prediction pedestrian and vehicle after following a period of time, it can accurately predict the position of acquisition pedestrian and vehicle, precision of prediction is higher, it can effectively predict the possibility to collide between vehicle, pedestrian, to effectively prevent the collision between pedestrian, vehicle, achieve the purpose that reduce traffic accident, can be widely applied in automobile intelligent technology.
Description
Technical field
The present invention relates to field of auxiliary is driven, more particularly to traffic crash protection method and system.
Background technology
With the substantial increase of automobile quantity, vehicle security drive technology is also more and more important.Wherein, traffic collides
Prevention technique is the major function in vehicle security drive, which can remind driver to watch for pedestrians, reduce collision accident
Occur, therefore, the accuracy of the technology has vital effect to vehicle security drive.
At present in technology, generally by the specific location of prediction pedestrian over time, then by judging the position
It sets and whether there is intersection point with the predetermined driving trace of vehicle, to predict whether to collide, to carry out alarm prompting.Due to
The movement of pedestrian has a uncertainty, then accurate prediction algorithm still can there are errors with physical location, therefore, technology at present
There is a problem of that forecasting inaccuracy is true, can not effectively prevent pedestrian, the case where vehicle collides.
Explanation of nouns：
UTM coordinate systems：UTM full name are UNIVERSAL TRANSVERSE MERCARTOR GRID SYSTEM, and Chinese is complete
Claim Universal Trans Meridian grid system, UTM coordinates are a kind of plane rectangular coordinates, this coordinate axiom system and its based on
Projection is widely used for topographic map, and the grid of reference and requirement as satellite image and natural resources database are accurately positioned
Application.
Invention content
In order to solve the above technical problems, the object of the present invention is to provide traffic crash protection method and system.
The technical solution adopted by the present invention to solve the technical problems is：
A kind of traffic crash protection method, includes the following steps：
According to the first real-time motion data of pedestrian, using pedestrian prediction model dynamic prediction pedestrian at setting time interval
The first predicted position afterwards；
According to the second real-time motion data of vehicle, using vehicle prediction model dynamic prediction vehicle at setting time interval
The second predicted position afterwards；
According to the first real time kinematics speed of pedestrian, pedestrian is depicted behind setting time interval in conjunction with the first predicted position
The first predicted motion region；
According to the coincidence situation in the first predicted motion region of pedestrian and the second predicted position of vehicle, prediction obtains vehicle
The collision that may occur with pedestrian, and carry out anti-collision warning.
Further, further comprising the steps of：
The second real-time motion data of the first real-time motion data of pedestrian and vehicle is synchronized based on timestamp.
Further, first real-time motion data and the first predicted position include position, speed and the movement of pedestrian
Direction, second real-time motion data and the second predicted position include position, speed and the direction of motion of vehicle.
Further, pedestrian's prediction model and vehicle prediction model are by using multi-layered perception neural networks or non-thread
Property recurrent neural networks dynamic training obtain, and the hidden layer of neural network use Bayes's canonical algorithm or back-propagation algorithm
It is calculated.
Further, pedestrian's prediction model and vehicle prediction model are real by corresponding first during dynamic prediction
When exercise data, the second real-time motion data carry out dynamic prediction in the following manner as training data：
Data cleansing is carried out to training data, filters out noise data；
According to the sample frequency and sampled point quantity of setting, the training data after cleaning is sampled, when obtaining different
Carve the input data of corresponding prediction model；
The input data of different moments is input to multi-layered perception neural networks or nonlinear-recurrent neural network in real time
In be trained, dynamic prediction obtains real-time predicted position behind setting time interval of pedestrian or vehicle.
Further, the dynamic prediction process is that prediction calculating is carried out under UTM coordinate systems, described to training data
It is further comprising the steps of before the step of carrying out data cleansing, filtering out noise data：
Training data is converted to the data of UTM coordinate systems by GPS coordinate system.
Further, first predicted motion region and the first real time kinematics speed are positive correlation.
Further, the prediction obtains the collision that vehicle and pedestrian may occur, and the step of carrying out anti-collision warning, specifically
Including：
According to the first predicted position and the second predicted position, calculates and obtain the distance between vehicle and pedestrian D；
According to the second real time kinematics speed of vehicle, the braking distance Ds for obtaining vehicle is calculated；
By the following conditions, prediction obtains the collision that vehicle and pedestrian may occur, and carries out corresponding anti-collision warning：
Condition one, as D-Ds-L ＞ Dw, judge that traffic will not collide；
Condition two works as Dw>D-Ds-L>When Dd, judge that vehicle has the possibility of collision with pedestrian, and alert vehicle and pay attention to going
People；
Condition three works as D-Ds-L<When Dd, judge that vehicle has the danger of collision with pedestrian, and alert vehicle and take defence
Measure, and/or actively take defensive measure；
Wherein, L indicates that vehicle body distance, Dw indicate that the radius of the corresponding warning zone of pedestrian, Dd indicate the corresponding danger of pedestrian
The radius in danger zone domain, and Dw and Dd are the description parameter in pedestrian's corresponding first predicted motion region.
The present invention solves another technical solution used by its technical problem：
Traffic collision avoidance system, including：
At least one processor；
At least one processor, for storing at least one program；
When at least one program is executed by least one processor so that at least one processor is realized
The traffic crash protection method.
The beneficial effects of the invention are as follows：This programme by predict first predicted position of the pedestrian behind setting time interval with
And after the second predicted position of the vehicle behind setting time interval, to combine the real time kinematics speed of pedestrian, correspondence to depict
Behind the first predicted motion region of the pedestrian behind setting time interval, according to the of the first predicted motion region of pedestrian and vehicle
The coincidence situation of two predicted positions can be predicted to obtain the collision that vehicle and pedestrian may occur, and carry out anti-collision warning.We
Specific location of the case by dynamic prediction pedestrian and vehicle after following a period of time can accurately be predicted to obtain pedestrian and Che
Position, precision of prediction is higher, can effectively predict the possibility to collide between vehicle, pedestrian, to effectively
Prevent the collision between pedestrian, vehicle, achievees the purpose that reduce traffic accident.
Description of the drawings
Fig. 1 is a kind of flow diagram of traffic crash protection method of the present invention；
Fig. 2 be the present invention a specific embodiment in pedestrian the first predicted motion region and vehicle between position relationship
Schematic diagram；
Fig. 3 is the structure diagram of the traffic collision avoidance system of the present invention.
Specific implementation mode
Embodiment of the method
Referring to Fig.1, the present invention provides a kind of traffic crash protection method, include the following steps：
According to the first real-time motion data of pedestrian, using pedestrian prediction model dynamic prediction pedestrian at setting time interval
The first predicted position after Δ t；
According to the second real-time motion data of vehicle, using vehicle prediction model dynamic prediction vehicle at setting time interval
The second predicted position after Δ t；
According to the first real time kinematics speed of pedestrian, pedestrian is depicted in setting time interval delta in conjunction with the first predicted position
The first predicted motion region after t；
According to the coincidence situation in the first predicted motion region of pedestrian and the second predicted position of vehicle, prediction obtains vehicle
The collision that may occur with pedestrian, and carry out anti-collision warning.
In the present solution, if vehicle periphery there are multiple pedestrians, this method is respectively adopted, prediction vehicle is sent out with each pedestrian
The possibility of raw collision, and carry out anti-collision warning.
This method obtains first predicted position and vehicle of the pedestrian after setting time interval of delta t by dynamic prediction
The second predicted position after setting time interval of delta t, to the real time kinematics speed in conjunction with pedestrian, correspondence depicts pedestrian and exists
The first predicted motion region after setting time interval of delta t, to according to the of the first predicted motion region of pedestrian and vehicle
The coincidence situation of two predicted positions can be predicted to obtain the collision that vehicle and pedestrian may occur, and carry out anti-collision warning.Due to
What in general vehicle running path was to determine, vehicle position after setting time interval of delta t be it is more determining, therefore, this
Scheme emphasis considers the uncertainty of pedestrian movement, carries out prediction of collision.This programme is existed by dynamic prediction pedestrian and vehicle
Specific location after following a period of time can accurately be predicted to obtain pedestrian and vehicle in conjunction with the uncertainty of pedestrian movement
The possibility to collide, precision of prediction is higher, can effectively predict the possibility to collide between vehicle, pedestrian, to
Effectively prevent the collision between pedestrian, vehicle, achievees the purpose that reduce traffic accident.
It is further used as preferred embodiment, it is further comprising the steps of：
The second real-time motion data of the first real-time motion data of pedestrian and vehicle is synchronized based on timestamp.Tool
In body processing procedure, if the sample frequency of the first real-time motion data and the second real-time motion data is different, it is also necessary to by data
It is converted into same frequency, so that vehicle and pedestrian data is corresponded, is convenient for mathematical operation.In actual test, only
The sample frequency of vehicle and pedestrian need to be allowed to be consistent.
It is further used as preferred embodiment, first real-time motion data and the first predicted position include pedestrian
Position, speed and the direction of motion, second real-time motion data and the second predicted position include position, the speed of vehicle
And the direction of motion.
After the first real-time motion data of pedestrian can be by the acquisition for mobile terminal of pedestrian, with automobile be attached to
It is sent to automobile, after the image that pedestrian can also be acquired in real time by the imaging sensor being arranged on automobile, is calculated by analyzing
It obtains.Second real-time motion data of vehicle is directly obtained by bus.
Preferably, in the present embodiment, pedestrian's prediction model is with vehicle prediction model by using Multilayer Perception god
Obtained through network or nonlinear-recurrent neural network dynamic training, and the hidden layer of neural network using Bayes's canonical algorithm or
Back-propagation algorithm is calculated.
It is noted that various ways foundation may be used with vehicle prediction model in pedestrian's prediction model, for example, it is simplest,
It is predicted using GPS inertial navigation technologies, it is assumed that the operating status before vehicle, pedestrian's maintenance passes through simple formula meter
Calculate the operating status after obtaining setting time interval of delta t.Can also realize pedestrian, vehicle motion prediction.
It is further used as preferred embodiment, pedestrian's prediction model is with vehicle prediction model in dynamic prediction process
In, using corresponding first real-time motion data, the second real-time motion data as training data, and in the following manner into action
State is predicted：
Data cleansing is carried out to training data, filters out noise data；
According to the sample frequency and sampled point quantity of setting, the training data after cleaning is sampled, when obtaining different
Carve the input data of corresponding prediction model；
The input data of different moments is input to multi-layered perception neural networks or nonlinear-recurrent neural network in real time
In be trained, dynamic prediction obtains real-time predicted position behind setting time interval of pedestrian or vehicle.
It in the present embodiment, is predicted using dynamic prediction mode, i.e. the input data trained each time of neural network
All it is dynamic, the predicted position of the vehicles or pedestrians after setting time interval of delta t to obtain each moment.
For example, setting sample frequency as 10Hz, sampled point quantity is 20, then when training each time, to current time
The training data in 2s is sampled before, and it is to use the historical track in 2s as dynamic prediction to be equivalent to training each time
Input data so moves in circles, at any time can dynamic prediction obtain pedestrian/vehicle real time position, overcome
The true problem of forecasting inaccuracy present in static prediction technology at present, can efficiently and accurately predict obtain pedestrian, vehicle reality
When position.
It is further used as preferred embodiment, the dynamic prediction process is that prediction calculating is carried out under UTM coordinate systems
, it is described that data cleansing is carried out to training data, it is further comprising the steps of before the step of filtering out noise data：
Training data is converted to the data of UTM coordinate systems by GPS coordinate system.By being calculated after conversion, Ke Yizhi
The coordinate distribution situation of ground observation training data is seen, and can be to avoid the interference of latitude, it is most important that, it is convenient for physics
The utilization of mathematical formulae calculates.
The first predicted motion region of pedestrian can be set as being obtained based on the description of probability theory algorithm, meet normal distribution
Rule, to which regional center represents the probability bigger that collides, edge represents the probability smaller to collide, when the of pedestrian
The overlapping positions of second predicted position of one predicted motion region and vehicle at edge, indicate the possibility to collide compared with
It is small, and when overlapping positions are relatively close to central area, it indicates that the possibility to collide is very big, therefore, calculating can be passed through
The coincidence situation in the first predicted motion region and the second predicted position, to judge possibility that pedestrian and vehicle collide.
Specifically, the first predicted motion region can be portrayed as round, sector etc., mainly according to pedestrian movement's data come
Described.Such as the randomness based on pedestrian movement, can be circle by the first predicted motion region description, or based on row
First predicted motion region description can be sector, subsequently be overlapped with the second predicted position by the advance characteristic of people's movement
Calculating, main calculating border circular areas or sector region overlap situation with the second predicted position, to carry out prediction of collision alarm.
As a preferred embodiment of the present invention, the first predicted motion region is set as circle.And specific description mode is such as
Under：
First predicted motion region and the first real time kinematics speed are positive correlation, and its positive correlation coefficient is from the
One coefficient data library, which is read, to be obtained, and first coefficient data library obtains in the following manner：
A large amount of pedestrian movement's speed is obtained with corresponding moving region as training data, after being trained, fitting obtains
The curved line relation between pedestrian movement's speed and moving region is obtained, to the different motion speed for obtaining pedestrian and corresponding movement
After positive correlation coefficient between region, foundation obtains the first coefficient data library.
As another preferred embodiment of the present invention, therefore, with reference to Fig. 2, the prediction, which obtains vehicle and pedestrian, to be occurred
Collision, and the step of carrying out anti-collision warning specifically include：
According to the first predicted position and the second predicted position, calculates and obtain the distance between vehicle and pedestrian D；
According to the second real time kinematics speed of vehicle, the braking distance Ds for obtaining vehicle is calculated；
By the following conditions, prediction obtains the collision that vehicle and pedestrian may occur, and carries out corresponding anti-collision warning：
Condition one, as D-Ds-L ＞ Dw, judge that traffic will not collide；
Condition two works as Dw>D-Ds-L>When Dd, judge that vehicle has the possibility of collision with pedestrian, and alert vehicle and pay attention to going
People；
Condition three works as D-Ds-L<When Dd, judge that vehicle has the danger of collision with pedestrian, and alert vehicle and take defence
Measure, and/or actively take defensive measure；
Wherein, L indicates that vehicle body distance, Dw indicate that the radius of the corresponding warning zone of pedestrian, Dd indicate the corresponding danger of pedestrian
The radius in danger zone domain, and Dw and Dd are the description parameter in pedestrian's corresponding first predicted motion region.
Preceding embodiment is mentioned, and the first predicted motion region of pedestrian meets probability distribution algorithm, and different zones represent hair
The possibility of raw collision it is of different sizes, therefore can when describing the first predicted motion region, describe a warning zone with
Danger zone, warning zone representative there is a possibility that collide, it should be noted that safe driving, danger zone represents and touches
The possibility hit is larger, and pedestrian is more dangerous into the region, therefore, it is necessary to alert vehicle to take defensive measure, and/or actively
Defensive measure is taken, such as can be braked with voice reminder, or is actively braked, travel direction is changed, changes traveling
Speed etc..
System embodiment
Reference Fig. 3, traffic collision avoidance system, including：
At least one processor 100；
At least one processor 200, for storing at least one program；
When at least one program is executed by least one processor 100 so that at least one processor
100 realize the traffic crash protection method.
The traffic collision avoidance system of the present embodiment, the traffic that executable the method for the present invention embodiment is provided
Crash protection method, the arbitrary combination implementation steps of executing method embodiment have the corresponding function of this method and beneficial to effect
Fruit.
It is to be illustrated to the preferable implementation of the present invention, but the invention is not limited to the implementation above
Example, those skilled in the art can also make various equivalent variations or be replaced under the premise of without prejudice to spirit of that invention
It changes, these equivalent modifications or replacement are all contained in the application claim limited range.
Claims (9)
1. a kind of traffic crash protection method, which is characterized in that include the following steps：
According to the first real-time motion data of pedestrian, using pedestrian prediction model dynamic prediction pedestrian behind setting time interval
First predicted position；
According to the second real-time motion data of vehicle, using vehicle prediction model dynamic prediction vehicle behind setting time interval
Second predicted position；
According to the first real time kinematics speed of pedestrian, of pedestrian behind setting time interval is depicted in conjunction with the first predicted position
One predicted motion region；
According to the coincidence situation in the first predicted motion region of pedestrian and the second predicted position of vehicle, prediction obtains vehicle and row
The collision that people may occur, and carry out anti-collision warning.
2. traffic crash protection method according to claim 1, which is characterized in that further comprising the steps of：
The second real-time motion data of the first real-time motion data of pedestrian and vehicle is synchronized based on timestamp.
3. traffic crash protection method according to claim 1, which is characterized in that first real-time motion data
Include position, speed and the direction of motion of pedestrian, second real-time motion data and the second prediction with the first predicted position
Position includes position, speed and the direction of motion of vehicle.
4. traffic crash protection method according to claim 1, which is characterized in that pedestrian's prediction model and vehicle
Prediction model is obtained by using multi-layered perception neural networks or nonlinear-recurrent neural network dynamic training, and nerve net
The hidden layer of network is calculated using Bayes's canonical algorithm or back-propagation algorithm.
5. traffic crash protection method according to claim 4, which is characterized in that pedestrian's prediction model and vehicle
Prediction model is during dynamic prediction, using corresponding first real-time motion data, the second real-time motion data as training
Data, and dynamic prediction is carried out in the following manner：
Data cleansing is carried out to training data, filters out noise data；
According to the sample frequency and sampled point quantity of setting, the training data after cleaning is sampled, obtains different moments pair
The input data for the prediction model answered；
In real time by the input data of different moments be input in multi-layered perception neural networks or nonlinear-recurrent neural network into
Row training, dynamic prediction obtain the real-time predicted position of pedestrian or vehicle behind setting time interval.
6. traffic crash protection method according to claim 5, which is characterized in that the dynamic prediction process is
Carry out under UTM coordinate systems prediction calculating, it is described that data cleansing is carried out to training data, the step of filtering out noise data it
Before, it is further comprising the steps of：
Training data is converted to the data of UTM coordinate systems by GPS coordinate system.
7. traffic crash protection method according to claim 5, which is characterized in that first predicted motion region
It is positive correlation with the first real time kinematics speed.
8. traffic crash protection method according to claim 1, which is characterized in that the prediction obtains vehicle and row
The collision that people may occur, and the step of carrying out anti-collision warning, specifically include：
According to the first predicted position and the second predicted position, calculates and obtain the distance between vehicle and pedestrian D；
According to the second real time kinematics speed of vehicle, the braking distance Ds for obtaining vehicle is calculated；
By the following conditions, prediction obtains the collision that vehicle and pedestrian may occur, and carries out corresponding anti-collision warning：
Condition one, as D-Ds-L ＞ Dw, judge that traffic will not collide；
Condition two works as Dw>D-Ds-L>When Dd, judge that vehicle has the possibility of collision with pedestrian, and alert vehicle and watch for pedestrians；
Condition three works as D-Ds-L<When Dd, judge that vehicle has the danger of collision with pedestrian, and alert vehicle and take defensive measure,
And/or actively take defensive measure；
Wherein, L indicates that vehicle body distance, Dw indicate that the radius of the corresponding warning zone of pedestrian, Dd indicate the corresponding danger area of pedestrian
The radius in domain, and Dw and Dd are the description parameter in pedestrian's corresponding first predicted motion region.
9. traffic collision avoidance system, which is characterized in that including：
At least one processor；
At least one processor, for storing at least one program；
When at least one program is executed by least one processor so that at least one processor is realized as weighed
Profit requires 1-8 any one of them traffic crash protection methods.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810483121.4A CN108674413B (en) | 2018-05-18 | 2018-05-18 | Vehicle and pedestrian collision prevention method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810483121.4A CN108674413B (en) | 2018-05-18 | 2018-05-18 | Vehicle and pedestrian collision prevention method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108674413A true CN108674413A (en) | 2018-10-19 |
| CN108674413B CN108674413B (en) | 2021-02-19 |
Family
ID=63805338
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810483121.4A Active CN108674413B (en) | 2018-05-18 | 2018-05-18 | Vehicle and pedestrian collision prevention method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108674413B (en) |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110293968A (en) * | 2019-06-18 | 2019-10-01 | 百度在线网络技术（北京）有限公司 | Control method, device, equipment and the readable storage medium storing program for executing of automatic driving vehicle |
| CN110443310A (en) * | 2019-08-07 | 2019-11-12 | 浙江大华技术股份有限公司 | Compare update method, server and the computer storage medium of analysis system |
| CN111081067A (en) * | 2019-12-27 | 2020-04-28 | 武汉大学 | Vehicle collision early warning system and method based on IGA-BP neural network under vehicle networking environment |
| CN111223302A (en) * | 2018-11-23 | 2020-06-02 | 明创能源股份有限公司 | External coordinate real-time three-dimensional road condition auxiliary device for mobile carrier and system |
| CN111634292A (en) * | 2020-05-18 | 2020-09-08 | 北京踏歌智行科技有限公司 | Collision prediction method for mining area |
| CN111739344A (en) * | 2020-06-29 | 2020-10-02 | 北京百度网讯科技有限公司 | Early warning method and device and electronic equipment |
| CN111746555A (en) * | 2019-03-26 | 2020-10-09 | 通用汽车环球科技运作有限责任公司 | Identification and avoidance of collision behavior |
| CN112349144A (en) * | 2020-11-10 | 2021-02-09 | 中科海微(北京)科技有限公司 | Monocular vision-based vehicle collision early warning method and system |
| CN113044028A (en) * | 2019-12-10 | 2021-06-29 | 本田技研工业株式会社 | Information presentation device for autonomous vehicle |
| CN113795874A (en) * | 2019-05-14 | 2021-12-14 | 大众汽车股份公司 | Method for detecting a potential collision of a vehicle with a living being and parking lot management system |
| CN114187764A (en) * | 2022-02-16 | 2022-03-15 | 深圳佑驾创新科技有限公司 | Method for rapidly detecting collision risk degree aiming at VRU passing scene |
| CN114228707A (en) * | 2021-12-22 | 2022-03-25 | 广西科技大学 | Anti-collision method and system for unmanned vehicle |
| CN114511999A (en) * | 2020-11-17 | 2022-05-17 | 郑州宇通客车股份有限公司 | Pedestrian behavior prediction method and device |
| CN118053322A (en) * | 2024-02-04 | 2024-05-17 | 天津大学 | Engineering vehicle collision detection early warning system and method based on computer vision |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110227713A1 (en) * | 2010-03-16 | 2011-09-22 | GM Global Technology Operations LLC | Method for the avoidance or mitigation of a collision, control apparatus for a driver Assistance system and vehicle |
| US20110249118A1 (en) * | 2010-04-13 | 2011-10-13 | Becker Research And Development (Proprietary) Limited | Apparatus, a system and a method for collission avoidance |
| CN102765365A (en) * | 2011-05-06 | 2012-11-07 | 香港生产力促进局 | Pedestrian detection method based on machine vision and pedestrian anti-collision warning system based on machine vision |
| US8868325B2 (en) * | 2010-04-05 | 2014-10-21 | Toyota Jidosha Kabushiki Kaisha | Collision judgment apparatus for vehicle |
| CN103782330B (en) * | 2011-09-14 | 2015-12-16 | 丰田自动车株式会社 | Drive assistance device and driving assistance method |
| CN105761488A (en) * | 2016-03-30 | 2016-07-13 | 湖南大学 | Real-time limit learning machine short-time traffic flow prediction method based on fusion |
-
2018
- 2018-05-18 CN CN201810483121.4A patent/CN108674413B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110227713A1 (en) * | 2010-03-16 | 2011-09-22 | GM Global Technology Operations LLC | Method for the avoidance or mitigation of a collision, control apparatus for a driver Assistance system and vehicle |
| US8868325B2 (en) * | 2010-04-05 | 2014-10-21 | Toyota Jidosha Kabushiki Kaisha | Collision judgment apparatus for vehicle |
| US20110249118A1 (en) * | 2010-04-13 | 2011-10-13 | Becker Research And Development (Proprietary) Limited | Apparatus, a system and a method for collission avoidance |
| CN102765365A (en) * | 2011-05-06 | 2012-11-07 | 香港生产力促进局 | Pedestrian detection method based on machine vision and pedestrian anti-collision warning system based on machine vision |
| CN103782330B (en) * | 2011-09-14 | 2015-12-16 | 丰田自动车株式会社 | Drive assistance device and driving assistance method |
| CN105761488A (en) * | 2016-03-30 | 2016-07-13 | 湖南大学 | Real-time limit learning machine short-time traffic flow prediction method based on fusion |
Cited By (22)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111223302A (en) * | 2018-11-23 | 2020-06-02 | 明创能源股份有限公司 | External coordinate real-time three-dimensional road condition auxiliary device for mobile carrier and system |
| CN111223302B (en) * | 2018-11-23 | 2021-12-03 | 明创能源股份有限公司 | External coordinate real-time three-dimensional road condition auxiliary device for mobile carrier and system |
| CN111746555A (en) * | 2019-03-26 | 2020-10-09 | 通用汽车环球科技运作有限责任公司 | Identification and avoidance of collision behavior |
| CN113795874B (en) * | 2019-05-14 | 2024-02-06 | 大众汽车股份公司 | Method for identifying potential collision of vehicle and living beings and parking lot management system |
| CN113795874A (en) * | 2019-05-14 | 2021-12-14 | 大众汽车股份公司 | Method for detecting a potential collision of a vehicle with a living being and parking lot management system |
| CN110293968A (en) * | 2019-06-18 | 2019-10-01 | 百度在线网络技术（北京）有限公司 | Control method, device, equipment and the readable storage medium storing program for executing of automatic driving vehicle |
| CN110443310B (en) * | 2019-08-07 | 2022-08-09 | 浙江大华技术股份有限公司 | Updating method of comparison analysis system, server and computer storage medium |
| CN110443310A (en) * | 2019-08-07 | 2019-11-12 | 浙江大华技术股份有限公司 | Compare update method, server and the computer storage medium of analysis system |
| CN113044028A (en) * | 2019-12-10 | 2021-06-29 | 本田技研工业株式会社 | Information presentation device for autonomous vehicle |
| CN111081067A (en) * | 2019-12-27 | 2020-04-28 | 武汉大学 | Vehicle collision early warning system and method based on IGA-BP neural network under vehicle networking environment |
| CN111634292A (en) * | 2020-05-18 | 2020-09-08 | 北京踏歌智行科技有限公司 | Collision prediction method for mining area |
| CN111634292B (en) * | 2020-05-18 | 2021-10-19 | 北京踏歌智行科技有限公司 | Collision prediction method for mining area |
| CN111739344A (en) * | 2020-06-29 | 2020-10-02 | 北京百度网讯科技有限公司 | Early warning method and device and electronic equipment |
| US11645909B2 (en) | 2020-06-29 | 2023-05-09 | Beijing Baidu Netcom Science Technology Co., Ltd. | Method, apparatus and electronic device for early-warning |
| CN112349144B (en) * | 2020-11-10 | 2022-04-19 | 中科海微(北京)科技有限公司 | Monocular vision-based vehicle collision early warning method and system |
| CN112349144A (en) * | 2020-11-10 | 2021-02-09 | 中科海微(北京)科技有限公司 | Monocular vision-based vehicle collision early warning method and system |
| CN114511999A (en) * | 2020-11-17 | 2022-05-17 | 郑州宇通客车股份有限公司 | Pedestrian behavior prediction method and device |
| CN114511999B (en) * | 2020-11-17 | 2023-09-01 | 宇通客车股份有限公司 | Pedestrian behavior prediction method and device |
| CN114228707A (en) * | 2021-12-22 | 2022-03-25 | 广西科技大学 | Anti-collision method and system for unmanned vehicle |
| CN114228707B (en) * | 2021-12-22 | 2023-04-25 | 广西科技大学 | Anti-collision method and system for unmanned vehicle |
| CN114187764A (en) * | 2022-02-16 | 2022-03-15 | 深圳佑驾创新科技有限公司 | Method for rapidly detecting collision risk degree aiming at VRU passing scene |
| CN118053322A (en) * | 2024-02-04 | 2024-05-17 | 天津大学 | Engineering vehicle collision detection early warning system and method based on computer vision |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108674413B (en) | 2021-02-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108674413A (en) | Traffic crash protection method and system | |
| Wu et al. | A novel method of vehicle-pedestrian near-crash identification with roadside LiDAR data | |
| Shi et al. | Key risk indicators for accident assessment conditioned on pre-crash vehicle trajectory | |
| Laureshyn et al. | Cross-comparison of three surrogate safety methods to diagnose cyclist safety problems at intersections in Norway | |
| Shirazi et al. | Looking at intersections: a survey of intersection monitoring, behavior and safety analysis of recent studies | |
| Alozi et al. | Evaluating the safety of autonomous vehicle–pedestrian interactions: An extreme value theory approach | |
| JP2014528118A (en) | System and method for determining highway throughput | |
| CN106133805A (en) | Driving assistance method and system for collision elimination | |
| Lv et al. | Automatic vehicle-pedestrian conflict identification with trajectories of road users extracted from roadside LiDAR sensors using a rule-based method | |
| CN112912883B (en) | Simulation method and related equipment | |
| CN112800166B (en) | Community correction object activity track supervision and early warning method, system and device | |
| Wei et al. | Trajectory-based identification of critical instantaneous decision events at mixed-flow signalized intersections | |
| Ma et al. | Left-turn conflict identification at signal intersections based on vehicle trajectory reconstruction under real-time communication conditions | |
| CN115100863B (en) | Road monitoring method, device, equipment and storage medium | |
| Mohammadnazar et al. | Incorporating driving volatility measures in safety performance functions: Improving safety at signalized intersections | |
| Chen et al. | Network-level safety metrics for overall traffic safety assessment: A case study | |
| Anwari et al. | Investigating surrogate safety measures at midblock pedestrian crossings using multivariate models with roadside camera data | |
| Li et al. | A probabilistic framework for estimating the risk of pedestrian-vehicle conflicts at intersections | |
| Saito et al. | Context-sensitive hazard anticipation based on driver behavior analysis and cause-and-effect chain study | |
| Detzer et al. | Analysis of traffic safety for cyclists: The automatic detection of critical traffic situations for cyclists | |
| Yin et al. | Detecting illegal pickups of intercity buses from their gps traces | |
| Swargiary et al. | A study on meta-analysis approach for pedestrian-vehicle interaction using LiDAR | |
| Gerogiannis et al. | Analysis of long-term observational data on pedestrian road crossings at unmarked locations | |
| Islam et al. | Understanding the impact of vehicle dynamics, geometric and non-geometric roadway attributes on surrogate safety measure using connected vehicle data | |
| Autey | Before and after traffic safety evaluations using computer vision techniques |
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 | ||
| CP02 | Change in the address of a patent holder |
Address after: 510000 No.8 Songgang street, Cencun, Tianhe District, Guangzhou City, Guangdong Province Patentee after: GUANGZHOU XPENG AUTOMOBILE TECHNOLOGY Co.,Ltd. Address before: Room 333, Guangzhou, China Patentee before: GUANGZHOU XPENG AUTOMOBILE TECHNOLOGY Co.,Ltd. |
|
| CP02 | Change in the address of a patent holder |