CN109557912A - A kind of decision rule method of automatic Pilot job that requires special skills vehicle - Google Patents
A kind of decision rule method of automatic Pilot job that requires special skills vehicle Download PDFInfo
- Publication number
- CN109557912A CN109557912A CN201811183181.0A CN201811183181A CN109557912A CN 109557912 A CN109557912 A CN 109557912A CN 201811183181 A CN201811183181 A CN 201811183181A CN 109557912 A CN109557912 A CN 109557912A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- trajectory
- decision
- mode
- planning
- 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 33
- 230000007613 environmental effect Effects 0.000 claims abstract description 14
- 230000008447 perception Effects 0.000 claims abstract description 7
- 238000000354 decomposition reaction Methods 0.000 claims abstract 3
- 230000001133 acceleration Effects 0.000 claims description 6
- 238000005259 measurement Methods 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 5
- 230000003068 static effect Effects 0.000 claims description 3
- 230000009897 systematic effect Effects 0.000 claims description 3
- 230000004927 fusion Effects 0.000 abstract description 3
- 230000008901 benefit Effects 0.000 abstract description 2
- 230000004888 barrier function Effects 0.000 description 64
- 230000009191 jumping Effects 0.000 description 17
- 238000010586 diagram Methods 0.000 description 6
- 230000001953 sensory effect Effects 0.000 description 5
- 238000001514 detection method Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 241001417527 Pempheridae Species 0.000 description 1
- 208000027418 Wounds and injury Diseases 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000001174 ascending effect Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 239000004568 cement Substances 0.000 description 1
- 230000008094 contradictory effect Effects 0.000 description 1
- 230000006378 damage Effects 0.000 description 1
- 238000006073 displacement reaction Methods 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 208000014674 injury Diseases 0.000 description 1
- 230000002147 killing effect Effects 0.000 description 1
- 230000000670 limiting effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000007115 recruitment Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000002829 reductive effect Effects 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 239000002689 soil Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
- G05D1/0223—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory involving speed control of the vehicle
Landscapes
- Engineering & Computer Science (AREA)
- Aviation & Aerospace Engineering (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Control Of Driving Devices And Active Controlling Of Vehicle (AREA)
- Traffic Control Systems (AREA)
- Steering Control In Accordance With Driving Conditions (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
本发明涉及一种自动驾驶特种作业车辆的决策规划方法。包括以步骤：1)自动驾驶作业模块获取自车的当前定位位姿；2)将感知系统发送的环境信息投影到栅格地图，并生成环境地图；3)自动驾驶作业模块获取当前作业执行器的控制指令并下发；4)自动驾驶作业模块获取任务参考路径，采用路径‑速度分解的轨迹规划方法结合车辆动力学约束进行轨迹簇规划，获取车辆可执行的基础轨迹簇，将基础轨迹簇和任务参考路径融合得到可执行轨迹簇；5)对规划的可执行轨迹簇进行安全性和高效性的择优，最终生成高收益轨迹。与现有技术相比，本发明具有提高避障成功率、自动决策、多模式的轨迹决策策略、实现自动驾驶安全性等优点。
The invention relates to a decision-making and planning method for an automatic driving special operation vehicle. It includes the following steps: 1) the automatic driving operation module obtains the current positioning pose of the vehicle; 2) the environmental information sent by the perception system is projected to the grid map, and the environment map is generated; 3) the automatic driving operation module obtains the current operation executor 4) The autonomous driving operation module obtains the task reference path, uses the path-velocity decomposition trajectory planning method combined with the vehicle dynamics constraints to plan the trajectory cluster, obtains the basic trajectory cluster executable by the vehicle, and combines the basic trajectory cluster Fusion with the task reference path to obtain the executable trajectory cluster; 5) The planned executable trajectory cluster is selected for safety and efficiency, and finally a high-yield trajectory is generated. Compared with the prior art, the present invention has the advantages of improving the success rate of obstacle avoidance, automatic decision-making, multi-mode trajectory decision-making strategy, and realizing the safety of automatic driving.
Description
Technical field
The present invention relates to track of vehicle planning fields, advise more particularly, to a kind of decision of automatic Pilot job that requires special skills vehicle
The method of drawing.
Background technique
In recent years, the developing by leaps and bounds of artificial intelligence technology, computer hardware operational capability is substantially improved, sensory perceptual system not
It is disconnected to improve and vehicle electric, reaching its maturity for line traffic control make it possible the landing of automatic Pilot technology.It is unmanned to multiply
It is current automatic Pilot technical application direction the fiercest with vehicle, including as Google Waymo, Baidu Apollo etc., however, class
There are the need more more eager to automatic Pilot technology than passenger car like the job that requires special skills vehicle of mine vehicle, sweeper, slag-soil truck etc.
It asks, the considerations of this is not only for traffic safety, of equal importance there are also the reduction of driver's work load, recruitment demands to increase
Add and the insufficient contradictory alleviation of experienced operator.Thus, the automatic Pilot skill of job that requires special skills vehicle will be the another of market competition
A focus.
Existing automatic Pilot job that requires special skills vehicle, trajectory planning often consider one curve of the primary system plan, in order to
Avoiding barrier, it is necessary to get around barrier beginning to planning apart from the farther away place of barrier；Track decision or use are protected
The strategy kept encounters barrier and just stops or take radical prescription, encounters barrier and directly detour.This all will greatly be dropped
The operating efficiency of low job that requires special skills vehicle.Further, since measurement caused by the FOV of sensory perceptual system, resolution ratio and measurement accuracy misses
Difference, the systematic error after rasterizing and barrier block, pass to the environmental map of decision rule there are it is serious not really
It is qualitative, consider for the operation safety to automatic Pilot job that requires special skills vehicle, programmed decision-making needs can be compatible with these uncertainties.
Therefore, how to provide a kind of automatic Pilot job that requires special skills vehicle decision rule method solve the above problems it is automatic
Driving special vehicle decision rule strategy is those skilled in the art's urgent problem to be solved.
Summary of the invention
It is extraordinary that it is an object of the present invention to overcome the above-mentioned drawbacks of the prior art and provide a kind of automatic Pilots
The track decision planing method of the decision rule method automatic Pilot job that requires special skills vehicle of working truck.
The purpose of the present invention can be achieved through the following technical solutions:
A kind of decision rule method of automatic Pilot job that requires special skills vehicle, comprising the following steps:
1) the automatic Pilot operation module of job that requires special skills vehicle is obtained from the current positioning pose of vehicle by GPS/IMU, packet
Include longitude, latitude, course and current positioning states；
2) environmental information that sensory perceptual system is sent is projected into grating map centered on from vehicle, and got the bid in grating map
Note static state and dynamic barrier build environment map；
3) according to the current positioning pose from vehicle, automatic Pilot operation module is by being stored in appointing for Local or Remote transmission
Business file, the control instruction for obtaining current work actuator are issued to kinetic control system；
4) automatic Pilot operation module obtains the task reference path for being stored in Local or Remote transmission, and task is referred to
Path projects to environmental map, is constrained using path-resolution of velocity method for planning track combination dynamics of vehicle and carries out track
Cluster planning obtains the executable basic track cluster of vehicle, and basic track cluster and task reference path are merged to obtain executable rail
Mark cluster；
The rail that the process of Trace Formation can be regarded as the linking of path segment and curvature smoothing process plans avoidance
Mark, path segment is successively connected by task reference locus, avoidance track, task reference path translation track, in connecting points
The continuous of curvature is realized with B-spline curve, and for simultaneously road/return planning track, path segment is successively referred to by task
Trajectory-offset track, simultaneously road/return track, task reference locus mark are connected, and are realized in connecting points with B-spline curve
Curvature it is continuous, what basic track cluster was completed is avoidance track or simultaneously road/return track planning；
5) uncertainty for considering environment sensing, to the executable track cluster progress safety of planning and selecting for high efficiency
It is excellent, ultimately generate the high yield track of the executable high-efficient homework of a vehicle and low risk of collision；
6) according to high yield track and the positioning pose current from vehicle, heading angle deviation and lateral deviation are obtained, and is issued
Real-time route control is carried out to kinetic control system.
The automatic Pilot operation module includes three functional areas, respectively decision rule, Operation control and driving control
System.
In path-resolution of velocity method for planning track, trajectory planning is decoupled as path planning and speed planning, it can
Advanced row speed planning, then carry out path planning, can also advanced row path planning, then carry out speed planning.
In the step 4), it includes vehicle minimum turning radius that dynamics of vehicle, which constrains, be most worth speed, most value longitudinal direction adds
Speed is most worth side acceleration and coefficient of road adhesion constraint.
In the step 4), the starting point of the executable basic track cluster of vehicle is the current pose point or vehicle of vehicle
Currently a bit in execution track safe distance；
The terminal of the executable basic track cluster of vehicle has target skewed popularity, it may be assumed that
It is avoiding barrier, in the travelable region where task reference path when vehicle is in task reference path
It is interior, inconsistent discrete of transverse and longitudinal granularity is carried out to task reference path and obtains target point set, when vehicle is in avoidance execution track
When upper, operation was carried out to return to task reference path, the travelable region where avoidance execution track and task reference path
Interior, inconsistent to avoidance execution track progress transverse and longitudinal granularity is discrete and consistent to the longitudinal granularity of task reference path progress
Discrete obtain target point set.
In the step 4), executable track cluster is made of the path segment of four kinds of modes, including task reference locus,
Avoidance track, task reference locus translate track and simultaneously road/return track.
In the step 5), consider that the uncertainty of environment sensing includes FOV, resolution ratio and the measurement of sensory perceptual system
Precision leads to the systematic error after measurement error and rasterizing.
In the step 5), track is preferentially presently in mode according to vehicle, is divided into avoidance decision and road/return
Decision and simultaneously road/return tracking Three models.
Inconsistent to task reference path progress transverse and longitudinal granularity is discrete, specifically:
Laterally more deviate reference path, discrete target point is more intensive, longitudinal target point remoter, discrete away from planning starting point
It is more sparse；
To task reference path carry out longitudinal granularity it is consistent it is discrete obtain target point set, specifically:
Along the longitudinal direction of reference path, equidistant discrete multiple target points out, all target points are made as automatic Pilot special type
Industry vehicle and road point.
The avoidance decision-making mode is divided into according to the distance from vehicle with respect to the barrier in task reference path by limited
State machine realizes 6 subpatterns of function switch, including automatic tracking subpattern, planning and adjusting subpattern, optimizing decision submodule
Formula, track execute subpattern, emergency braking subpattern and urgent avoidance and turn to subpattern；
Described and road/return decision-making mode is divided into according to and road/reentry point distance relatively optimal from vehicle by limited shape
State machine realizes 3 subpatterns of function switch, including the subpattern of avoidance tracking and road/return decision subpattern, emergency braking
Mode；
Described and road/return tracking mode is divided into respect to the distance of barrier and locating path segment by having according to from vehicle
Limit 2 modes that state machine realizes function switch, including tracking subpattern and emergency braking subpattern.
The avoidance decision-making mode specifically:
When being in automatic tracking subpattern, shows that barrier is very remote or there is no barrier, vehicle continues to hold at this time
Operation track before the trade；When being in planning and adjusting mode, show to need to adjust there are closer barrier on current work track
With planning module, the basic path cluster of avoidance is planned, after the completion of planning, the cross of current work track is deviateed with path cluster terminal
It is index to distance, carries out descending sequence, then ascending successively to carry out collision detection with barrier, first is not sent out
The path of raw collision is as path candidate, if there is collision, select terminal it is laterally most from operation track that as candidate
Path candidate is stored in sequence Q1 by path；When being in optimizing decision subpattern, candidate road all in sequence Q1 is transferred
Diameter, if there is path candidate, selects terminal in path candidate horizontal from operation track if execution track does not change without path candidate
To the farthest candidate execution route of that conduct, and carry out speed planning；When being in track execution subpattern, by candidate rail
Mark is transferred, and is executed；When being in emergency braking subpattern, show promptly to be made at this time without the execution track of effective avoidance
It is dynamic, it to avoid collision；When being in emergency braking subpattern, shows that emergency braking is not still avoided that and collide, it can only
By killing steering wheel, bring loss is avoided collision as far as possible.
Described and road/return decision-making mode specifically:
When being in avoidance tracking subpattern, show currently can't backtracking track, need first translating sections operation
On track to current execution track, so that there are also subsequent reference tracks can be performed for vehicle, and backtracking track is constantly planned
Path cluster, sort from the near to the distant with a distance from vehicle by the switching point of backtracking track, successively carry out collision detection, first
The path that item does not collide is stored in sequence Q2 as path candidate；When being in simultaneously road/return decision subpattern, sequence is transferred
All path candidates in Q2 select that paths that switching point is nearest in path candidate as path candidate, scanning frequency of going forward side by side
Metric is drawn, using the track as simultaneously road/return decision track；When being in emergency braking subpattern, show to execute without effective
Track carries out emergency braking at this time, to avoid collision.
Described and road/return tracking mode specifically:
When being in tracking subpattern, simultaneously road/return decision-making mode decision comes out and road/return decision track is executed；
When being in emergency braking subpattern, shows to carry out emergency braking at this time without effective execution track, to avoid collision.
Compared with prior art, the present invention has the advantage that
One, operating personnel only needs to open automatic Pilot work pattern switch, system near operating area in the present invention
The information such as job task and work route can be transferred automatically, and hereafter planning acts, participate in judgement without the mankind；
Two, the decision rule module in the present invention is obtaining location information, environmental information, mission mode and task reference arm
Job that requires special skills can be completely automatically carried out after diameter, by operation process and driving conditions sequencing and automation, improve extraordinary make
Industry efficiency reduces the work load of operation.
Three, the present invention provides a kind of target skewed popularity planing methods, mainly include that transverse and longitudinal granularity is inconsistent and longitudinal
The consistent basic track cluster planing method of granularity, and the planning strategy of fusion task reference path and basic track cluster, are improved
The success rate of avoidance, while the possibility for quickly returning to task reference path is increased, promote operating efficiency.
Four, the present invention uses the track decision strategy of multi-mode, keeps away by the continuous detection to optimal trajectory, and promptly
Barrier and the mode of emergency turn access, and track may be implemented and be compatible with to the uncertainty of perception environment, realize automatic Pilot
Safety.
Detailed description of the invention
Fig. 1 is flow chart of the method for the present invention.
The discrete schematic diagram of target point Fig. 2 inconsistent for the starting point and transverse and longitudinal granularity of avoidance planning of the invention.
Fig. 3 be the planning of of the invention and road/return starting point and transverse and longitudinal granularity it is inconsistent with the consistent mesh of longitudinal granularity
The discrete schematic diagram of punctuate.
Fig. 4 is trajectory set of the invention into fragmentary views.
Fig. 5 is that the finite state machine of track decision of the invention switches schematic diagram.
Fig. 6 is six sub- pattern diagrams that avoidance decision of the invention is segmented.
Fig. 7 is three sub- pattern diagrams of of the invention and road/return decision subdivision.
Fig. 8 is two sub- pattern diagrams of of the invention and road/return tracking subdivision.
Specific embodiment
The present invention is described in detail with specific embodiment below in conjunction with the accompanying drawings.Note that the following embodiments and the accompanying drawings is said
Bright is substantial illustration, and the present invention is not intended to be applicable in it object or its purposes is defined, and the present invention does not limit
In the following embodiments and the accompanying drawings.
Embodiment
As shown in Figure 1, the present invention provides a kind of decision rule method of automatic Pilot job that requires special skills vehicle, specific steps packet
It includes:
Step 1: being controlled by driver to automatic Pilot operation area, open automatic Pilot work pattern, automatic Pilot
Operation module adapter tube full-vehicle control, including Operation control and thermoacoustic prime engine；
Step 2: automatic Pilot operation module determines the current pose from vehicle by GPS/IMU, mainly includes longitude, latitude
Degree, course and current positioning states；
Step 3: centered on from vehicle, the environmental information that sensory perceptual system is sent being projected in a grating map, nothing
The grid of barrier is uniformly labeled as 0；For there is the grid of barrier, can be labeled as according to velocity information static or dynamic
State；For the type of barrier, equally cement road edge, shrub etc. can be labeled as according to the recognition result of perception.
Step 4: according to the current positioning pose from vehicle, automatic Pilot operation module calling, which is stored in, locally or remotely to be sent out
The assignment file sent, decision go out the control instruction of current work actuator, and issue；
Step 5: according to the current positioning pose from vehicle, automatic Pilot operation module calling, which is stored in, locally or remotely to be sent out
The task reference path sent；
Step 6: the task reference path that step 5 obtains being projected into the environmental map obtained in step 3, and carries out track
Cluster planning is constrained using path-resolution of velocity strategy in conjunction with dynamics of vehicle, and the executable basic track of vehicle is cooked up
Then cluster merges basic track cluster and task reference path to obtain executable track cluster, specific implementation step:
Step 61: using path-resolution of velocity trajectory planning strategy, trajectory planning being decoupled as path planning and speed
Plan two parts.
Step 62: being planning starting point with vehicle current point as shown in Figures 2 and 3；
Step 63: as shown in Fig. 2, when vehicle is in task reference path, in can travel where task reference path
In region, inconsistent discrete of transverse and longitudinal granularity is carried out to task reference path and obtains terminal target point set；As shown in figure 3, working as
Vehicle is when in avoidance execution track, in the travelable region where avoidance execution track and task reference path, to avoidance
Execution track carries out inconsistent discrete of transverse and longitudinal granularity and carries out that longitudinal granularity is consistent discrete to be obtained to task reference path
Terminal target point set；
Step 64: being constrained in conjunction with dynamics of vehicle, including vehicle minimum turning radius, most value speed, the longitudinal acceleration of most value
Degree is most worth side acceleration, coefficient of road adhesion etc. constraint relevant to dynamics of vehicle, calculates the curvature of outbound path about
Beam cooks up N number of discreet paths point by optimal method, and smooth by cubic B-spline；
Step 65: task reference path is cut, remainder is laterally displaced to the basic path target point of planning,
And combination dynamics of vehicle constraint, including vehicle minimum turning radius, most value speed, most value longitudinal acceleration, most value laterally add
Speed, coefficient of road adhesion etc. constraint relevant to dynamics of vehicle, calculate the curvature limitation of outbound path, by optimization side
Method cooks up the discrete velocity point of corresponding number, and smooth by cubic B-spline；
Step 6: as shown in figure 4, the track cluster of fusion task reference path and basic track carries out segment division, from
The path segment of task reference path is divided into task reference locus (ID1), and the path segment from avoidance planning is divided into
Avoidance track (ID2), the track from task reference path translation are divided into task reference locus translation track (ID3), come
Simultaneously road/return track (ID4) is divided into from Yu Bingdao/return trajectory planning path segment；
Step 7: from the track cluster that step 6 is cooked up, consider the uncertainty of environment sensing, constantly to planning can
Execution track cluster carries out safety and high efficiency preferentially, and final decision exports the executable high-efficient homework of a vehicle and low touches
Hit the high yield track of risk.As shown in figure 5, track decision quartile avoidance decision and road/return decision and road/return tracking
Three bulks, wherein as shown in fig. 6, avoidance decision be successively automatic tracking mode (S1), planning and adjusting mode (S2), it is optimal certainly
Plan mode (S3), track execution pattern (S4), emergency brake modes (S5) and urgent avoidance steering pattern (S6)；As shown in fig. 7,
And road/return decision is successively avoidance tracking mode (B1) and road/return decision-making mode (B2), emergency brake modes (B3)；Such as
Shown in Fig. 8, and road/return tracking is successively tracking (C1), emergency braking (C2), specific implementation step:
Step 71: along task reference path or current execution track, with vehicle width plus certain lateral safety margin, forward
Barrier is searched for, if barrier is not present, executes the tracking of task reference path or current execution track, if it exists barrier,
According to mode is presently in, step 72, step 73 or step 74 are jumped to respectively.
Step 72: as shown in the avoidance decision-making state machine of Fig. 5, if avoidance decision, specific steps are as follows:
Step 721: as shown in figure 5, defaulting since the area automatic tracking mode (S1), if clear or obstacle distance
The section S1 is fallen in, then executes the tracking of task reference path or current execution track；If there is barrier, distance falls in the section S2
(jumping condition E1) then jumps to the area planning and adjusting mode (S2)；If there is barrier, distance falls in the area S5 (jumping condition E5),
Then jump to the area emergency brake modes (S5)；If there is barrier, distance falls in the area S6 (jumping condition E6), then jumps to and promptly keep away
Hinder the area steering pattern (S6).
Step 722: as shown in figure 5, executing step if vehicle falls in the area planning and adjusting mode (S2) at a distance from barrier
The planning of track cluster described in rapid 6, and the tracking of task reference path or current execution track is continued to execute, it is completed in each planning
Afterwards, collision detection is carried out one by one from the distant to the near according to the lateral distance of target point, if without track, optimum point Best_ is touched
Point_queue queue is denoted as in vain, and if it exists without track is touched, then updating Best_Point_queue queue is near obstacle
That track serial number of object；If barrier disappears or the distance of barrier falls in the area S1 (jumping condition E2), executes and freely follow
Mark mode (S1)；If the distance of barrier falls in the area S3 (jumping condition E3), the area optimizing decision mode (S3) is jumped to；If having
Barrier, distance fall in the area S5 (jumping condition E5), then jump to the area emergency brake modes (S5)；If there is barrier, distance is fallen
The area S6 (jumping condition E6), then urgent avoidance steering pattern (S6) area is jumped to.
Step 723: as shown in figure 5, if vehicle falls in the area optimizing decision mode (S3) at a distance from barrier, by step
All available points of 722 queue Best_Point_queue take out, and select to deviate that farthest point of transverse barrier as optimal
Point Best_Point；If the distance of barrier falls in the area S4 (jumping condition E4), the area track execution pattern (S4) is jumped to；If
There is barrier, distance falls in the area S5 (jumping condition E5), then jumps to the area emergency brake modes (S5)；If there is barrier, distance
The area S6 (jumping condition E6) is fallen in, then jumps to urgent avoidance steering pattern (S6) area.
Step 724: as shown in figure 5, if vehicle falls in the area track execution pattern (S4) at a distance from barrier, with step
723 Best_Point is terminal, cooks up a complete track, and in this, as execution track (jumping condition H1), is jumped
Go to simultaneously road/return decision；If there is barrier, distance falls in the area S5 (jumping condition E5), then jumps to emergency brake modes
(S5) area；If there is barrier, distance falls in the area S6 (jumping condition E6), then jumps to urgent avoidance steering pattern (S6) area.
Step 725: as shown in figure 5, falling in the area if vehicle falls in the area emergency brake modes (S5) at a distance from barrier
Show that barrier is emergent (dynamic barrier) or barrier can not evade that (barrier is excessive, covers Operation Van
The section that can be detoured), to avoid collision, it is necessary to take emergency braking；If there is barrier, distance falls in the area S6 and (jumps condition
E6), then urgent avoidance steering pattern (S6) area is jumped to；If clear or the distance of barrier fall in the section S1~S4 and (jump
Turn condition E7), then jump to the area planning and adjusting mode (S2).
Step 726: as shown in figure 5, being fallen in if vehicle falls in urgent avoidance steering pattern (S6) area at a distance from barrier
The area shows that barrier is emergent (dynamic barrier, it is most likely that be people), and emergency braking has been difficult to avoid that collision, is
Injury is reduced as far as possible, steering wheel is killed by avoidance direction, urgent avoiding obstacles；If clear or barrier away from
From the section S1~S5 (jumping condition E8) is fallen in, then the area emergency brake modes (S5) is jumped to.
Step 73: if as shown in figure 5, simultaneously road/return decision, specific steps are as follows:
Step 731: as shown in figure 5, default enters the area avoidance tracking mode (B1), the tracking of avoidance track is executed, and
Constantly planning and road/return track, updating Best_Point_queue queue is at first and road/return track serial number；?
If vehicle falls in simultaneously road/return decision-making mode area (B2) or clear at a distance from barrier and vehicle reaches the track ID3
Segment (jumps condition F1), then jumps to simultaneously road/area return decision-making mode (B2)；If vehicle falls in the area B3 at a distance from barrier
(jumping condition F3) then jumps to the area emergency brake modes (B3).
Step 732: as shown in figure 5, if vehicle fallen at a distance from barrier and road/return decision-making mode area (B2) or
Clear and vehicle arrival ID3 path segment, all available points of step 731 queue Best_Point_queue are taken out, choosing
It selects that maximum point of serial number and cooks up a complete track as optimum point Best_Point, and in this, as execution rail
Mark (jumps condition H2), jumps to simultaneously road/return tracking；If clear and vehicle reach ID3 path segment and (jump condition
F2), then the area avoidance tracking mode (B1) is jumped to；If vehicle falls in the area B3 (jumping condition F3) at a distance from barrier, jump
Go to the area emergency brake modes (B3).
Step 733: as shown in fig. 7, falling in the area if vehicle falls in the area emergency brake modes (B3) at a distance from barrier
Show that barrier is emergent (dynamic barrier) or barrier can not evade that (barrier is excessive, covers Operation Van
The section that can be detoured), to avoid collision, it is necessary to take emergency braking；If clear or obstacle distance fall in B1~B2
Area (jumps condition F4), then jumps to the area avoidance tracking mode (B1).
Step 74: if as shown in figure 5, simultaneously road/return tracking, specific steps are as follows:
Step 731: default enters the area tracking (C1) as shown in Figure 5, simultaneously road/return track tracking is executed, as vehicle is worked as
Front position then jumps to avoidance decision ID1 path segment (jumping condition H3)；If having barrier and distance falling in the area C2 and (jumps
Turn condition G1), then jump to the area emergency braking (C2).
Step 732: as shown in figure 5, falling in the area if vehicle falls in the area emergency braking (C2) at a distance from barrier and showing
Barrier is emergent (dynamic barrier) or barrier can not be evaded, and (barrier is excessive, and covering Operation Van can be around
Capable section), to avoid collision, it is necessary to take emergency braking；If clear or obstacle distance fall in the area C1 and (jump condition
G2), then the area tracking (C1) is jumped to.
Step 8: according to execution track and the pose current from vehicle, calculating heading angle deviation and lateral deviation, and issue
To kinetic control system.
The present invention successively executes the location information for being obtained from vehicle；The environmental information being obtained from around vehicle；According to determining from vehicle
Position information determines job task mode；Task reference path is obtained according to from the location information of vehicle；It is true according to job task mode
Determine Work implement control instruction；In conjunction with task reference path and from the environmental information around vehicle, trajectory planning is carried out, obtaining can
Execution track cluster；Track decision goes out the execution track for meeting safety and high efficiency from decision in executable track cluster；
According to execution track and the pose current from vehicle, course angle and lateral deviation are calculated, and be handed down to kinetic control system.Its
In, trajectory planning module uses path-resolution of velocity strategy, constrains in conjunction with dynamics of vehicle, cooks up what vehicle can be performed
Then basic track cluster merges basic track cluster and task reference path to obtain executable track cluster；Track decision module is examined
Planning is considered to the compatibility of environmental uncertainty, by constantly carrying out safety and high efficiency to the executable track cluster of planning
Preferentially, final decision exports the high yield track of the executable high-efficient homework of a vehicle and low risk of collision.
Above embodiment is only to enumerate, and does not indicate limiting the scope of the invention.These embodiments can also be with other
Various modes are implemented, and can make in the range of not departing from technical thought of the invention it is various omit, displacement, change.
Claims (10)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811183181.0A CN109557912B (en) | 2018-10-11 | 2018-10-11 | A decision-making planning method for autonomous driving special operation vehicles |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811183181.0A CN109557912B (en) | 2018-10-11 | 2018-10-11 | A decision-making planning method for autonomous driving special operation vehicles |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109557912A true CN109557912A (en) | 2019-04-02 |
| CN109557912B CN109557912B (en) | 2020-07-28 |
Family
ID=65864885
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811183181.0A Active CN109557912B (en) | 2018-10-11 | 2018-10-11 | A decision-making planning method for autonomous driving special operation vehicles |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109557912B (en) |
Cited By (28)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109976387A (en) * | 2019-05-07 | 2019-07-05 | 江苏盛海智能科技有限公司 | The method and terminal of a kind of unmanned vehicle detection track |
| CN110196592A (en) * | 2019-04-26 | 2019-09-03 | 纵目科技（上海）股份有限公司 | A kind of update method of trajectory line, system, terminal and storage medium |
| CN110502004A (en) * | 2019-07-01 | 2019-11-26 | 江苏大学 | A modeling method of driving area importance weight distribution for intelligent vehicle lidar data processing |
| CN110568841A (en) * | 2019-08-05 | 2019-12-13 | 西藏宁算科技集团有限公司 | Automatic driving decision method and system |
| CN111142532A (en) * | 2020-01-03 | 2020-05-12 | 三一汽车制造有限公司 | Unmanned road roller group obstacle avoidance method and system based on 5G network communication |
| CN111473795A (en) * | 2020-04-08 | 2020-07-31 | 江苏盛海智能科技有限公司 | Obstacle avoidance method and terminal for unmanned vehicle |
| CN111679678A (en) * | 2020-06-30 | 2020-09-18 | 安徽海博智能科技有限责任公司 | Track planning method and system for transverse and longitudinal separation and computer equipment |
| CN111766870A (en) * | 2020-05-29 | 2020-10-13 | 广州极飞科技有限公司 | Transition path and operation path planning method and related device |
| CN112232558A (en) * | 2020-10-09 | 2021-01-15 | 中国铁建重工集团股份有限公司 | TBM spraying and mixing system operation path planning method |
| CN112578788A (en) * | 2019-09-30 | 2021-03-30 | 北京百度网讯科技有限公司 | Vehicle obstacle avoidance quadratic programming method, device, equipment and readable storage medium |
| CN112721914A (en) * | 2020-12-23 | 2021-04-30 | 同济大学 | Intelligent electric vehicle drifting and warehousing sectional type control method with supervision mechanism |
| CN112829765A (en) * | 2021-01-23 | 2021-05-25 | 湖南喜宝达信息科技有限公司 | Electric bicycle control method and system with unmanned function and storage medium |
| CN112859829A (en) * | 2019-11-28 | 2021-05-28 | 北京百度网讯科技有限公司 | Vehicle control method and device, electronic equipment and medium |
| CN113050633A (en) * | 2021-03-12 | 2021-06-29 | 苏州挚途科技有限公司 | Method and device for determining cleaning track and automatic cleaning equipment |
| CN113135195A (en) * | 2020-01-17 | 2021-07-20 | 现代自动车株式会社 | Turn-around path generation device and method for autonomous vehicle |
| CN113204242A (en) * | 2021-07-05 | 2021-08-03 | 北京理工大学 | Reconfigurable unmanned vehicle three-section type butt joint control method |
| CN113238564A (en) * | 2021-06-07 | 2021-08-10 | 江苏理工学院 | Trajectory planning method and equipment for pure electric unmanned mining dump truck |
| CN113465590A (en) * | 2021-06-29 | 2021-10-01 | 三一专用汽车有限责任公司 | Path planning method and device, automatic driving method and device and operation machine |
| CN113963027A (en) * | 2021-10-28 | 2022-01-21 | 广州文远知行科技有限公司 | Uncertainty detection model training method and device, and uncertainty detection method and device |
| WO2022021712A1 (en) * | 2020-07-28 | 2022-02-03 | Huawei Technologies Co., Ltd. | Predictive motion planning system and method |
| CN114056351A (en) * | 2021-11-26 | 2022-02-18 | 文远苏行(江苏)科技有限公司 | Automatic driving method and device |
| CN114115283A (en) * | 2021-12-02 | 2022-03-01 | 深圳市证通电子股份有限公司 | Robot Path Planning Method and System Based on Improved Opposite Parallel Algorithm |
| CN114200939A (en) * | 2021-12-10 | 2022-03-18 | 江苏集萃智能制造技术研究所有限公司 | Robot anti-collision path planning method |
| WO2023103692A1 (en) * | 2021-12-07 | 2023-06-15 | 阿里巴巴达摩院(杭州)科技有限公司 | Decision planning method for autonomous driving, electronic device, and computer storage medium |
| CN116381748A (en) * | 2023-04-06 | 2023-07-04 | 黑龙江惠达科技发展有限公司 | Agricultural machinery straight line automatic driving track comparison method and device |
| CN117622227A (en) * | 2024-01-24 | 2024-03-01 | 广汽埃安新能源汽车股份有限公司 | Vehicle lane changing obstacle avoidance control method and device |
| CN117698769A (en) * | 2024-02-05 | 2024-03-15 | 上海鉴智其迹科技有限公司 | Automatic driving track planning method and device, electronic equipment and storage medium |
| DE102024104674B3 (en) | 2024-02-20 | 2025-03-27 | Audi Aktiengesellschaft | Method for providing a collision-free rescue trajectory for at least one emergency vehicle, as well as control device and emergency vehicle |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107168305A (en) * | 2017-04-01 | 2017-09-15 | 西安交通大学 | Unmanned vehicle method for planning track based on Bezier and VFH under the scene of crossing |
| CN107340772A (en) * | 2017-07-11 | 2017-11-10 | 清华大学 | It is a kind of towards the unpiloted reference locus planing method that personalizes |
| CN107817509A (en) * | 2017-09-07 | 2018-03-20 | 上海电力学院 | Crusing robot navigation system and method based on the RTK Big Dippeves and laser radar |
| CN107992050A (en) * | 2017-12-20 | 2018-05-04 | 广州汽车集团股份有限公司 | Pilotless automobile local path motion planning method and device |
| CN108088456A (en) * | 2017-12-21 | 2018-05-29 | 北京工业大学 | A kind of automatic driving vehicle local paths planning method with time consistency |
| CN108501911A (en) * | 2018-03-09 | 2018-09-07 | 同济大学 | A kind of vehicle integrated type electrical brake fluid system and stability control method |
| US20180284791A1 (en) * | 2017-03-31 | 2018-10-04 | Panasonic Intellectual Property Management Co., Ltd. | Automatic driving control method, automatic driving control device using the same, and non-transitory storage medium |
-
2018
- 2018-10-11 CN CN201811183181.0A patent/CN109557912B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180284791A1 (en) * | 2017-03-31 | 2018-10-04 | Panasonic Intellectual Property Management Co., Ltd. | Automatic driving control method, automatic driving control device using the same, and non-transitory storage medium |
| CN107168305A (en) * | 2017-04-01 | 2017-09-15 | 西安交通大学 | Unmanned vehicle method for planning track based on Bezier and VFH under the scene of crossing |
| CN107340772A (en) * | 2017-07-11 | 2017-11-10 | 清华大学 | It is a kind of towards the unpiloted reference locus planing method that personalizes |
| CN107817509A (en) * | 2017-09-07 | 2018-03-20 | 上海电力学院 | Crusing robot navigation system and method based on the RTK Big Dippeves and laser radar |
| CN107992050A (en) * | 2017-12-20 | 2018-05-04 | 广州汽车集团股份有限公司 | Pilotless automobile local path motion planning method and device |
| CN108088456A (en) * | 2017-12-21 | 2018-05-29 | 北京工业大学 | A kind of automatic driving vehicle local paths planning method with time consistency |
| CN108501911A (en) * | 2018-03-09 | 2018-09-07 | 同济大学 | A kind of vehicle integrated type electrical brake fluid system and stability control method |
Non-Patent Citations (2)
| Title |
|---|
| YU, ZHUOPING等: "Design of Unmanned Park Vehicle Decision-making System Based on Uncertainty Information", 《 5TH INTERNATIONAL-FEDERATION-OF-AUTOMATIC-CONTROL (IFAC) CONFERENCE ON ENGINE AND POWERTRAIN CONTROL, SIMULATION AND MODELING (E-COSM)》 * |
| 梁广民: "一种新的自动驾驶轨迹规划方法", 《电子科技大学学报》 * |
Cited By (42)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110196592A (en) * | 2019-04-26 | 2019-09-03 | 纵目科技（上海）股份有限公司 | A kind of update method of trajectory line, system, terminal and storage medium |
| CN110196592B (en) * | 2019-04-26 | 2023-07-28 | 纵目科技（上海）股份有限公司 | Track line updating method, system, terminal and storage medium |
| CN109976387B (en) * | 2019-05-07 | 2022-05-27 | 福建盛海智能科技有限公司 | Method and terminal for detecting track by unmanned vehicle |
| CN109976387A (en) * | 2019-05-07 | 2019-07-05 | 江苏盛海智能科技有限公司 | The method and terminal of a kind of unmanned vehicle detection track |
| CN110502004A (en) * | 2019-07-01 | 2019-11-26 | 江苏大学 | A modeling method of driving area importance weight distribution for intelligent vehicle lidar data processing |
| CN110568841A (en) * | 2019-08-05 | 2019-12-13 | 西藏宁算科技集团有限公司 | Automatic driving decision method and system |
| CN112578788A (en) * | 2019-09-30 | 2021-03-30 | 北京百度网讯科技有限公司 | Vehicle obstacle avoidance quadratic programming method, device, equipment and readable storage medium |
| CN112578788B (en) * | 2019-09-30 | 2023-05-02 | 北京百度网讯科技有限公司 | Vehicle obstacle avoidance secondary planning method, device, equipment and readable storage medium |
| CN112859829A (en) * | 2019-11-28 | 2021-05-28 | 北京百度网讯科技有限公司 | Vehicle control method and device, electronic equipment and medium |
| CN111142532A (en) * | 2020-01-03 | 2020-05-12 | 三一汽车制造有限公司 | Unmanned road roller group obstacle avoidance method and system based on 5G network communication |
| CN113135195A (en) * | 2020-01-17 | 2021-07-20 | 现代自动车株式会社 | Turn-around path generation device and method for autonomous vehicle |
| CN111473795A (en) * | 2020-04-08 | 2020-07-31 | 江苏盛海智能科技有限公司 | Obstacle avoidance method and terminal for unmanned vehicle |
| CN111766870A (en) * | 2020-05-29 | 2020-10-13 | 广州极飞科技有限公司 | Transition path and operation path planning method and related device |
| CN111679678A (en) * | 2020-06-30 | 2020-09-18 | 安徽海博智能科技有限责任公司 | Track planning method and system for transverse and longitudinal separation and computer equipment |
| US11835958B2 (en) | 2020-07-28 | 2023-12-05 | Huawei Technologies Co., Ltd. | Predictive motion planning system and method |
| WO2022021712A1 (en) * | 2020-07-28 | 2022-02-03 | Huawei Technologies Co., Ltd. | Predictive motion planning system and method |
| CN112232558A (en) * | 2020-10-09 | 2021-01-15 | 中国铁建重工集团股份有限公司 | TBM spraying and mixing system operation path planning method |
| CN112232558B (en) * | 2020-10-09 | 2024-02-20 | 中国铁建重工集团股份有限公司 | TBM (Tunnel boring machine) spraying and mixing system operation path planning method |
| CN112721914A (en) * | 2020-12-23 | 2021-04-30 | 同济大学 | Intelligent electric vehicle drifting and warehousing sectional type control method with supervision mechanism |
| CN112829765A (en) * | 2021-01-23 | 2021-05-25 | 湖南喜宝达信息科技有限公司 | Electric bicycle control method and system with unmanned function and storage medium |
| CN113050633A (en) * | 2021-03-12 | 2021-06-29 | 苏州挚途科技有限公司 | Method and device for determining cleaning track and automatic cleaning equipment |
| CN113050633B (en) * | 2021-03-12 | 2024-05-31 | 苏州挚途科技有限公司 | Method and device for determining cleaning track and automatic cleaning equipment |
| CN113238564B (en) * | 2021-06-07 | 2025-04-25 | 江苏理工学院 | A trajectory planning method and device for a pure electric unmanned mining dump truck |
| CN113238564A (en) * | 2021-06-07 | 2021-08-10 | 江苏理工学院 | Trajectory planning method and equipment for pure electric unmanned mining dump truck |
| CN113465590A (en) * | 2021-06-29 | 2021-10-01 | 三一专用汽车有限责任公司 | Path planning method and device, automatic driving method and device and operation machine |
| CN113465590B (en) * | 2021-06-29 | 2024-03-15 | 三一专用汽车有限责任公司 | Path planning method and device, automatic driving method and device and working machine |
| CN113204242B (en) * | 2021-07-05 | 2021-09-07 | 北京理工大学 | Three-stage docking control method for reconfigurable unmanned vehicles |
| CN113204242A (en) * | 2021-07-05 | 2021-08-03 | 北京理工大学 | Reconfigurable unmanned vehicle three-section type butt joint control method |
| CN113963027A (en) * | 2021-10-28 | 2022-01-21 | 广州文远知行科技有限公司 | Uncertainty detection model training method and device, and uncertainty detection method and device |
| CN114056351A (en) * | 2021-11-26 | 2022-02-18 | 文远苏行(江苏)科技有限公司 | Automatic driving method and device |
| CN114115283B (en) * | 2021-12-02 | 2023-09-19 | 深圳市证通电子股份有限公司 | Robot path planning method and system based on improved oppositional parallel algorithm |
| CN114115283A (en) * | 2021-12-02 | 2022-03-01 | 深圳市证通电子股份有限公司 | Robot Path Planning Method and System Based on Improved Opposite Parallel Algorithm |
| WO2023103692A1 (en) * | 2021-12-07 | 2023-06-15 | 阿里巴巴达摩院(杭州)科技有限公司 | Decision planning method for autonomous driving, electronic device, and computer storage medium |
| CN114200939B (en) * | 2021-12-10 | 2024-02-27 | 江苏集萃智能制造技术研究所有限公司 | Robot anti-collision path planning method |
| CN114200939A (en) * | 2021-12-10 | 2022-03-18 | 江苏集萃智能制造技术研究所有限公司 | Robot anti-collision path planning method |
| CN116381748B (en) * | 2023-04-06 | 2023-11-21 | 黑龙江惠达科技股份有限公司 | Agricultural machinery straight line automatic driving track comparison method and device |
| CN116381748A (en) * | 2023-04-06 | 2023-07-04 | 黑龙江惠达科技发展有限公司 | Agricultural machinery straight line automatic driving track comparison method and device |
| CN117622227A (en) * | 2024-01-24 | 2024-03-01 | 广汽埃安新能源汽车股份有限公司 | Vehicle lane changing obstacle avoidance control method and device |
| CN117622227B (en) * | 2024-01-24 | 2024-04-16 | 广汽埃安新能源汽车股份有限公司 | Vehicle lane changing obstacle avoidance control method and device |
| CN117698769A (en) * | 2024-02-05 | 2024-03-15 | 上海鉴智其迹科技有限公司 | Automatic driving track planning method and device, electronic equipment and storage medium |
| CN117698769B (en) * | 2024-02-05 | 2024-04-26 | 上海鉴智其迹科技有限公司 | Automatic driving track planning method and device, electronic equipment and storage medium |
| DE102024104674B3 (en) | 2024-02-20 | 2025-03-27 | Audi Aktiengesellschaft | Method for providing a collision-free rescue trajectory for at least one emergency vehicle, as well as control device and emergency vehicle |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109557912B (en) | 2020-07-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109557912A (en) | A kind of decision rule method of automatic Pilot job that requires special skills vehicle | |
| CN112882469B (en) | Deep reinforcement learning obstacle avoidance navigation method integrating global training | |
| CN110268457B (en) | Method for determining a collective maneuver of at least two vehicles, computer program product, computer-readable medium, controller and vehicle comprising the controller | |
| WO2018176593A1 (en) | Local obstacle avoidance path planning method for unmanned bicycle | |
| CN112284393B (en) | Global path planning method and system for intelligent mobile robot | |
| CN108225364A (en) | A kind of pilotless automobile driving task decision system and method | |
| WO2019042295A1 (en) | Path planning method, system, and device for autonomous driving | |
| CN113916246A (en) | A method and system for unmanned obstacle avoidance path planning | |
| CN110260867A (en) | Method, equipment and the device that pose is determining in a kind of robot navigation, corrects | |
| EP4320020B1 (en) | Lane changing based only on local information | |
| CN113670305B (en) | Parking trajectory generation method, device, computer equipment and storage medium | |
| Chen et al. | Foad: Fast optimization-based autonomous driving motion planner | |
| WO2022040748A1 (en) | Method and apparatus for coordinating multiple cooperative vehicle trajectories on shared road networks | |
| CN112622924B (en) | Driving planning method and device and vehicle | |
| JP5666327B2 (en) | Unmanned moving body and control method of unmanned moving body | |
| CN111006667A (en) | Automatic driving track generation system under high-speed scene | |
| CN105752154A (en) | Vehicle steering control system and method | |
| Hosseini et al. | Interactive path planning for teleoperated road vehicles in urban environments | |
| CN117885764B (en) | Vehicle track planning method and device, vehicle and storage medium | |
| CN110254422A (en) | A car obstacle avoidance method based on multi-objective reinforcement learning and Bezier curve | |
| Kala et al. | Planning autonomous vehicles in the absence of speed lanes using an elastic strip | |
| WO2023039707A1 (en) | Motion planning using spatio-temporal convex corridors | |
| CN116300971B (en) | Traction sliding control method and device for civil aircraft, tractor and storage medium | |
| CN113515111A (en) | Vehicle obstacle avoidance path planning method and device | |
| KR20180040787A (en) | Method and device for autonomous driving based on route adaptation |
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 |