CN102591332A - Device and method for local path planning of pilotless automobile - Google Patents

Abstract

The invention relates to a device and a method for local path planning of a pilotless automobile. The device comprises an environment sensing device, a repulsion force calculating device, a gravitation force calculating device, a resultant force direction angle calculating device and a steering wheel turning angle calculating device, wherein the environment sensing device is used for detecting obstacles and building a road boundary model and a road center line model, the repulsion force calculating device is used for building a repulsion force point function and calculating the repulsion force, the gravitation force calculating device is used for building a gravitation force point function and calculating the gravitation force, the resultant force direction angle calculating device Is used for calculating the resultant force direction angle of the repulsion force and the gravitation force, and the steering wheel turning angle calculating device is used for determining the steering wheel turning angle according to the direction angle of the resultant force and the transmission ratio of a steering system. The method has the advantages that the problems of path oscillation and minimum sunk local part because the repulsion force and the gravitation force are in the same direction in a manual potential field method are eliminated, and in addition, the running path deflection of vehicles caused by uncertain factors can be corrected in real time.

Description

The device and method that is used for the pilotless automobile local paths planning

Technical field

The invention belongs to the intelligent automobile technical field, be specifically related to be used for the device and method of unmanned local paths planning.

Background technology

Pilotless automobile system (Autonomous Ground Vehicle is called for short AGV) a kind ofly obtains environmental informations and vehicle-state, position according to various sensors; Through the understanding of environment being controlled automatically the intelligence control system of vehicle drive behavior; Mainly by sensor; Processor, device such as controller is formed.

Local paths planning is one of gordian technique of pilotless automobile research.Local paths planning is meant: pilotless automobile in uncertain road environment, information, the global path planning that control system provides according to environment sensing system and vehicle-state detection system provide the target that will reach etc. cook up the current driving path of vehicle in real time.

The Artificial Potential Field method is a comparative maturity and real-time planing method preferably in the local paths planning research; To be that environmental information with vehicle ' is abstract be gravitational field function and repulsion field function for it, through the composite force field function cook up one from starting point to gravitation the collisionless path of point (impact point).

Summary of the invention

The object of the present invention is to provide a kind of device and method that is used for the pilotless automobile local paths planning, it is need not set up under the situation of complex environment model, calculate the driving path of pilotless automobile according to environmental characteristic.

For reaching above purpose, the solution that the present invention adopted is:

A kind of device that is used for the pilotless automobile local paths planning, it comprises:

Environmental perception device is used for the detecting obstacles thing, sets up road boundary model and road-center line model;

The repulsion calculation element is used to set up the repulsion point function and calculates repulsion;

The gravitation calculation element is used to set up the gravitation point function and calculates gravitation;

Resultant direction angle calculation device is used to calculate the orientation angle of making a concerted effort of repulsion and gravitation;

The steering wheel angle calculation element is used for confirming steering wheel angle according to the orientation angle and the steering ratio of gear of making a concerted effort.

Further; Said environmental perception device is vision sensor and radar; Vision sensor is surveyed road boundary and is calculated road axis; The radar detection obstacle information, the path coordinate dot information that vision sensor is provided fits to repeatedly curve, sets up road boundary model and road-center line model.

Said environmental perception device is a radar, the radar detection curb, and match road boundary information is extrapolated road axis information, sets up road boundary model and road-center line model.

The real-time information of the environment that said repulsion calculation element provides according to environmental perception device confirms that the border, two road that vehicle will go sets up the repulsion point function; And according to the change of repulsion point function calculating along with environment, the position of automatic driving car repulsion point and calculating repulsion point are to the repulsion size of pilotless automobile.

Said gravitation calculation element through the deviation calculation device calculate with new boundary line be the road axis of road with the road axis of vehicle place road at present between lateral separation, be subordinate to composite function through Gauss simultaneously the information real-time that clear, obstacle distance pilotless automobile distance are arranged in the road be reflected on the gravitation point function; And according to the change of gravitation point function calculating along with environment, the position of automatic driving car gravitation point, and calculate the gravitation size of gravitation point to pilotless automobile.

A kind of method that is used for the pilotless automobile local paths planning, it comprises:

The environment sensing step, the detecting obstacles thing is set up road boundary model and road-center line model;

The repulsion calculation procedure is set up the repulsion point function and is calculated repulsion;

The gravitation calculation procedure is set up the gravitation point function and is calculated gravitation;

Resultant direction angle calculation step, the orientation angle of making a concerted effort of calculating repulsion and gravitation;

The steering wheel angle calculation procedure is confirmed steering wheel angle according to orientation angle of making a concerted effort and steering ratio of gear.

Further; Said environment sensing step is that vision sensor is surveyed road boundary and calculated road axis; The radar detection obstacle information, the path coordinate dot information that vision sensor is provided fits to repeatedly curve, sets up road boundary model and road-center line model.

Said environment sensing step is the radar detection curb, and match road boundary information is extrapolated road axis information, sets up road boundary model and road-center line model.

The real-time information of the environment that said repulsion calculation procedure provides according to environmental perception device confirms that the border, two road that vehicle will go sets up the repulsion point function; And according to the change of repulsion point function calculating along with environment, the position of automatic driving car repulsion point and calculating repulsion point are to the repulsion size of pilotless automobile.

Said gravitation calculation procedure through the deviation calculation device calculate with new boundary line be the road axis of road with the road axis of vehicle place road at present between lateral separation, be subordinate to composite function through Gauss simultaneously the information real-time that clear, obstacle distance pilotless automobile distance are arranged in the road be reflected on the gravitation point function; And according to the change of gravitation point function calculating along with environment, the position of automatic driving car gravitation point, and calculate the gravitation size of gravitation point to pilotless automobile.

Owing to adopted such scheme, the present invention to have following characteristics:

1), the invention solves in traditional Artificial Potential Field method because the problem that is absorbed in local minimum and path concussion that repulsion and gravitation produce when same direction through utilizing Gauss to make up the method for subordinate function real time reaction environmental change;

2) selection through the repulsion point function; When vehicle because of the interference that receives uncertain factor when for example side direction wind, road roughness etc. depart from target line and sail the path; Repulsion and gravitation with joint efforts guided vehicle is got back to the target driving path, improved the robust performance of vehicle tracking dreamboat driving path;

3) owing to need not set up the complex environment model; Only need the constructing environment characteristic model to realize local paths planning; Therefore required sensor data information amount requires real-time littler, processing data information better; Help to reduce the sensor cost on the one hand, also help satisfying the requirement of the high real-time responsiveness of vehicle control syetem on the other hand;

4) the present invention has good environmental suitability, in embodiment of the present invention, lifts two kinds of running environments, and the present invention is for other running environments, and for example the intersection is stopped up but had and can pass through path situation etc., can realize vehicle automatic obstacle-avoiding under the multiple environment;

5) make up the setting of subordinate function coefficient through Gauss; The target path function that goes also can satisfy the vehicle minimal curve radius; Vehicle kinematics and dynamic (dynamical) constraint conditions such as front-wheel steering locking angle speed; Make controlled device can be good at following the tracks of the expected path that provides, thereby realize the optimum control of expection.

Description of drawings

Fig. 1 is a local paths planning device synoptic diagram.

Fig. 2 is an environmental characteristic modelling device synoptic diagram.

Fig. 3 is a gravitation point function apparatus for establishing synoptic diagram.

Fig. 4 gets the track of vehicle comparison diagram of different value for σ.

Fig. 5 is a resultant direction angle calculation device synoptic diagram.

Fig. 6 is automatic driving car driving trace simulation result figure under the clear environment.

Fig. 7 is automatic driving car expectation steering wheel angle figure under the clear environment.

Fig. 8 is automatic driving car driving trace simulation result figure under the obstacle environment.

Fig. 9 is automatic driving car expectation steering wheel angle figure under the obstacle environment.

Embodiment

Below in conjunction with the accompanying drawing illustrated embodiment the present invention is further described.

The present invention is on the basis of Artificial Potential Field APF (Artificial Potential Field) method; Utilize Gauss to make up subordinate function and set up the target path function that goes; It is the gravitation point function of Artificial Potential Field method; The variation real-time embodying of environment in the variation of gravitation point function, and is calculated gravitation with this; Obtain the repulsion point function according to obstacle information and road side information, calculate repulsion with this.Calculate the suffered resultant direction of pilotless automobile through gravitation and repulsion again, thereby cook up automatic driving car the path of going.When vehicle when the target driving path goes; Two repulsion are cancelled each other, and gravitation plays a leading role, in case vehicle is because the interference that receives uncertain factor side direction wind for example; Road roughnesss etc. depart from target line and sail the path, repulsion and gravitation with joint efforts guided vehicle is got back to the target driving path.

A kind of device that is used for gravitation calculating among the present invention; Comprise gravitation point function apparatus for establishing and gravitation calculation element; Wherein this gravitation point function device comprises: deviation calculation, and being used to calculate with new boundary line is the road axis of road and the lateral separation between the road axis of vehicle place road at present; Be subordinate to composite function with having clear, obstacle distance pilotless automobile distance etc. to be reflected in real time on the gravitation point function in the road through Gauss; Wherein this gravitation calculation element clicks according to gravitation on the gravitation point function and fetches calculating gravitation.

Wherein, this deviation calculation device is according to the obstacles borders dot information and the road-center line computation of radar (laser or millimetre-wave radar) acquisition.

Wherein, This gravitation point function provides road axis according to vision system; Through the maximum change calculations of finding the solution the road axis deviation and the taking place value that deviates, and utilize Gauss to be subordinate to the information of composite function real-time embodying barrier according to the variation of real time environment, thereby obtain the gravitation point function.

Wherein, this Gauss coefficient of being subordinate to composite function will make the target driving path satisfy the kinematics and the dynamics constraint condition of vehicle.

Wherein, this gravitation point is chosen deciding according to the curvature of road and the speed of vehicle.

A kind ofly be used for the device that repulsion calculates, comprise repulsion point function apparatus for establishing and repulsion calculation element, wherein the real-time information of the environment that provides according to vision sensor and radar of repulsion point function confirms that the border, two road that vehicle will go sets up; The repulsion calculation element is used for calculating the repulsion size according to repulsion point.

The present invention also comprises and is converted into the steering wheel angle device with making a concerted effort; Comprise the steps: to establish bodywork reference frame; Direction big or small according to gravitation and under bodywork reference frame is decomposed into gravitation the component of orthogonal axis; According to repulsion size and the direction under bodywork reference frame repulsion is decomposed into the component of orthogonal axis, thereby calculates front wheel angle, calculate steering wheel angle according to the steering ratio of gear again.

The present invention intends and uses vision sensor detection road boundary and calculate road axis; Use radar (laser or millimetre-wave radar) detecting obstacles thing information; The path coordinate dot information that vision system is provided fits to repeatedly curve and (considers the dynamics constraint of vehicle; One be three times and with upper curve) set up road boundary model and road-center line model (also can be only with laser radar as the environment sensing system; Utilize laser radar to survey curb, match road boundary information is extrapolated road axis information).Obtain the environmental characteristic model based on sensor; Gauss in the gravitation point function is subordinate to the variation of composite function item with environment in the real-time embodying road; The distance that barrier, barrier and unmanned workshop promptly whether occur, and according to the gravitation point function calculating gravitation that obtains in real time; Set up the repulsion point function according to environment change, and calculate the repulsion size according to the repulsion point function.Local paths planning device synoptic diagram such as Fig. 1.

1, environment sensing:

According to vision system, on bodywork reference frame, set up boundary's function model on both sides of the road

$\left\{\begin{array}{c}{\mathrm{<figure-callout\; id="1"\; label="y"\; filenames="HDA0000043738410000011.png,HDA0000043738410000052.png"\; state="\{\{state\}\}">y</figure-callout>}}_1=a_1{x}^3+b_1{x}^2+c_{1}x+d_1\\ {\mathrm{<figure-callout\; id="2"\; label="y"\; filenames="HDA0000043738410000011.png"\; state="\{\{state\}\}">y</figure-callout>}}_2=a_1{x}^3+b_1{x}^2+c_{1}x+d_2\end{array}\right.---\left(1\right)$

A wherein ₁, b ₁, c ₁, d ₁, d ₂Be the cubic curve coefficient.

Release the road axis function model by formula (1)

y _centre＝a ₁x ³+b ₁x ²+c ₁x+(d ₁+d ₂)/2 (2)

When in the road barrier being arranged, the road route information data that provides through radar scanning and vision system merges, and obtains the real-time coordinate (X of the most dangerous frontier point of barrier under the same coordinate system in the current road _Ob, Y _Ob), promptly obtain the data message on two borders that can be through road.Obtain the basic model of environmental characteristic through above raw data.Environmental characteristic modelling device synoptic diagram such as Fig. 2.

2, the gravitation point function is set up

Vehicle goes on structured road; Must have the road boundary constraint; And possibly there is barrier; Exist the situation of barrier have two kinds maybe, a kind of be that barrier and road boundary are the wideest path of passing through during as the boundary line, a kind of is that two barriers are the wideest path of passing through during as the border.Gravitation point function apparatus for establishing such as Fig. 3.

1) there is not barrier in the road

At this moment, the constraint that vehicle receives the boundary line, two road can not drive to beyond the road, and therefore, boundary line, twice roadside is the repulsion point function, and the gravitation point function is the road-center line function.

2) when in the road barrier being arranged

When in the road barrier being arranged, at first select the wideest road that passes through, thereby confirm road boundary point and Z-factor.

Gravitation point function under the situation of barrier is arranged in the road; Be divided into two kinds of situation; A kind of situation barrier is as a side boundary line, and lane boundary line is as the opposite side boundary line, and promptly the repulsion point function is respectively lane boundary line and the most dangerous frontier point of barrier; Another kind of situation be two barriers as the boundary line, specify with first kind of situation at present.

The repulsion point function is the real-time coordinate points (X of lane boundary line and the dangerous point of barrier in this case _Ob, Y _Ob).

Barrier is formed and can be produced new road axis through the border in path with road, one side route, then the required maximum offset of original path center line function

Δs＝|(a ₁x _ob ³+b ₁x _ob ²+c ₁x _ob+(d ₁+d ₂)/2)-(y _ob+a ₂x _ob ³+b ₂x _ob ²+c ₂x _ob+d ₂)/2| (3)

Promptly for the gravitation point function under the edge-restraint condition of obstacles borders do when vehicle is on one side

y _goal＝a ₁x ³+b ₁x ²+c ₁x+(d ₁+d ₂)/2+Δs (4)

Can cause the discontinuous of path curvature but directly add a side-play amount, not satisfy the kinematical constraint condition of vehicle, to sum up analyze and consider vehicle constraint condition, adopt Gauss's subordinate function to come smooth excessive two objective functions, therefore revise the gravitation point function and do

y _goal＝a ₁x ³+b ₁x ²+c ₁x+(d ₁+d ₂)/2+Δs*exp(-(X-X _ob) ²/2*σ ²) (5)

Wherein σ is the relevant coefficient of curvature with Gauss's subordinate function curve, in the gravitation point function of gravitation, is path curvature bounded and path curvature inverse bounded through regulating the dynamics constraint that its value can satisfy vehicle, gets the path locus of different value like Fig. 4 σ.The situation that does not as seen from the above analysis have barrier in the road is the special case that the barrier situation is arranged in the road, is not promptly having Δ s*exp ((X-X under the situation of barrier _Ob) ²/ 2* σ ²To go to zero.

To sum up, establishing the gravitation point function is

y _goal＝a ₁x ³+b ₁x ²+c ₁x+(d ₁+d ₂)/2+Δs*exp(-(X-X _ob) ²/2*σ ²) (6)

3, the repulsion point function is set up

The repulsion point function

1) when not having barrier, it is formula (1) that the repulsion point function is road boundary

2) when in the road barrier being arranged, will introduce barrier as a side, boundary through the road boundary situation according to the wideest

4, gravitation calculates

In the method pilotless automobile (controlled device) is reduced to a particle, its space, place is two-dimentional theorem in Euclid space.The position X=[xy] of controlled device in the space ^T, in the present invention, because be under bodywork reference frame, so X=[00] ^TControlled device is in the suffered gravitational field function U of X _Att(X) be defined as and target location X _g=[x _gy _g] ^TRelevant function:

$U_{\mathrm{att}}\left(X\right)=\frac{1}{2}\mathrm{\&kappa;}{(X-X_g)}^2---\left(7\right)$

In the formula: κ is the gravitational field gain.Corresponding gravitation F _Att(X) be the negative gradient of gravitational field:

$F_{\mathrm{att}}\left(X\right)=-\&dtri;U_{\mathrm{att}}\left(X\right)=-\mathrm{\&kappa;}{\mathrm{\&rho;}}_g{a}_G---\left(8\right)$

In the formula: a _GBe the targeted vector of unit length of controlled device; ρ _g=|| X-X _g|| be the distance between controlled device and the gravitation point.

5, repulsion calculates

The repulsion field function does

$U_{\mathrm{rep}}\left(X\right)=\left\{\begin{array}{cc}\frac{1}{2}\mathrm{\&eta;}{(\frac{1}{{\mathrm{\&rho;}}_{\mathrm{ob}}}-\frac{1}{{\mathrm{\&rho;}}_0})}^2{{\mathrm{\&rho;}}_g}^n& {\mathrm{\&rho;}}_{\mathrm{ob}}\&le;{\mathrm{\&rho;}}_0\\ 0& {\mathrm{\&rho;}}_{\mathrm{ob}}>{\mathrm{\&rho;}}_0\end{array}\right.---\left(9\right)$

In the formula n be one greater than any real number of zero.

When controlled device during not at gravitation point, then corresponding repulsion is:

$F_{\mathrm{rep}}\left(X\right)=\left\{\begin{array}{cc}({\mathrm{<figure-callout\; id="1"\; label="F\; rep"\; filenames="HDA0000043738410000011.png,HDA0000043738410000052.png"\; state="\{\{state\}\}">F</figure-callout>}}_{\mathrm{rep}}+F_{\mathrm{rep}2})a_o& {\mathrm{\&rho;}}_{\mathrm{ob}}\&le;{\mathrm{\&rho;}}_0\\ 0& {\mathrm{\&rho;}}_{\mathrm{ob}}>{\mathrm{\&rho;}}_0\end{array}\right.---\left(10\right)$

Wherein:

${\mathrm{<figure-callout\; id="1"\; label="F\; rep"\; filenames="HDA0000043738410000011.png,HDA0000043738410000052.png"\; state="\{\{state\}\}">F</figure-callout>}}_{\mathrm{rep}}=\mathrm{\&eta;}(\frac{1}{{\mathrm{\&rho;}}_{\mathrm{ob}}}-\frac{1}{{\mathrm{\&rho;}}_0})\frac{{{\mathrm{\&rho;}}_g}^n}{{{\mathrm{\&rho;}}_{\mathrm{<figure-callout\; id="2"\; label="ob"\; filenames="HDA0000043738410000011.png"\; state="\{\{state\}\}">ob</figure-callout>}}}^2}$

${\mathrm{<figure-callout\; id="2"\; label="F\; rep"\; filenames="HDA0000043738410000011.png"\; state="\{\{state\}\}">F</figure-callout>}}_{\mathrm{rep}}=\frac{\mathrm{<figure-callout\; id="2"\; label="n"\; filenames="HDA0000043738410000011.png"\; state="\{\{state\}\}">n</figure-callout>}}{2}\mathrm{\&eta;}{(\frac{1}{{\mathrm{\&rho;}}_{\mathrm{ob}}}-\frac{1}{{\mathrm{\&rho;}}_0})}^2{{\mathrm{\&rho;}}_g}^{n-1}$

In the formula: η is the gain of repulsion field function, ρ _Ob=|| X-X _Ob|| be the bee-line of controlled device and barrier, constant ρ ₀The distance that influences for the barrier set according to the speed of a motor vehicle.

A wherein _OPoint to the vector of unit length of controlled device for barrier.

6, resultant direction is calculated

Resultant direction has promptly determined the direction of motion of controlled device.Under bodywork reference frame, gravitation and repulsion are decomposed into the component on two coordinate axis respectively.Resultant direction angle calculation device synoptic diagram such as Fig. 5.

On the gravitation point function that with the bodywork reference frame is coordinate axis foundation, choose gravitation point X _g=[x _g, y _g], the angle between automatic driving car and the gravitation point then

α＝arctan(y _g/x _g) (11)

The component of gravitation on horizontal stroke, ordinate does

$\left\{\begin{array}{c}{F}_{\mathrm{att}}\left(x_g\right)=F_{\mathrm{att}}*\cos \left(\mathrm{\&alpha;}\right)\\ F_{\mathrm{att}}\left(y_g\right)=F_{\mathrm{att}}*\sin \left(\mathrm{\&alpha;}\right)\end{array}\right.---\left(12\right)$

On bodywork reference frame, repulsion point X _Ob(i)=[x _Ob(i), y _Ob(i)], the angle between automatic driving car and the barrier then

β _i＝arctan(y _ob(i)/x _ob(i)) (13)

The component of repulsion on horizontal stroke, ordinate does

$\left\{\begin{array}{c}{F}_{\mathrm{rep}}\left(x_{\mathrm{ob}}\left(i\right)\right)=F_{\mathrm{rep}}\left(i\right)*\cos \left({\mathrm{\&beta;}}_i\right)\\ F_{\mathrm{rep}}\left(y_{\mathrm{ob}}\left(i\right)\right)=F_{\mathrm{rep}}\left(i\right)*\sin \left({\mathrm{\&beta;}}_i\right)\end{array}\right.---\left(14\right)$

The then pilotless automobile and the angle of making a concerted effort, the i.e. course angle of expectation

θ＝arctan((F _att(y)+F _rep(y(i)))/(F _att(x)+F _rep(x(i)))) (15)

Steering wheel angle

δ _sw＝θ*i _s (16)

I wherein _sBe the steering ratio of gear.

Simulating, verifying is carried out in above invention under varying environment, each parameter of system adopts following value:

κ=6, ρ ₀=10, η=0.7, n=2, σ=4.5, speed of a motor vehicle v=18km/h

When clear in the road or barrier (are not ρ in the coverage scope _Ob＞ρ ₀), as repulsion, as gravitation, this moment, correction term was zero to utilization improvement Artificial Potential Field method with road axis with lane boundary line.Fig. 6 is the pilotless automobile driving trace, can find out from driving trace, and pilotless automobile can be good at following road axis.Fig. 7 is a steering wheel angle, can find out from the order of the steering wheel angle that provides, expectation corner kinematics smooth and that satisfy vehicle retrain with dynamics (for example, turned to topworks to limit, being constrained to of the steering wheel angle of pilotless automobile | δ _Sw|＜=500 °, being constrained to of steering wheel angle speed | V _δ|＜=200 °/s.As can beappreciated from fig. 7, under the clear environment, when utilizing automatic driving vehicle of the present invention in having the road of certain curvature, to go, steering wheel angle can remain in 200 degree, and steering wheel angle speed also can remain in the restriction range of 200 degree/seconds).

Barrier appears in the road of the place ahead, and automatic driving car and obstacle distance (ρ in effective range _Ob≤ρ ₀), as the repulsion point function, this moment, correction term was relevant with maximum offset with barrier and boundary line, a side line roadside, was gravitation with the center line of obstacles borders line and lane boundary line.Fig. 8 is the pilotless automobile driving trace, can find out from driving trace, and when in the road barrier being arranged, the cut-through thing that pilotless automobile can be level and smooth, and behind the cut-through thing, the level and smooth road axis of getting back to.Fig. 9 is a steering wheel angle; Can find out from the steering wheel angle that provides; Expect that corner is smooth and satisfy dynamics of vehicle constraint (having under the obstacle environment, utilizing steering wheel angle of the present invention can remain in 120 degree, steering wheel angle speed also can remain in the restriction range of 200 degree/seconds).

The description of the foregoing description is can understand and use the present invention for ease of the those of ordinary skill of this technical field.The personnel of skilled obviously can easily make various modifications to these embodiment, and needn't pass through performing creative labour being applied in one principle of this explanation among other embodiment.Therefore, the invention is not restricted to the embodiment here, those skilled in the art should be within protection scope of the present invention for improvement and modification that the present invention makes according to announcement of the present invention.

Claims (10)

1. device that is used for the pilotless automobile local paths planning, it is characterized in that: it comprises:

Environmental perception device is used for the detecting obstacles thing, sets up road boundary model and road-center line model;

The repulsion calculation element is used to set up the repulsion point function and calculates repulsion;

The gravitation calculation element is used to set up the gravitation point function and calculates gravitation;

Resultant direction angle calculation device is used to calculate the orientation angle of making a concerted effort of repulsion and gravitation;

The steering wheel angle calculation element is used for confirming steering wheel angle according to the orientation angle and the steering ratio of gear of making a concerted effort.

2. the device that is used for the pilotless automobile local paths planning as claimed in claim 1; It is characterized in that: said environmental perception device is vision sensor and radar; Vision sensor is surveyed road boundary and is calculated road axis; The radar detection obstacle information, the path coordinate dot information that vision sensor is provided fits to repeatedly curve, sets up road boundary model and road-center line model.

3. the device that is used for the pilotless automobile local paths planning as claimed in claim 1; It is characterized in that: said environmental perception device is a radar, radar detection curb, match road boundary information; Extrapolate road axis information, set up road boundary model and road-center line model.

4. the device that is used for the pilotless automobile local paths planning as claimed in claim 1 is characterized in that: the real-time information of the environment that said repulsion calculation element provides according to environmental perception device confirms that the border, two road that vehicle will go sets up the repulsion point function; And according to the change of repulsion point function calculating along with environment, the position of automatic driving car repulsion point and calculating repulsion point are to the repulsion size of pilotless automobile.

5. the device that is used for the pilotless automobile local paths planning as claimed in claim 1; It is characterized in that: said gravitation calculation element through the deviation calculation device calculate with new boundary line be the road axis of road with the road axis of vehicle place road at present between lateral separation, be subordinate to composite function through Gauss simultaneously the information real-time that clear, obstacle distance pilotless automobile distance are arranged in the road be reflected on the gravitation point function; And according to the change of gravitation point function calculating along with environment, the position of automatic driving car gravitation point, and calculate the gravitation size of gravitation point to pilotless automobile.

6. method that is used for the pilotless automobile local paths planning, it is characterized in that: it comprises:

The environment sensing step, the detecting obstacles thing is set up road boundary model and road-center line model;

The repulsion calculation procedure is set up the repulsion point function and is calculated repulsion;

The gravitation calculation procedure is set up the gravitation point function and is calculated gravitation;

Resultant direction angle calculation step, the orientation angle of making a concerted effort of calculating repulsion and gravitation;

The steering wheel angle calculation procedure is confirmed steering wheel angle according to orientation angle of making a concerted effort and steering ratio of gear.

7. the method that is used for the pilotless automobile local paths planning as claimed in claim 6; It is characterized in that: said environment sensing step is that vision sensor is surveyed road boundary and calculated road axis; The radar detection obstacle information; The path coordinate dot information that vision sensor is provided fits to repeatedly curve, sets up road boundary model and road-center line model.

8. the method that is used for the pilotless automobile local paths planning as claimed in claim 6; It is characterized in that: said environment sensing step is the radar detection curb; Match road boundary information is extrapolated road axis information, sets up road boundary model and road-center line model.

9. the method that is used for the pilotless automobile local paths planning as claimed in claim 6 is characterized in that: the real-time information of the environment that said repulsion calculation procedure provides according to environmental perception device confirms that the border, two road that vehicle will go sets up the repulsion point function; And according to the change of repulsion point function calculating along with environment, the position of automatic driving car repulsion point and calculating repulsion point are to the repulsion size of pilotless automobile.

10. the method that is used for the pilotless automobile local paths planning as claimed in claim 6; It is characterized in that: said gravitation calculation procedure through the deviation calculation device calculate with new boundary line be the road axis of road with the road axis of vehicle place road at present between lateral separation, be subordinate to composite function through Gauss simultaneously the information real-time that clear, obstacle distance pilotless automobile distance are arranged in the road be reflected on the gravitation point function; And according to the change of gravitation point function calculating along with environment, the position of automatic driving car gravitation point, and calculate the gravitation size of gravitation point to pilotless automobile.

Images