Abstract: The potential eld method is widely used for autonomous mobile robot path planning due to its elegant mathematical analysis and simplicity. However, most researches were focused on solving the motion planning problem in a stationary environment, w
stant velocity and its position ptar and velocity vtar are known.
Assumption 3 The obstacles are assumed to be balls of radius ri centered at pobsi, with i= 1 2::: nobs, where nobs is the number of obstacles. The positions pobsi and velocities vobsi of the obstacles can be measured accurately. obstacle is close enough to the robot and needs to be avoided. The rest are assumed to be farther away and their in uences are neglected for that instant.
which does not contain the velocity information of the robot or the target. The corresponding conventional virtual attractive force is de ned as the negative gradient of the attractive potential in terms of position (3) F (p)=;rU (p)=;@Uatt(p) att att
@p
Assumption 4 At each time instant, only one
As the new attractive potential Uatt (p v) is a function of both the position p and velocity v of the robot, we shall de ne the corresponding virtual attractive force as the negative gradient of the attractive potential with respect to both position and velocity as follows:
3 ATTRACTIVE POTENTIAL FUNCTION Conventionally, the attractive potential is de ned as a function of the relative distance between the robot and the target only where the target is a xed point in space. In our research, it was found that it is bene cial to have the velocities of the robot and the target considered in the construction of the potential eld. In doing so, additional degrees of freedom are introduced for different tracking performances. When the target is moving, the conventional pure position based potential function is not directly applicable and has to be modi ed. In this paper, the potential eld functions are presented as follows+ v kvtar (t); v(t)kn (1) where p(t) and ptar (t) denote the positions of the robot and the target, respectively p= x y z]T in a 3-dimensional space or p= x y]T in a 2dimensional space v(t) and vtar (t) denote the velocities of the robot and the target, respectively kptar (t); p(t)k is the Euclidean distance between the robot and the target at time t kvtar (t); v(t)k is the magnitude of the relative velocity between the target and the robot at time t p and v are scalar positive parameters and m and n are positive constants which satisfy m n> 1. From (1), the attractive potential Uatt (p v) approaches its minimum zero if and only if the relative distance and velocity between the robot and the target are zero. The attractive potential Uatt (p v) increases as the relative distance or velocity between the robot and the target increases. It is easy to see that if v= 0 and m= 2, the new attractive potential function (1) degenerates to the conventional quadratic form
Fatt (p v)=;rUatt(p v)=;rp Uatt (p
v); rv Uatt (p v)(4) where
(6) with the subscripts p and v denoting the gradient with respect to position and velocity, respectively. Substituting (1) into (4), we have
rp Uatt (p v)=@Uatt (p v)@p rv Uatt (p v)=@Uatt (p v)@v
(5)
Fatt (p v)= Fatt1 (p)+ Fatt2 (v) where
(7) (8)
Uatt (p v)=
p kptar (t); p(t)km
Fatt1(p)= m pkptar (t); p(t)km;1 nRT
Fatt2(v)= n v kvtar (t); v(t)kn;1 nV RT (9) with nRT being the unit vector pointing from the robot to the target and nV RT being the unit vec-
tor denoting the relative velocity direction of the target with respect to the robot. The relationship between the attractive force and the position and velocity of the robot and the target in a 2-dimensional space is illustrated in Figure 1. The attractive force Fatt consists of two components: while the rst component, Fatt1 (p), pulls the robot to the target and shortens the distance between them, the second component, Fatt2 (v),\tries" to drive the robot to move at the same velocity of the target. From equations (8) and (9), since m> 1 and n> 1, we can see that when the robot approaches the target, i.e. kptar (t); p(t)k approaches zero, Fatt1 approaches zero, and when the velocity of the robot approaches that of the target, Fatt2 approaches zero. Then, when both the position and velocity of the robot approach those of the target, the attractive force Fatt approaches zero, i.e. when the robot catches the target and at the same
Uatt (p v)= Uatt (p)= p kptar (t); p(t)k2
(2)
Motion Planning for Humanoid Robots James Kuffner1...Using standard dynamic programming techniques, we ...Ahuja. A potential ?eld approach to path ...
Using Potential Field for Mobile Ad-hoc Networks_...By traditional method, the robots who lost the ...Planning," IEEE Transactions on Robotics and ...
3.S S GE,Y J CUI.Dyn咖i c Motion Planning for Mobile Robots Using Potential Field Method〔J〕.Autonomous Rootbs,20o2;(13):2O7—222 4.Zheng ...
On-line real-time path planning of mobile robots in dynamic uncertain environment_专业资料。维普资讯 http://www.cqvip.com 56l Zunt1/hjagUiSINC 20t11-...
An Evolutionary Method of Traditional Artificial Potential Field WANG Shunhong1...“Dynamic Motion Planning for Mobile Robots Using Potential Field Method”. ...
using two instances of the dynamic wave expansion...motion planning for mobile robots, many methods ...4. Discretized potential ?eld over a 60x60 ...
dynamic robot motion planning[外文期刊] 2000(02) 5.Moravec H P Sensor ...for mobile robots 1988(02) 6.Koren Y;Borenste in J Potential field ...
dynamic robot motion planning[外文期 刊] 2000(02) 5.Moravec H P Sensor ...for mobile robots 1988(02) 6.Koren Y;Borenste in J Potential field ...
Motion Planning for Industrial Robots Based on ...dynamic limits of maximum velocities, accelerations ...A new method for smooth trajectory planning of ...
Cui.Dynamic Motion Planning for Mobile Robots Using Potential Field Method[J].Autonomous Robots,2002,13(3). 六、指导教师意见 同意 指导教师签字: 2010 年...
热门文档