JOURNAL OF COMPUTER SCIENCE AND ENGINEERING, VOLUME 17, ISSUE 1, JANUARY 2013 11

Virtual Distance Sensor Information for Effective Obstacle Avoidance of Mobile Robots
Marwa T. Yousef, Shahira M. Habashy, Hosam Eldin I. Ali and Elsayed M. Saad
AbstractIn this paper, an optimization technique for the virtual sensors calculations is proposed to reduce the computation time during the motion of robot with obstacles avoidance. This less computation has a good effect on the reaction of the robot. Where, it gives the robot a high reactionspeed. Advanced Artificial Potential Field (AAPF) controller is used to guide the robot towards the best next step. Genetic Algorithms is also used to select the optimum values of the forces factors at AAPF controller. These optimum values make the robot's path much smoother. Simulation on Windows Vista using Matlab Software is executed to illustrate the proposed approach. Results are compared to previous work illustrating the superiority of the proposed work. Index TermsMobile Robot, Obstacle Avoidance, Virtual Sensor, Advanced Artificial Potential Field Controller.

1 INTRODUCTION

obile robots is improved from path guidance methods to autonomous mobility, resulting on obstacle avoidance of robots have emerged steadily over the years including directional command methods, such as artificial potential fields [1-5] and speed-space commands, such as the curvature foundation system [69]. However, most previous results remain a challenging problem as most existing methods have not considered the mobility of robots and obstacles. These algorithms cause a robot to move very slowly for obstacle avoidance. If a robot moves very slowly, most of the established algorithms can be applied to avoid obstacles. As it moves faster and faster, avoidance control is more difficult and the robot tends to collide more frequently with obstacles. The virtual sensor concept that was introduced in [10] is modified in our work to less computation. This concept is similar to that of the Doppler Effect. When a robot heads to an obstacle, the distance on the robot sensor is longer than the virtual sensor. Likewise, the physical distance on the robot sensor is shorter than the virtual one when it goes away from an obstacle. The introduced modification of virtual sensor at computation permits the robot to acquire a high speed reaction. The philosophy of the artificial potential field approach which is used in the current work can be schematically described as: the robot moves in a field of forces. The position to be reached is an attractive pole for the robot and obstacles are repulsive surfaces for the manipu

lator parts. The net force determines the behavior of the robot against the objects in its environment. The theory of the advanced artificial potential field (AAPF) approach is described in [11]. The Genetic Algorithm (GA) is used as in [12] to select the optimum factors of the repulsive and the attractive forces in the offline state which can be then applied on the robot in the online state. The optimum factors are those that make the robot motion much smoother. Simulation on Matlab is executed for testing the performance of the proposed system in robot motion with obstacle avoidance. The simulation is done at four workspaces as in [10], [12]. The simulation results are then used to compare the performance of the proposed system and the established system in [10], [12] to evaluate the proposed system effectiveness. The odometry of the robot is calculated using the angle approach [13]. This paper is organized as follows. Section 2 illustrates the proposed scheme of virtual sensor calculations. Section 3 shows the Advanced Artificial Potential Field (AAPF). Section 4 describes the determination of the optimum factors of the potential field controller forces using GA and illustrates the GA fitness function which is used in our work. Section 5 shows the performance of the proposed system by simulation in four different cases. In addition, comparison with the established system in [10], [12] will be reviewed. Conclusion and future work are given in Section 6.

Marwa T. Yousef is with the Automatic Control Department, Tibben Institute for Metallurgical Studies, 11913. Shahira M. Habashy is with Faculty of Engineering, Helwan University, 11792. Hosam Eldin I. Ali is with Faculty of Engineering, Helwan University, 11792. Elsayed M. Saad is with Faculty of Engineering, Helwan University, 11792.

2 VIRTUAL SENSORCALCULATIONS
The concept is used in [10] needs to large computation which decreases the speed of robot reaction during its motion. So, a modification to obtain simple computation to increase the speed of robot reaction is proposed. To make the distance on the robot sensor longer than the virtual sensor when a robot heads to an obstacle and

12

shorter than the virtual one when it goes away from an obstacle, the square of the actual sensor reading should be applied to give the virtual value that achieves our goal. When the actual distance is less than 1, 1 is the threshold reading between the robot and the obstacle, it is squared. This makes the virtual reading smaller than the actual. But when the actual is more than the threshold reading, squaring it makes the virtual reading bigger than the actual. This idea is illustrated by the following characteristics curve. Using higher power than 2 makes the virtual reading devolve zero.

Where: Pr t is the position of the robot which is calculated by using the angle approach as in [13], Po t is the position of the obstacle, Pg is the goal position, K r is the repulsive force factor, is the attractive force factor, is the virtual sensor reading of the distance between the robot and the obstacle which is calculated as in [10], and o is the threshold distance between the robot and the obstacle. The forces on the robot can be calculated from the gradient function of the artificial potential field as follows [10].

= , , = , , = Fo P +Fg P (4)
Where, P=Pr t , Fo Pr t is the repulsive force and Fg Pr t is the attractive force.

4 DETERMINING THE OPTIMUM FACTORS AND FITNESS FUNCTION

The attractive and repulsive coefficients Ka, Kr of the forces are determined empirically in [10], that shouldn't grantee a fast and smooth robot path. To overcome that empirical factors determination, one of the optimization techniques such as (GA) can be used to better determine those factors. The fitness function which is used in GA optimizer is the smoothness (SM) that is defined as a criterion of the evaluation to measure the various robot trajectories as:

Fig. 1. The characteristics of the virtual sensor reading

Of course the proposed modification with this simple computation leads to increases the speed of robot reaction.

=0

1
1 1

(5) (6)

= 1

3 ADVANCED ARTIFICAIL POTENTIAL FIELD (AAPF)

The controller generates a virtual force field with repulsion between obstacle and robot and attraction between target and robot. As the robot approaches the obstacle the repulsive force increases. On the other hand, the attractive force to the goal will be decreased as it approaches its goal. So, the artificial potential field Uart consists of attractive Ug and repulsive Uo forces. Where, the repulsive force is inversely proportional to the distance between the robot and the obstacles. And the attractive force is proportional to the distance between the robot and the goal. These logic rules are stated by the following equations [10], [11].

Uart Pr t ,Po t ,Pg = Uo Pr t ,Po t +Ug Po t ,Pg (1) Uo Pr t ,Po t Ug Po t ,Pg = =

1 2

K r ln 0
2

if o if >o

(2)

Wherem= the last pose number of robot trajectory, xn and yn represent the robot position at nth sampling time. n is the angle between current and former position at nth sampling time. First the optimum parameters for the potential field controller are obtained in offline state. It includes guided random search operation using GA to determine the optimum factors for Kr and Ka which are the factors of the repulsive and the attractive forces respectively. Four different environments for testing the improved potential field controller are used. The first environment has one obstacle, the second environment has two obstacles, the third one has three obstacles, finally the fourth has multiknee corridor. The start point of the robot at the first three workspaces are (0,0), while at the fourth, the start point of the robot is (2,0). The end points (goals) at the four workspaces are (4,2), (4,3), (4,3) and (6.5,6.5) respectively. These cases are those typically used in [10]. Ka and Kr values for each workspace according to using GA are denoted in Table 1.

Pr t Pg

(3)

13

TABLE 1 THE OPTIMUM OALUES OF FORCES FACTORS AT THE FOUR CASES CASE 1 KA KR 14.9226 0.0511 CASE 2 24.8609 0.0631 CASE 3 12.096 0.0084 CASE 4 8.3922 0.0013

5 SIMULATION RESULTS
The robot paths at the four cases are shown in Fig. 2, 3, 4, 5 respectively. Where: Fig. 2.a, Fig. 3.a, Fig. 4.a and Fig. 5.a represent the simulation results in [10].

but with complicated calculations. It didn't use the GA to optimize the forces factors.Fig. 2.b, Fig. 3.b, Fig. 4.b and Fig. 5.b represent the simulation results in [11]. This has used the AAPF controller which has less computation than APF controller with the GA to optimize the forces factors. And it used the virtual sensor concept which is proposed in [10]. Fig. 2.c, Fig. 3.c, Fig. 4.c and Fig. 5.c represent the simulation results of the proposed system which contains the improvement of virtual sensor calculations.

(a)

(a)

(b) (b)

(c) Fig. 2.The robot path (a) in [10], (b) in [11], and (c) the proposed system

(c) Fig. 3.The robot path (a) in [10], (b) in [11], and (c) the proposed system

Dong Jin Seo, Nak Yong Ko, and Jung Eun Son in [10] used the APF controller and the virtual sensor concept

14

(a) (b)

(b) (c) Fig. 5. the robot path (a) in [10], (b) in [11], and (c) the proposed system

At figures 2, 3, 4, and 5, the motion paths (a) at the four cases present that the robot with empirical values of Ka and Kr leads to slow evasive action than (b), (c). The robot in (a) of each figure makes abrupt changes in its paths while the robot at (b), (c) display smoother changes. The smoothness achieved with the proposed system is better than that achieved with the previous work. TABLE 2 THE COMPARISON OF THE SMOOTHNESS BETWEEN THE PROPOSED METHOD AND THE PREVIOUS W ORKS SMOOTHNESS

(c) Fig. 4. the robot path (a) in [10], (b) in [11], and (c) the proposed system

CASE 1 133.731

CASE 2 303.495

CASE 3 510.116

CASE 4 1420.295

APF W ITHOUT GA [10] AAPF W ITH GA [11] THE PROPROPOSED SCHEME

0.627

0.961

0.980

54.051

0.672

1.083

0.924

11.602

(a)

The smoothness achieved at each case is compared with the previous works and is shown in Table 2. As shown from the table 2, an optimization of the smooth-

15

ness is occurred with the proposed approach. When the path of the robot is smoother it means that the robot takes a short path and less time to go to the goal. And the reaction speed of the robot will be higher at the proposed system than [10], [11] due to simplifying the calculations of virtual sensor reading.

6 CONCLUSION
In this paper, Modification of virtual sensor calculations is introduced. Of course this modification gives less computation. Then, the reaction of the robot will achieve higher speed than the robot that uses a method that was introduced in [10]. Advanced artificial potential field controller is used. The potential fields controlling the robot motion are: attractive one (from the goal), and repulsive one from the obstacles. The GA optimizer was used in different robot environments to select the optimum factors of this controller to reach the goal while avoiding obstacles. The optimized parameters are tested in four different environments. The results show the superiority of the proposed controller generally and inclusively compared to the works in [10], [11]. Graceful optimization regarding the path motion is achieved. The real implementation of the proposed work and robots formation can be done as a future work.

[11] Marwa Taher, Elsayed Mostafa, Shahira Mahmoud, Obstacle Avoidance with Virtual Sensor in Mobile Robots Motion using the Advanced Potential Field Controller, International Journal of Computer Applications, (0975 8887) Volume 53 No.4, September 2012. [12] Marwa Taher, Hosam Eldin Ibrahim, Shahira Mahmoud, Elsayed Mostafa, Tracking of a Moving Target by Improved Potential Field Controller in Cluttered Environments, IJCSI International Journal of Computer Science Issues, Vol. 9, Issue 2, No 3, March 2012, pp. 472 480. [13] Philip Machler, "Robot Positioning by Supervised and Unsupervised Odometry Correction", PhD thesis, Department of Information, cole Poly Technique Federal of Lausanne, 1998. Marwa T. Yousefreceived her B.Sc. degree in Computer Engineering and her M.Sc. degree in Electronic Engineering from HelwanUniversity, Cairo, Egypt in 2001 and 2006, respectively. She has twopublished papers at M.Sc. degree. She is currently a Research Assistantat Tibben Institute for Metallurgical Studies, Cairo, Egypt. Herresearch interests include robot motion, tracking and their industrialapplications. Shahira M. Habashyreceived her B.Sc. degree in Communications& Electronics Engineering, her M.Sc. degree in Communications &Electronics Engineering, and her Ph.D. degree in Communications &Electronics Engineering from Helwan University, Cairo, Egypt, in1997, 2000, and 2006 respectively. She has seven published papers.She is currently a Teacher at Electronics, Communication &Computer Department, Faculty of Engineering, Helwan University. Hosam Eldin I. Alireceived his B.Sc. degree in Communications &Electronics Engineering, his M.Sc. degree in Computer Engineering, and his Ph.D. degree in Computer Engineering from Helwan University,Cairo, Egypt, in 2000, 2004, and 2009 respectively. He hasthree published papers at M.Sc. degree, and seven published papersat Ph.D. degree. He is currently a Teacher at Electronics, Communication& Computer Department, Faculty of Engineering,HelwanUniversity. Elsayed M. Saadis a Professor of Electronic Circuits, Faculty ofEngineering,Univ. of Helwan.He received his B.Sc. degree in ElectricalEngineering( Communication section) from Cairo Univ. ,his Dipl.-Ing degree and Dr.-Ing degree from Stuttgart Univ. , West Germany,at 1967,1977 and 1981 respectively . He became an Associate Prof.and a Professor in 1985, and 1990 respectively. He was an Internationalscientific member of the ECCTD, 1983. He is Author and/orCo-author of 132 scientific papers. He is a member of the nationalRadio Science Comittee, member of the scientific consultant committeein the Egyptian Eng. Syndicate for Electrical Engineers, till May1995, Member of the Egyptian Eng. Sydicate, Member of the EuropeanCircuit Society (ECS), and Member of the Society of Electrial Engineering (SEE).

REFERENCES
[1] O. Khatib, "Real-Time Obstacle Avoidance for Manipulators and Mobile Robots", Int. J. Robotics Research, Vol. 5, No. 1, Spring 1986, pp. 90-98. [2] J. Borenstein and Y. Koren, "The Vector Field Histogram-Fast Obstacle Avoidance for Mobile Robots", IEEE Trans. on Robotics Automation, Vol. 7, No. 3, June 1991, pp. 278-288. [3] J. Borenstein and Y. Koren, "Obstacle Avoidance with Ultrasonic Sensors", IEEE Journal of Robotics, Vol. 4, No. 2, 1988, pp. 213-218. [4] E. Shi, T. Cai, C. He and J. Guo, "Study of the New Method for Improving Artificial Potential Field in Mobile Robot Obstacle Avoidance", Proc. of IEEE International Conference on Automation and Logistics, 2007, pp 282286. [5] J. Gong, Y. Duan, Y. Man, and G. Xiong, "VPH+: An Enhanced Vector Polar Histogram Method for Mobile Robot Obstacle Avoidance", Proc. of International Conference on Mechatronics and Automation, 2007, pp. 27842788. [6] R. G. Simmons, "The Curvature-Velocity Method for Local Obstacle Avoidance", Proc. of IEEE International Conference on Robotics and Automation, April 1996, pp.3375-3382. [7] N. Y. Ko, R. G. Simmons, and K. S. Kim, "A Lane Based Obstacle Avoidance Method for Mobile Robot Navigation", Journal of Mechanical Science and Technology, Vol. 17, No. 11, 2003, pp. 1693-1703. [8] A. Kelly, "An Intelligent Predictive Control Approach to the High Speed Cross Country Autonomous Navigation Problem", Tech Report CMU-CS-TR-95-33, School of Computer Science, Carnegie Mellon University, 1995. [9] C. Schlegel, "Fast Local Obstacle Avoidance under Kinematic and Dynamic Constrains for a Mobile Robot", Proc. of IEEE/RSJ International Conference on Intelligent Robots and Systems, 1998, Vol. 1, pp. 594- 599. [10] Dong Jin Seo, Nak Yong Ko, and Jung Eun Son, "A Method for Combining Odometry and Distance Sensor Information for Effective Obstacle Avoidance of Autonomous Mobile Robots", International Journal of Control, Automation, and Systems, Vol. 8, No. 3, 2010, pp. 597-603.