See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/267204334

Optimal path planning of mobile robots: A review

Article  in  International Journal of Physical Sciences · February 2012

DOI: 10.5897/IJPS11.1745

CITATIONS READS
232 4,317

2 authors:

P. Raja S. Pugazhenthi
SASTRA University SASTRA University
33 PUBLICATIONS   567 CITATIONS    15 PUBLICATIONS   445 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Robotic Disease Guidance System for selected Crops View project

UAV systems and applications View project

All content following this page was uploaded by P. Raja on 09 July 2015.

The user has requested enhancement of the downloaded file.

International Journal of Physical Sciences Vol. 7(9), pp. 1314 - 1320, 23 February, 2012
Available online at http://www.academicjournals.org/IJPS
DOI: 10.5897/IJPS11.1745
ISSN 1992 - 1950 ©2012 Academic Journals

Review

Optimal path planning of mobile robots: A review

P. Raja* and S. Pugazhenthi
School of Mechanical Engineering, SASTRA University, Thanjavur – 613401, Tamilnadu, India.
Accepted 30 January, 2012

Mobile robots are increasingly used in automated industrial environments. There are also other
applications like planet exploration, surveillance, landmine detection, etc. In all these applications, in
order that the mobile robots perform their tasks, collision-free path planning is a prerequisite. This
article provides an overview of the research progress in path planning of a mobile robot for off-line as
well as on-line environments. Commonly used classic and evolutionary approaches of path planning of
mobile robots have been addressed. Review shows that evolutionary optimization algorithms are
computationally efficient and hence are increasingly being used in tandem with classic approaches
while handling Non-deterministic Polynomial time hard (NP-hard) problems. Also, challenges involved
in developing a computationally efficient path planning algorithm are addressed.

Key words: Path planning, mobile robot, off-line environment, on-line environment, classic, evolutionary
algorithms.

INTRODUCTION

Mobile robots are increasingly being employed in many hierarchical routes for networks in wireless mobile
automated environments. Potential applications of mobile communication (Manousakis et al., 2005), planning the
robots include a wide range such as service robots for path for digital artists in computer graphics and in
elderly persons, automated guided vehicles for computational biology to understand probable protein
transferring goods in a factory, unmanned bomb disposal folding paths (Choset et al., 2005).
robots and planet exploration robots. In all these The objective of this paper is to provide an overview of
applications, the mobile robots perform their navigation path planning algorithms used in mobile robots. This
tasks using the building blocks (Siegwart and article provides an overview of the research progress in
Nourbakhsh, 2004) as shown in Figure 1. path planning of a mobile robot for off-line as well as on-
Navigation of a mobile robot involves perception of line environments. Commonly used classic and
environment, localization and map building, cognition and evolutionary approaches of path planning of mobile
path planning and motion control. While perception refers robots have been addressed. Further scope and
to understanding its sensory data, finding its pose or challenges involved in developing computationally
configuration in the surroundings is localization and map efficient path planning algorithms are also identified.
building. Planning the path in accordance with the task by
using cognitive decision making is an essential phase
before actually accomplishing the preferred trajectory by CATEGORIES OF PATH PLANNING ALGORITHMS
controlling the motion. As each of the building blocks is
by itself a vast research field, this paper reviews path Path planning of a mobile robot is to determine a
planning approaches. collision-free path from a starting point to a goal point
Apart from robotic applications, path planning finds use optimizing a performance criterion such as distance, time
in planning the routes on circuit boards, obtaining the or energy, distance being the most commonly adopted
criterion. Based on the availability of information about
environment, there are two categories of path planning
algorithms, namely off-line and on-line. Off-line path
*Corresponding author. E-mail: raja@mech.sastra.edu. Tel: +91 planning of robots in environments where complete
264101 136. Fax: +91 4362 264120. information about stationary obstacles and trajectory of
 Raja and Pugazhenthi 1315

Environment

Motion
Perception
control

Path Local map

“Cognition” and Position/Map Localization and

path planning map building

Task/Mission

Figure 1. Building blocks of mobile robot navigation.

moving obstacles are known in advance is also known as roadmap approach, networks of collision-free paths are
global path planning. When complete information about constructed connecting start and target points. The well
environment is not available in advance, mobile robot known roadmap methods are visibility graph and Voronoi
gets information through sensors, as it moves through the diagram.
environment. This is known as on-line or local path Visibility graph (Lozano-Perez and Wesley, 1979) is
planning. Essentially, on-line path planning begins its drawn by joining two vertices of mutually visible polygonal
initial path off-line but switches to on-line mode when it obstacles that are present between start and target
discovers new changes in obstacle scenario. points. The shortest path is then identified through the
roads obtained from the visibility graph. The method is
efficient in sparse environments as the number of roads
OFF-LINE PATH PLANNING ALGORITHMS is dependent on the number of polygonal obstacles and
their edges (Li et al., 2002; Siegwart and Nourbakhsh,
Examples of path planning in off-line environments are 2004). Another roadmap approach, the Voronoi diagram
service robots operating during maintenance period of a (Dunlaing and Yap, 1985) is constructed using via points
nuclear power plant, automated guided vehicles in a which are equidistant from two or more obstacles. As a
factory, etc. where there may not be any change in the result, the obtained path is safer but normally not shorter
captured environment map. (Masehian and Amin-Naseri, 2007; Garrido et al., 2011).
The cell decomposition approach (Lozano-Perez, 1983)
computes the C-space of the mobile robot decomposes
Classic approaches the resulting space into cells and then searches for a
route in the free space cell graph. Grid method (Brooks
A fundamental approach for formulating and solving the and Lozano-Perez, 1983; Zhu and Latombe, 1989;
path planning problem is the configuration space (C- Payton et al., 1993; Likhachev et al., 2005; Hachour,
space) approach (Lozano-Perez and Wesley, 1979). 2008a, b) is a popular cell decomposition approach
Though the idea was also exercised by Udupa (1977) in where grids are used to generate the map of the
his doctoral thesis, it was Lozano-Perez who extensively environment. The main difficulty is how to find the size of
used this in the perspective of path planning. The central the grids, the lesser the size of grids, the more accurate
idea of this approach is the representation of the robot as will be the representation of the environment. However,
a single point. Thus, the C-space of the mobile robot path using lesser grids will result in exponential rise in memory
planning problem is reduced to a 2-dimensional problem. space and search range (Zheng et al., 2007).
As robot is reduced to a point, each obstacle is enlarged
by the size of the robot to compensate. The path planning
literature was united around this approach by Latombe’s Evolutionary approaches
book (1991). Using C-space as the fundamental concept,
there are many classic path planning approaches like Classic approaches though found to be effective, take
roadmap approach, cell decomposition approach, etc. In more time in the determination of feasible collision-free
1316 Int. J. Phys. Sci.

path. Also, classic approaches tend to get locked in local and fish resulted in the possibilities of utilizing this
optimal solution which may be far inferior to the global behavior as an optimization tool. Compared to GA, the
optimal solution. Moreover, path planning of a mobile advantages of PSO are that PSO is easier to implement
robot in the presence of multiple obstacles is found to be and there are fewer parameters to be adjusted (Kennedy
non-deterministic polynomial time hard (NP-hard) and Eberhart, 1995). Qin et al. (2004) proposed an
problem (Canny and Reif, 1987). It becomes even more algorithm which finds out shortest path using graph
complicated when the environment is dynamic. These based approach and PSO. They used graph based
drawbacks make the classic approaches to be approach to obtain the collision–free paths in static
incompetent in complex environments (Sugihara and environments and then PSO along with mutation operator
Smith, 1997). Hence, evolutionary approaches such as to arrive at the shortest path.
Genetic Algorithm (GA), Particle Swarm Optimization Zhang and Gu (2008) used variable path length which
(PSO), etc. Similarly, Ant Colony Optimization (ACO) and depends upon number of vertices of polygonal obstacles.
Simulated Annealing (SA) are employed to solve the path Binary PSO along with genetic like mutation operator is
planning problem quickly. used to optimize the path. Nasrollahy and Javadi (2009)
GA is an optimization tool based on the mechanics of presented a PSO based planner for dynamic
natural genetics and selection (Holland, 1975; Goldberg, environments in which populations are generated
2000). The first step in path planning using GA is random containing invalid paths and then they are subjected to
generation of population containing alternative paths. penalty function evaluation. Recently, Gong et al. (2011)
Dozier et al. (1997) from NASA presented a hybrid proposed a model which uses a multi objective PSO and
planner which makes use of visibility-based repair genetic like mutation operator. The multi objectives are
approach and evolutionary technique. The visibility-based shortest distance and danger of path from the obstacles.
repair approach is used to quickly transform the ones The mutation operator is used to repair the invalid paths.
which interfere with the obstacles (invalid paths) into valid The other optimization algorithms which have been
paths and then subject to binary coded GA. GA with used in path planning to a smaller percentage are ant
binary string is computationally costly for the reason that colony optimization (ACO) and simulated annealing (SA)
before each evaluation of function, chromosomes are algorithms (Masehian and Sedighizadeh, 2007). ACO is
transformed to phenotypes (Ripon et al., 2007). Xiao et inspired by the foraging behavior of ants for finding the
al. (1997) proposed an evolutionary planner for path shortest path to the food source. A method for optimal
planning. This planner has comparatively simple path planning based on improved Dijkstra algorithm and
genotype structures that can represent valid paths, but ACO is proposed by Guan-zheng et al. (2006). The
necessitates complex decoders and fitness functions to improved Dijkstra algorithm consists of the standard
obtain the optimal path. Moreover, there may be loss of Dijkstra algorithm and removal of unnecessary path
accuracy in transforming to binary mode. nodes algorithm to get the sub-optimal paths. ACO is
Further, Sugihara and Smith (1997), Gallardo et al. used to obtain the global optimal path from the sub-
(1998), Nagib and Gharieb (2004) and Al-Taharwa et al. optimal paths. Garcia et al. (2009) presented simple ACO
(2008) used fixed-length path consisting of binary strings. – distance memory (SACO-dm) algorithm for global path
A fixed-length path gives a quick solution for planning among static and moving obstacles. In SACO-
environments with few obstacles and it takes hours to dm, optimal path is influenced by current distance
evolve a solution for a complex environment. In order to between robot and target nodes and the memory
reach the target in a complex environment, variable capability of ants remembering the visited nodes. Results
length chromosomes are needed. Tu and Yang (2003) show that optimal paths are achieved with lesser
presented variable length binary coded GA in which gene computation time compared to SACO.
indicates the subsequent movement direction and SA is a type of heuristic random search method and it
distance. The main limitation of such algorithms is that resembles the cooling process of molten metals through
they direct to some invalid results like paths that may not annealing. A method employing SA for collision-free path
reach to the target point at all (Shahidi et al., 2004). amid static polygonal obstacles in C-space setting is
Wang et al. (2006) presented a genetic based path presented in Martinez-Alfaro and Gomez-Garcia (1998).
planning algorithm, in which populations are generated Binary setting is used which may result in non-optimal
including the ones which interfere with the obstacles path when shorter computation time is desired. Miao and
(invalid paths also). Later such invalid path sequences Tian (2008) developed SA algorithm based approach for
are subjected to penalty function evaluation. This dynamic environments. Their approach uses vertices of
increases computation load resulting in higher execution the static and dynamic obstacles as search space. Then
time (Raja and Pugazhenthi, 2008, 2009a, b, 2011). the SA algorithm is used to find the optimal path.
PSO is another very widely used evolutionary algorithm
in path planning. It is an evolutionary computation ON-LINE PATH PLANNING ALGORITHMS
technique inspired by social behavior of bird flocking or
fish schooling. Years of study on the dynamics of bird In recent times, on-line path planning has received more
 Raja and Pugazhenthi 1317

attention from researchers (Masehian and Katebi, 2007) Khatib, 1999). The dynamic window contains the feasible
since autonomous mobile robots must be capable of linear and angular velocities taking into consideration
operating in dynamic environments. Applications of path acceleration capability of robot. Then the velocity at the
planning in on-line environments include planet next instant is optimized for obstacle avoidance, subject
exploration, mine industry, reconnaissance robots, etc. to vehicle dynamics.
(Hachour, 2008a). On-line path planning approaches like
potential field approach and collision-cone approach have
been traditionally followed. Nowadays, evolutionary Evolutionary approaches
approaches are increasingly being used along with
classic approaches. Although classic approaches are found to be effective,
computation time is crucial for the success of any on-line
path planning algorithms. But, with classic approaches,
Classic approaches optimum result can hardly be achieved in quick
computation time owing to incomplete information of the
Pioneering works have been initiated by Khatib (1986) environment. Further, due to NP-hard complexity of path
who proposed Artificial Potential Field (APF) approach planning problem, classic approaches are often
which is popular in mobile robotics. By this approach, a combined with evolutionary approaches like GA, PSO,
point robot in C-space moves under the influence of an etc. to overcome their drawbacks.
APF in which obstacles are assumed to generate Vadakkepat et al. (2000, 2001) proposed evolutionary
repulsive forces and the target is assumed to generate APF (EAPF) algorithm to derive optimal potential field
attractive forces. The robot moves as per the resultant of functions using GA. When the robot is trapped, a
these forces. This approach is known for its mathematical separate algorithm named escape-force is introduced to
elegance and simplicity as path is found with very little recover from trap. Potential field immune network
computation. However, the drawback of this algorithm is proposed by Luh and Liu (2008) used velocity obstacle
that robot may become stagnant or trapped when there is method to identify the most imminent collision of the
a cancellation of equal magnitudes of attractive and obstacle. Potential field approach coupled with
repulsive forces. One solution to overcome this problem biologically inspired immune network is used to avoid the
is to complement with influential algorithms to escape most imminent obstacle. The overall response of the
from trap (Latombe, 1991) and till date many variants of immune network is calculated using GA. Adaptive virtual
potential field approach like escape-force algorithm target algorithm is proposed to direct the robot out of the
(Vadakkepat et al., 2000, 2001), trap recovery model trap. It was assumed that robot is being trapped if it
used by Lv and Feng (2006), adaptive virtual target moves beyond 90° off-target.
algorithm (Luh and Liu, 2007, 2008), etc. have been Min et al. (2005) employed a mathematical model using
proposed. collision cone approach and PSO for on-line path
Path planning problem can also be solved by vector planning. To facilitate reduction of computational burden,
field histogram approach (Borenstein and Koren, 1991). they ignored instantaneous changes in obstacle velocities
At every instant, a polar histogram is generated to in the motion model. Therefore, their algorithm is suited
represent the polar density of obstacles around a robot. to sparse environments having obstacles with slow
The robot’s steering direction is chosen based on the velocities. Also, PSO in combination with binary coded
least polar density and closeness to the goal. In a given genetic operators is used as optimization tool without
environment, the polar histogram must be regularly considering dynamic constraints. Nevertheless, recent
regenerated for every instant and hence the method is studies show that real coded evolutionary algorithms
suited for environments with sparse moving obstacles. execute better than the binary coded.
Another commonly used on-line approach is based on Hu et al. (2007) addressed an approach to steer the
collision cone concept (Chakravarthy and Ghose, 1998; mobile robot in static or dynamic environments based on
Qu et al., 2004). Collision of robot can be averted if the PSO and stream functions (or potential flows). Stream
relative velocity of robot with regard to a particular functions, derived from hydrodynamics, are engaged to
obstacle falls exterior to the collision cone. Fiorini and steer an autonomous robot to avoid the obstacles.
Shiller (1998) proposed a velocity obstacle approach, However, their model also does not consider
which is in resemblance to collision cone approach. It instantaneous changes in obstacle velocities. Park and
consists of choosing avoidance maneuvers to avoid static Kim (2008) proposed a PSO algorithm based on the
and moving obstacles in the velocity space. They used potential field approach. The potential field is
basic heuristic strategy for prioritizing objectives such as mathematically modeled by particles’ fitness value. The
averting collisions, attaining the goal or accomplishing PSO particles are designed to move with Newtonian
trajectories with preferred topologies. dynamics. Lu and Gong (2008) proposed an on-line path
Another on-line approach for obstacle avoidance is planning algorithm using PSO technique for unknown
dynamic windows approach (Fox et al., 1997; Brock and environments. Their algorithm is entirely based on
1318 Int. J. Phys. Sci.

distance information of the environment without any 3) Path containing variable number of segments can be
mathematical model featuring velocity of nearing used in generation of population considering the
obstacle. Recently, Hong et al. (2011) presented a model complexity (number of vertices) of environment.
using classic APF considering dynamic model of velocity 4) Simple decoders and fitness functions can be used to
potential field obtained by a variant of PSO called decrease the computational time of algorithms.
quantum PSO. Inspired by the nature of motion of
microscopic particles (quantum mechanics), quantum Compared to off-line, developing a computationally
PSO updates state of a particle by wave function instead efficient on-line algorithm is more challenging. One of the
of position and velocity. difficulties in working with incomplete information of
Mei et al. (2006) proposed a hybrid algorithm which environment is that the path cannot be pre-planned and
combines APF and ACO for dynamic environments. ACO therefore global optimum solutions can hardly be
is applied to plan the global path and then APF is achieved. However, better quality paths can be achieved
employed to guide the robot for local route. Lv and Feng by addressing the challenges as follows:
(2006) proposed numerical potential field to model the
environment and applied ACO to search for optimal path. 1) More accurate mathematical models can be developed
A trap recovery solution was also discussed by them. The which feature instantaneous velocity of robot as well as
main problem of ACO is difficulty in obtaining the quick nearing obstacles to tackle even cluttered moving
solution convergence. Lee et al. (2008) presented obstacles.
improved ACO using potential field approach to get quick 2) The influence of other constraining obstacles while
solution convergence by tuning the control parameters of negotiating the most imminent obstacle can yield overall
ACO. Improved ACO makes use of altered pheromone (a better result. Therefore, mathematical model should
substance secreted by ants) to update the position simultaneously consider the effect of constraining
vector. obstacles which may cause further deviation more than
Zhang et al. (2004) proposed an APF approach in avoiding the imminent threat.
combination with SA which considers the problems of 3) Avoiding trap of a robot when there exists a path is
goal non-reachable with obstacles nearby (GNRON) and another vital issue in obstacle avoidance. The need for
local minima in soccer robots. New potential functions trap recovery can be eliminated.
have been derived by considering the distance 4) Any on-line path planning algorithm becomes truly
information of start and target points for GNRON successful if the next instant of the robot is planned within
problem. Miao (2010) presented a multi operator based the bounds of the kinematic and dynamic constraints of
SA approach for path planning. Switching, deleting, the mobile robot. So the mathematical models should
mutating and repairing operators are introduced along incorporate the upper bounds for the dynamic constraints
with SA. The parameters of SA are fine tuned for moving in tough time varying environments.
obstacles also.
Further, multiple optimization objectives, multiple robot
coordination, uncertainties in sensing, prediction, motion
SCOPE AND CHALLENGES control, etc. pose many other challenges in mobile
robotics.
In the domain of path planning, evolutionary methods
have proved to yield better results than pure classical
approaches (Garcia MAP et al., 2009). Several CONCLUSION
algorithms have been proposed to overcome the complex
nature of NP-hard path planning problem as efficient At present, development in path planning is progressively
functioning of mobile robot is influenced by better quality more inspired by new applications such as circuit board
of paths. The review of literature shows that there is still designs, network routings, computer animations,
scope for developing more efficient off-line path planning pharmaceutical drug designs, computational biology, etc.
algorithms that will yield better quality paths by The research community puts forward many approaches
addressing some of the issues as follows: for solving the path planning problem. This article reflects
the research progress that has taken place in path
1) Valid paths which do not interfere with the obstacles planning of mobile robots, including on-line planning.
should alone be considered in the initial generation of Although many efficient algorithms have been developed,
population eliminating the use of penalty function the diversity of path planning problems has been
evaluation. constantly increasing. Up to 90s, determination of
2) Optimization techniques with real strings are collision-free path remained the main objective. Currently,
computationally less expensive because before the though collision-free path is a necessary condition, other
evaluation of objective function, details of alternative significant issues such as modeling of dynamic
paths need not be expressed in binary codes. environment, multiple optimal functions, dynamic
 Raja and Pugazhenthi 1319

constraints, etc. are also to be addressed. These Hachour O (2008b). Path planning of autonomous mobile robot. Int.
constraints craft path planning problems to be more J. Syst. Appl. Eng. Dev., 4(2): 178-190.
Holland JH (1975). Adaptation in natural and artificial systems.
challenging and necessitate more robust and efficient University of Michigan Press, Ann Arbor, USA.
algorithms. Hong Z, Liu Y, Zhongguo G, Yi C (2011). The dynamic path
planning research for mobile robot based on artificial
potential field. Proceedings of the International Conference on
REFERENCES consumer Electronics, Communications and Networks (CECNet),
XianNing, pp. 2736-2739.
Al-Taharwa I, Sheta A, Al-Weshah M (2008). A mobile robot path Hu C, Wu X, Liang Q, Wang Y (2007). Autonomous robot path
planning using genetic algorithm in static environment. J. planning based on swarm intelligence and stream functions. Lect.
Comput. Sci., 4(4): 341-344. Notes Comp. Sci., Springer, pp. 277-284.
Borenstein J, Koren Y (1991). The vector field histogram – Fast Kennedy J, Eberhart RC (1995). Particle swarm optimization.
obstacle avoidance for mobile robots. IEEE T. Robot. Autom., Proceedings of the IEEE International Conference on Neural
3(7): 278-288. Networks, Perth, Australia, pp. 1942-1948.
Brock O, Khatib O (1999). High-speed navigation using the global Khatib O (1986). Real time obstacle avoidance for manipulators
dynamic window approach. Proceedings of IEEE International and mobile robots. Int. J. Robot. Res., 5(1): 90-98.
Conference on Robotics and Automation, Detroit, USA, pp. 341- Latombe JC (1991). Robot motion planning. Kluwer Academic
346. Publisher, Boston.
Brooks RA, Lozano-Perez T (1983). A subdivision algorithm in Lee JW, Kim JJ, Choi BS, Lee JJ (2008). Improved ant colony
configuration space for find path with rotation. Proceedings of optimization algorithm by potential field concept for optimal path
eighth International Joint Conference on Artificial Intelligence, planning. Proceedings of the eighth IEEE-RAS International
Karlsruhe, Germany, pp. 799-806. Conference on Humanoid Robots, Daejeon, pp. 662–667.
Canny J, Reif J (1987). New lower bound techniques for robot Li L, Ye T, Tan M, Chen X (2002). Present state and future
motion planning problems. Proceedings of IEEE Symposium on development of mobile robot technology research. Robot., 24(5):
the Foundations of Computer Science, Los Angeles, California, 475-480.
pp. 49-60. Likhachev M, Ferguson D, Gordon G, Stentz A, Thrun S (2005).
Chakravarthy A, Ghose D (1998). Obstacle avoidance in a dynamic Anytime dynamic A*: An anytime, replanning algorithm.
environment: A collision cone approach. IEEE T. Syst. Man Cy. Proceedings of the International Conference on Automated
A., 28(5): 562-574. Planning and Scheduling, pp. 262-271.
Choset H, Lynch K, Hutchinson S, Kantor G, Burgard W, Kavraki L, Lozano-Perez T, Wesley MA (1979). An algorithm for planning
Thrun S (2005). Principles of Robot Motion: Theory, Algorithms, collision-free paths among polyhedral obstacles. Commun.
and Implementations. Massachusetts Institute of Technology ACM., 22(10): 560–570.
Press, Cambridge, U.S.A. Lozano-Perez T (1983), Spatial planning: A configuration approach.
Dozier G, Esterline A, Homaifar A, Bikdash M (1997). Hybrid IEEE T. Comput. C, 32(3): 108-120.
Evolutionary Motion Planning via Visibility-Based Repair. Lu L, Gong D (2008). Robot path planning in unknown
Proceedings of the IEEE International Conference on environments using particle swarm optimization. Proceedings of
Evolutionary Computation, Indianapolis, pp. 507-511. the fourth International Conference on Natural Computation,
Dunlaing CO, Yap CK (1985). A retraction method for planning the Jinan, pp. 422-426.
motion of a disc. J. Algorithms, 6: 104-111. Luh GC, Liu WW (2007). Motion planning for mobile robots in
Fiorini P, Shiller Z (1998). Motion planning in dynamic environments dynamic environments using a potential field immune network.
using velocity obstacles. Int. J. Robot. Res., 17(7): 760-772. IMechE J. Syst. Control Eng., 221(7): 1033-1045.
Fox D, Burgard W, Thrun S (1997). The dynamic window approach Luh GC, Liu WW (2008). An immunological approach to mobile
to collision avoidance. IEEE Robot. Autom. Mag., 4(1): 23-33. robot reactive navigation. Appl. Soft Comput., 8(1): 30-45.
Gallardo D, Colomina O, Florez F, Rize R (1998). A genetic Lv N, Feng Z (2006). Numerical potential field and ant colony
algorithm for robust motion planning. Proceedings of the optimization based path planning in dynamic environment.
Eleventh International Conference on Industrial and Engineering Proceedings of the Sixth World Congress on Intelligent Control
Applications of Artificial Intelligence and Expert Systems, and Automation, Dalian, pp. 8966-8970.
Benicassim, Spain, pp. 115-121. Manousakis K, McAuley T, Morera R, Baras J (2005). Using multi-
Garcia MAP, Montiel O, Castillo O, Sepulveda R, Melin P (2009). objective domain optimization for routing in hierarchical networks.
Path planning for autonomous mobile robot navigation with ant Proceedings of the International Conference on Wireless
colony optimization and fuzzy cost evaluation. Appl. Soft Networks, Communication and Mobile Computing, pp. 1460-
Comput., 9: 1102-1110. 1465.
Garrido S, Moreno L, Blanco D, Jurewicz P (2011). Path Planning Martinez-Alfaro H, Gomez-Garcia S (1998). Mobile robot path
for mobile robot navigation using Voronoi diagram and fast planning and tracking using simulated annealing and fuzzy logic
marching. Int. J. Robot. Autom., 2(1): 42-64. control. Expert Syst. Appl., 15(3): 421-429.
Goldberg DE (2000). Genetic algorithms in search, optimization and Masehian E, Katebi Y (2007). Robot motion planning in dynamic
machine learning. Addison-Wesley, Delhi, India. environments with moving obstacles and target. Int. J. Mech.
Gong DW, Zhang JH, Zhang Y (2011). Multi-objective particle Syst. Sci. Eng., 1(1): 20-25.
swarm optimization for robot path planning in environment with Masehian E, Sedighizadeh D (2007). Classic and heuristic
danger sources. J. Comput., 6(8): 1554-1561. approaches in robot motion planning – A chronological review.
Guan-zheng T, Huan HE, Sloman A (2006). Global optimal path World Acad. Sci. Eng, Technol.. 23: 101-106.
planning for mobile robot based on improved Dijkstra algorithm Masehian E, Amin-Naseri MR (2007). Composite models for mobile
and ant system algorithm. J. Cent. South Univ. Technol., 13(1): robot offline path planning, Mobile robots: Perception and
80-86. navigation. InTech Publisher, Croatia.
Hachour O (2008a). The proposed genetic FPGA implementation Mei H, Tian Y, Zu L (2006). A hybrid ant colony optimization
for path planning of autonomous mobile robot. Int. J. Circ. Syst. algorithm for path planning of robot in dynamic environment. Int.
Sig. Proc., 2(2): 151-167. J. Inf. Tech., 12(3): 78-87.
1320 Int. J. Phys. Sci.

Miao H (2010). A multi-operator based simulated annealing Shahidi N, Esmaeilzadeh H, Abdollahi M, Lucas C (2004). Memetic
approach for robot navigation in uncertain environments. Int. J. algorithm based path planning for a mobile robot. Int. J. Inf.
Comput. Sci. Secur., 4(1): 50-61. Tech., 1(4): 174-177.
Miao H, Tian YC (2008). Robot path planning in dynamic Siegwart R, Nourbakhsh IR (2004). Introduction to autonomous
environments using a simulated annealing based approach. mobile robot. Massachusetts Institute of Technology press,
Proceedings of the tenth International Conference on Control, Cambridge, U.S.A.
Automation, Robotics and Vision, Hanoi, Vietnam, pp. 1253- Sugihara K, Smith J (1997). Genetic algorithms for adaptive motion
1258. planning of an autonomous mobile robot. Proceedings of the
Min HQ, Zhu JH, Zheng XJ (2005). Obstacle Avoidance with Multi- IEEE International Symposium on Computational Intelligence in
Objective Optimization by PSO in Dynamic Environment. Robotics and Automation, Monterey, California, pp. 138-143.
Proceedings of the IEEE International Conference on Machine Tu J, Yang SX (2003). Genetic algorithm based path planning for a
Learning and Cybernetics, Guangzhou, pp. 2950-2956. mobile robot. Proceedings of the IEEE International Conference
Nagib G, Gharieb W (2004). Path planning for a mobile robot using on Robotics and Automation, Taipei, Taiwan, pp. 1221-1226.
genetic algorithms. Proceedings of the International Conference Udupa S (1977). Collision detection and avoidance in computer
on Electrical, Electronic and Computer Engineering (ICEEC'04), controlled manipulators. PhD thesis, California Institute of
Cairo, Egypt., pp. 185-189. Technology, California, USA.
Nasrollahy AZ, Javadi HHS (2009). Using particle swarm Vadakkepat P, Lee TH, Xin L (2001). Application of evolutionary
optimization for robot path planning in dynamic environments artificial potential field in robot soccer system. Proceedings of the
with moving obstacles and target. Third European Symposium on Joint ninth IFSA World Congress and twentieth NAFIPS
computer modeling and simulation, Athens, Greece, pp. 60-65. International Conference, Vancouver, BC, Canada, pp. 2781-
Park H, Kim JH (2008). Potential and dynamics-based Particle 2785.
Swarm Optimization. Proceedings of the IEEE World Congress Vadakkepat P, Tan KC, Ming-Liang W (2000). Evolutionary artificial
on Computational Intelligence, Hong Kong, pp. 2354-2359. potential fields and their application in real time robot path
Payton DW, Rosenblatt JK, Keirsey DM (1993). Grid-based planning. Proceedings of the 2000 Congress on Evolutionary
mapping for autonomous mobile robot. Robot. Auton. Syst., Computation, Piscataway, New Jersey, pp. 256-263.
11(1): 13-21. Wang Y, Mulvaney D, Sillitoe I (2006). Genetic-based mobile robot
Qin YQ, Sun DB, Li N, Cen YG (2004). Path Planning for mobile path planning using vertex heuristics. Proceedings of the 2006
robot using the particle swarm optimization with mutation IEEE International Conference on Cybernetics and Intelligent
operator. Proceedings of the IEEE International Conference on Systems, Bangkok, pp. 463-468.
Machine Learning and Cybernetics, Shanghai, pp. 2473-2478. Xiao J, Michalewicz Z, Zhang L (1997). Trojanowski K, Adaptive
Qu Z, Wang J, Plaisted CE (2004). A new analytical solution to evolutionary planner/navigator for mobile robots. IEEE T. Evolut.
mobile robot trajectory generation in the presence of moving Comput., 1(1): 18-28.
obstacles. IEEE T. Robot., 20(6): 978-993. Zhang PY, Lu TS, Song LB (2004). Soccer robot path planning
Raja P, Pugazhenthi S (2008). Path planning for a mobile robot to based on the artificial potential field approach with simulated
avoid polyhedral and curved obstacles. Int. J. Assist. annealing. Robotica, 22: 563-566.
Robotics. Mechatronics, 9(2): 31-41. Zhang QR, Gu GC (2008). Path planning based on improved binary
Raja P, Pugazhenthi S (2009a). Path planning for a mobile robot particle swarm optimization algorithm. Proceedings of the IEEE
using real coded genetic algorithm. Int. J. Assist. Robot. Syst., International Conference on Robotics, Automation and
10(1): 27-39. Mechatronics, Chendu, China, pp. 462-466.
Raja P, Pugazhenthi S (2009b). Path planning for mobile robots in Zheng TG, Huan H, Aaron S (2007). Ant Colony System Algorithm
dynamic environments using particle swarm optimization. IEEE for Real-Time Globally Optimal Path Planning of Mobile Robots.
International Conference on Advances in Recent Technologies Acta. Automatica. Sinica, 33(3): 279-285.
in Communication and Computing (ARTCom 2009), Kottayam, Zhu DJ, Latombe JC (1989). New heuristic algorithms for efficient
India, pp. 401-405. hierarchical path planning. Technical report STAN-CS-89-1279,
Raja P, Pugazhenthi S (2011). Path Planning for a mobile robot in Computer Science Department, Stanford University, USA.
dynamic environments. Int. J. Phys. Sci., 6(20): 4721-4731.
Ripon KSN, Kwong S, Man KF (2007). A real-coding jumping gene
genetic algorithm (RJGGA) for multiobjective optimization. Int. J.
Inf. Sci., 177(2): 632-654.

View publication stats