Mobile robot path planning using ant colony optimization

IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308
_______________________________________________________________________________________
Volume: 03 Special Issue: 11 | NCAMESHE - 2014 | Jun-2014, Available @ http://www.ijret.org 1
MOBILE ROBOT PATH PLANNING USING ANT COLONY OPTIMIZATION T. Mohanraj1, S. Arunkumar2, M. Raghunath3, M. Anand4 1Assistant Professor, Department of Mechatronics Engineering, Kongu Engineering College, Erode-638052 2Assistant Professor, Department of Mechatronics Engineering, Kongu Engineering College, Erode-638052 3PG Scholar, Department of Mechatronics Engineering, Kongu Engineering College, Erode-638052 4PG Scholar, Department of Mechatronics Engineering, Kongu Engineering College, Erode-638052 Abstract Currently Mobile Robot has been widely used in examination and navigation particularly where static and unknown surroundings are involved. Path planning is a crucial problem in mobile robotics. Path planning of robot refers to the determination of a path, a robot takes in order to carry out the necessary task with a given set of key parameters. To find best possible path from starting point to target point, that reduces time and distance, in a given environment, avoiding collision with obstacles is a current potential research area. This paper presents SACO and ACO-MH algorithm to solve the problem of mobile robot path planning such that to reach the target station from source station without collision. The SACO and ACO-MH algorithm will give the collision free optimal path. The result obtained with ACO-MH was compared with SACO. The mobile robot environment is treated as a grid based environment in which each grid can be represented by an ordered pair of row number and column number. The mobile robot is considered as a point in the environment, to reduce the computational complexities. The ACO-MH results show better convergence speed and reduction in computational time than that of SACO through multiple MATLAB experiments. Keywords: Mobile Robot, Path Planning, SACO, ACO-MH, Collision free and optimal path
--------------------------------------------------------------------***---------------------------------------------------------------------- 1. INTRODUCTION
Mobile robot path planning is an essential issue of mobile robotics in the past two decades. Path planning is the determination of a path that a robot must take in order to pass over each point in a location in a given environment while minimizing the total cost associated with the path [1-2] and path is a plan of geometric locus of the points in a known space where the robot has to pass through. In the past two decades, various conventional methods have been developed to solve the path planning problem; some methods are cell decomposition, road map and potential field [7]. The evolution of robot global path planning algorithm from 1980 until today shows that there are numerous types of path planning algorithms proposed by researchers to solve Mobile Robot path planning problem [11, 12,16]. These techniques show lack of alteration and a non-vigorous performance [8]. The drawbacks of potential field methods are discussed by [9]. To overcome the limitation of these approaches researchers explored to Heuristic methods. Heuristic methods have been developed progressively more over the past two decades to deal with high computational costs and complexities of classical methods, particularly for complex environments. While using heuristic algorithms, it is not sure to come across a solution, but if a solution is originated, it will be done much faster than classical methods. The main heuristic approaches employed in Mobile Robot Path Planning are Genetic Algorithm (GA), Particle Swarm Optimization (PSO), Ant Colony Optimization (ACO), Simulated Annealing (SA) and Tabu Search (TS) [6]. In case of static environment, the location of the obstacle is fixed and does not transform with time [14]. However in case of dynamic environment, the position of the obstacle changes with time [13, 16]. In case of robot navigation, a mobile robot reaches the target destination from source station, avoiding collision with obstacles and upon iterations gives an optimal path without any human involvement. In the last decade, genetic algorithms have been extensively used to generate the optimum path by taking the advantage of its strong optimization capability [14]. Since its appearance in approximately 1992 [5], ACO has been used for solve many optimization problems such as the Travelling Salesman problem (TSP) [4], Vehicle routing, Job shop Scheduling and Flow shop Scheduling. ACO is a search technique inspired by the foraging behaviour of real ants. ACO has the ability to solve a hard combinatorial optimizations problem, the use of Ant Colony Optimization has contribute to the achievement of many investigation on Robot path planning. Any Colony Optimization algorithm is used to solve the mobile robot path planning problem in such a way that the artificial ant reaches the target point from source point avoiding obstacles [15]. A new method SACOdm to solve the navigational problem of mobile robot which is based on Simple Ant Colony Optimization Meta Heuristic (SACO-MH) was proposed in [10]. The results are optimistic for mobile robot path planning. In this study, we present an initial idea based on Simple Ant Colony optimization to select the shortest path in static known environment. The rest of this paper is organized as follows. The description of the environment and problem is described in section II, whereas section III presents the methodology of ACS and ACO-MH Algorithm in detail.

_______________________________________________________________________________________
Section IV illustrates the effectiveness of the ACS and ACO-MH
algorithm through some simulations and results. Finally,
section V concludes the paper.
2. PROBLEM DEFINITION
The mobile robot environment is treated as a grid based
environment in which each grid can be represented by an
ordered pair of row number and column number. The
environment of the mobile robot is represented in 2-D grid
model. The map consists of a 20x20 square grids of identical
pattern which is shown in Fig 1. The X-axis is divided
equally into 20 parts and the Y-axis is also divided into 20
equal parts. The size of grids is considered in such a manner
that they can be used to accommodate obstacles of variable
size and shape. The edge of each cell is of unit length. The
origin or the source station (S) of the robot is at the bottom
left corner with coordinate (1, 1). The target station (G) of
the robot is at the top right corner with coordinate (20, 20).
The mobile robot is considered as a point in the environment
to reduce the computational complexities. We need to plan a
path between two specified locations, a start and goal point.
The path should have no collision and satisfies certain
optimization criteria (shortest distance).
Fig- 1Grid model for the problem
3. METHODOLOGY
The objective is to construct a shortest path from the source
station S, to the target station (goal point) G, which avoids
every obstacle in the map; (a path which not touch any
obstacle). To solve this problem, Simple Ant Colony
optimization (SACO) and Ant colony optimization Meta–
Heuristic (ACO-MH) Algorithm is used [5]. Ant Colony
Optimization (ACO) simulates the behavior of ant colony in
nature when they are foraging for food and finding the most
efficient routes from their nests to food sources, it is a
stochastic search algorithm, and good effect has been
obtained in solving function and combination optimization,
identification of various system, Path planning of mobile
robots, Mining of data, network routing by using the
algorithm. According to the intensity of the pheromone trail
and evaporation the probability of the motion path is chosen
by an ant. The probability is called as the transition
probability [17]. It is also known as Ant Colony System.
The following terms used in ACO.
Artificial ant: In this work simulation agents are the mobile
robots that are inspired from the real ants which makes
movements based on the attraction of pheromone.
Pheromone: Chemical essence deposited by an ant when
walking; each ant probabilistically prefers to follow a
direction rich in pheromone rather than a poorer one.
The equation for calculating the transition probability in the
ant colony algorithm is following [4]:
{ }
{ }







j
j
k
i
k i
ij m
i
j N
p (1)
Where, ij = pheromone trail, m = No. of Ants
 =Weight value (Positive Constant) [4]
In the equation (1)
k
ij  represents the transition probability
in which ant k will navigate from node i to node j
The numerator on the right side of the equation represents
the intensity of the pheromone trail {ij} between nodes i
and j with a corresponding weight value of α. The
denominator on the right side of the equation is a summation
of the products of the pheromone intensity for all possible
moving paths [14].
Travelling paths have been constructed to be defined by all
ants. The pheromone trails are updated in all iteration. This
is done by first evaporative the pheromone value on all
paths by a constant factor, and adds pheromone on the paths.
Pheromone evaporation is implemented by [3].
ij =(1-ρ)* ij (2)
Where 0 < ρ ≤1 is the pheromone evaporation rate. The
parameter ρ is utilized to avoid limitless accumulation of the
pheromone trails and it enables the algorithm to “forget” bad
decision previously taken [14]. After evaporation all ants
deposit pheromone on the paths they have crossed in the
motion paths. The change of the pheromone is following
[5]:
ij ij ij
1
  

 
m
k
(3)
Where, Δij is the amount of pheromone ant k deposits on
the paths it has visited. It is given as:
ij
1
  k C
(4)

_______________________________________________________________________________________
Where
Ck, the length of the path build by the kth ant.
3.1 Pseudo Code of ACS Algorithm for Mobile
Robot Path Planning
The Pseudo Code sequence of ACS is given below [18].
Represent the solution space by constructing a grid map
Locate the static obstacles, start and goal point
Initialize ACO parameters
If iteration (i) =1, 2, 3, 4, 5, 6, ………………… + n
Else if ant (m) =1, 2, 3, 4, 5, 6, ……………… + n
Else if nodes (n) =1, 2, 3, 4, ……………… + n
Compute the probability of the mth ant next nodes
Move to the next nodes by computed probability
Store history of past location of nodes in an array
If current location of nodes is equal to destination
Break the nodes (n) loop
End
End
Evaluate fitness and store path distance of mth ant
Compute pheromone amount generated by mth ant
End
Update pheromone amount of the entire map
End
4. SIMULATION RESULTS AND DISCUSSION
The result of applying the Ant Colony Algorithm (ACS and
ACO-MH Algorithm) is presented in this section. In order to
make objective conditions three experiments were
conducted with different test conditions. Simple experiments
are conducted with varying the number of ants. All
simulations were done with MATLAB R2009 (b). The
parameter specifications are shown in Table 1. For all the
experiments the pheromone evaporation rate (ρ) is fixed as
0.10.
4.1 Experiment 1
In this experiment the obstacles are located in random
manner for obtain the optimal path. The Figures 2 and 3
shows the results of SACO and ACO-MH algorithms. The
feasible path cost 31.799 cm was found for SACO and for
ACO-MH 28.6274 cm was obtained which has the
minimum path cost compare to the SACO algorithm. While
increasing the number of ants the algorithm founds the
optimal path. If the number of ants increases beyond certain
level there is no considerable change in the path distance.
Table 1 Parameter Specification for ACO algorithm
Parameters SACO ACO-MH
No. of Ants (m) 10 10
Weight value (α) 0.25 0.25
Heuristic Factor (β) 0 1
Radius of the obstacles 1 1
pheromone evaporation
rate (ρ)
0.10 0.10
Fig- 2 Optimal Path found with SACO for Experiment 1
Fig- 3 Optimal Path found with ACO-MH for Experiment1

_______________________________________________________________________________________
4.2 Experiment 2 In this experiment the obstacles are located in vertical line to block optimal diagonal path. The Figure 4 and 5 shows the optimal results of SACO and ACO-MH algorithms. The optimal path found with ACO-MH is 32.1421 cm but in the case of SACO it is found that 33.8995 cm. Both of the algorithms are unable to follow the diagonal path because of the obstacles location on the grid map.
Fig- 5 Optimal Path found with ACO-MH for Experiment 2 4.3 Experiment 3
The static obstacles are placed in horizontal line to block the easiest diagonal path. The proposed ACO-MH algorithm utilizes the heuristic function to forget the bad decisions which previously taken by the ants (mobile Robots). The obtained results with ACS and ACO-MH algorithms are shown in figure 6 and 7 respectively. For this experiment there is no significant variation in the optimal path distance. Both the ant algorithms follow the path which is similar to each other. Hence the diagonal path was blocked by the obstacles already.
Fig- 7 Optimal Path found with ACO-MH for Experiment3 The Table2 shows the comparison of SACO and ACO-MH algorithms. From the Table 2 the ACO-MH algorithm performs better than SACO in both execution time and path cost. Table 2 Comparison of Results
Ex No
SACO (ACS)
ACO -MH
Distance (cm)
Time (sec)
Distance (cm)
Time (sec)
1
31.799
3.669818
28.6274
3.266521
2
33.8995
6.568700
32.1421
6.903165
3
32.7279
2.398471
32.1421
2.369687

_______________________________________________________________________________________
5. CONCLUSIONS
This paper presents the path planning of mobile robots using
the ant colony system algorithm. It reach the target using the
ability of the optimization of ant colony algorithm, and
control the mobile robot moving to the target (food) from
the starting point (nest) in the given static environment. The
Number of experiments has been conducted by varying the
type of obstacle and number of ants. The output is found to
be optimal and satisfying for the given problem with the
parameters α = 0.25, ρ= 0.1 and number of ants m =10.
Further the experiments were extended to change the
obstacle shape and location. With increase in complexity of
the problem, i.e. with increase in number of obstacles and
large number of obstacles exist in the map; this algorithm
will require a significant amount of computation time.
REFERENCES
[1] Abiyev, R., Ibrahim, D., & Erin, B. (2010).
Navigation of mobile robots in the presence of
obstacles. Advances in Engineering Software, 41(10),
1179-1186.
[2] Buniyamin, N., Wan Ngah, W. A. J., Sariff, N., &
Mohamad, Z. (2011). A simple local path planning
algorithm for autonomous mobile robots.
International journal of systems applications,
Engineering & development, 5(2), 151-159.
[3] Cong Y.Z., & Ponnambalam S.G., (2009). Mobile
robot path planning using ant colony optimization.
IEEE International Conference on Advanced
Intelligent Mechatronics, pp. 851-856
[4] Dorigo, M., & Gambardella, L. M. (1997). Ant
colony system: a cooperative learning approach to
the traveling salesman problem. Evolutionary
Computation, IEEE Transactions on, 1(1), 53-66.
[5] Dorigo, M., & Blum, C. (2005). Ant colony
optimization theory: A survey. Theoretical computer
science, 344(2), 243-278.
[6] M. Dorigo and T. Stutzle (2004), “Ant Colony
Optimization”, MIT Press, England.
[7] Masehian, E., & Sedighizadeh, D. (2007). Classic
and heuristic approaches in robot motion planning-a
chronological review. World Academy of Science,
Engineering and Technology, 29(1), 101-106.
[8] Gu, D., & Hu, H. (2002). Neural predictive control
for a car-like mobile robot. Robotics and
Autonomous Systems, 39(2), 73-86.
[9] Hu, L., Gu, Z. Q., Huang, J., Yang, Y., & Song, X.
(2008). Research and realization of optimum route
planning in vehicle navigation systems based on a
hybrid genetic algorithm. Proceedings of the
Institution of Mechanical Engineers, Part D: Journal
of Automobile Engineering, 222(5), 757-763.
[10] Koren, Y., & Borenstein, J. (1991, April). Potential
field methods and their inherent limitations for
mobile robot navigation. Proceedings of IEEE
International Conference on In Robotics and
Automation, pp. 1398-1404.
[11] Garcia, M. A., Montiel, O., Castillo, O., Sepúlveda,
R., & Melin, P. (2009). Path planning for
autonomous mobile robot navigation with ant colony
optimization and fuzzy cost function evaluation.
Applied Soft Computing, 9(3), pp. 1102-1110.
[12] Sariff, N., & Buniyamin, N. (2006, June). An
overview of autonomous mobile robot path planning
algorithms. Proceedings of 4th Student Conference on
Research and Development, IEEE,(SCOReD 2006).
pp. 183-188.
[13] Sivanandam, S. N., & Deepa, S. N. (2008).
Introduction to Genetic Algorithms. (Vol. 2).
Springer.
[14] Soh Chin Yun, Veleppa Ganapathy & Lim Ooi
Chong (2010), “Improved Genetic Algorithms based
Optimum Path Planning for Mobile Robot”,
Proceedings of IEEE 11th International Conference
on Control, Automation, Robotics and Vision
Singapore, pp. 1565 -1570
[15] Chia, S. H., Su, K. L., Guo, J. H., & Chung, C. Y.
(2010, December). Ant colony system based mobile
robot path planning. Proceedings of Fourth
International Conference on Genetic and
Evolutionary Computing IEEE, (ICGEC) pp. 210-
213.
[16] Stentz, A. (1994, May). Optimal and efficient path
planning for partially-known environments.
Proceedings of IEEE International Conference on
Robotics and Automation, pp. 3310-3317.
[17] Brand, M., Masuda, M., Wehner, N., & Yu, X. H.
(2010, June). Ant colony optimization algorithm for
robot path planning. International Conference on In
Computer Design and Applications (ICCDA), 2010
(Vol. 3, pp. V3-436). IEEE.
[18] Dong, J., Liu, B., Peng, K., & Yin, Y. (2009, May).
Robot Obstacle Avoidance based on an Improved
Ant Colony Algorithm. WRI Global Congress on
Intelligent Systems, 2009. GCIS'09. IEEE. Vol. 3, pp.
103-106).
BIOGRAPHIES
T.Mohanraj graduated from the Anna
University, Chennai with a Master Degree
in Mechatronics Engineering (Hons.).
Currently he is as an Assistant Professor at
Kongu Engineering College, Erode. His
current research interest is in Condition
monitoring, Industrial Automation & Robotics and
Optimization techniques.
S.Arunkumar graduated from the Anna
University, Chennai with a Master Degree
in Mechatronics Engineering. He is now an
Assistant Professor at Kongu Engineering
College, Erode. His current research
interest is in MEMS, and Industrial
Automation & Robotics.

_______________________________________________________________________________________
M.Raghunath graduated from the Anna University, Chennai with a Bachelor Degree in Electrical and Electronics Engineering. He is now pursuing his Master degree in Mechatronics Engineering in Kongu Engineering College, Erode.
M.Anand graduated from the Anna University, Chennai with a Bachelor Degree in Electrical and Electronics Engineering. He is now pursuing his Master degree in Mechatronics Engineering in Kongu Engineering College, Erode.

Mobile robot path planning using ant colony optimization

Recommended

More Related Content

What's hot (17)

Viewers also liked (20)

Similar to Mobile robot path planning using ant colony optimization (20)

More from eSAT Publishing House (20)

Recently uploaded (20)

Mobile robot path planning using ant colony optimization