Application of particle swarm optimization in path planning of mobile robot

Yong Wang, Feng Cai, and Ying Wang

Citation: AIP Conference Proceedings 1864, 020023 (2017); doi: 10.1063/1.4992840

View online: http://dx.doi.org/10.1063/1.4992840
View Table of Contents: http://aip.scitation.org/toc/apc/1864/1
Published by the American Institute of Physics

Articles you may be interested in

Dynamic path planning for mobile robot based on particle swarm optimization
AIP Conference Proceedings 1864, 020024 (2017); 10.1063/1.4992841
 Application of Particle Swarm Optimization in Path
Planning of Mobile Robot
Yong Wang, Feng Cai a) and Ying Wang

School of Computers, Guangdong University of Technology, Guangzhou, China

a)
Corresponding author e-mail: [email protected]

Abstract. In order to realize the optimal path planning of mobile robot in unknown environment, a particle swarm
optimization algorithm based on path length as fitness function is proposed. The location of the global optimal particle is
determined by the minimum fitness value, and the robot moves along the points of the optimal particles to the target
position. The process of moving to the target point is done with MATLAB R2014a.Compared with the standard particle
swarm optimization algorithm, the simulation results show that this method can effectively avoid all obstacles and get the
optimal path.

INTRODUCTION
At present, there are many classical algorithms for path planning, uch as Ant colony algorithm (ACA) [1-2],
rticle swarm optimization (POS)[3-4], nitic Algorithms[5]. In these methods, OS is a simple and effective method.
ut it has some shortcomings: the relatively poor local search ability, the speed updating formula is relatively simple,
which seriously affects the efficiency. Fitness function can solve this.

POS
The calculation formula of particle velocity and position update function are as follows:

­ N 
°9 L  
Z9 LN  FU 3L  ; LN  F  U ¨ 3J  ; LN ¸
§ ·
® © ¹ (1)
°; N   ; N
9 N 
¯ L L L

L is particle number, N represents iteration counter, Z represents inertial weight, F  and F  is acceleration
coefficients konwn as cognitive and social parameters, U and U is random values in [0,1].
In PSO, the constraint conditions of velocity and position are:

 vmax d vid d vmax , xmin d xid d xmax (2)

Where max v is the maximum velocity, which allows actually serves as a constraint that controls the maximum
global exploration ability PSO can have; min x and max x are the lower boundary and upper boundary of the
solution space.
In this paper a new fitness function is generated to get the desired result.

Green Energy and Sustainable Development I

AIP Conf. Proc. 1864, 020023-1–020023-4; doi: 10.1063/1.4992840
Published by AIP Publishing. 978-0-7354-1542-3/$30.00

020023-1
 *5  *2$/
) \ (3)
*5  2%67$&/(

Where F is fitness function. *5  *2$/ Is the distance between robot and goal position, *5  REVWDFOH is the distance
between robot and obstacle. By tuning the value of Z , you can get different possible optimal paths.
POS flowchart is as follows:

start

move to the destination

No

No
presence of obstacle destination reached

Yes Yes

end
Run pso

compute global best position

robot moves to global best

position

No
destination reached
Yes

end

FIGURE 1. Flowchart of the Algorithm

First,move the robot from source to the destination.Then,if the robot senses any obstacle in front of it,using
minimum fitness value criteria to find the global best position.In the end,the robot will move to these global best
positions until it reaches the destination.

PARAMETERS SELECTION OF THE SIMULATION

In order to verify the validity of the algorithm, simulations has been done between the standard particle swarm
optimization and the improved particle swarm optimization.All the parameters in this simulation are selected as
follows:

020023-2
 TABLE 1. Parameters selection

Parameter value Explanation

1 300 the size of the population
Q PD[ 1 the maximum velocity
Q PLQ 0 the minimum velocity
N PD[ 500 the maximum number of iterations
Z PD[ 0.95 the maximum inertia weight
Z PLQ 0.36 the minimum inertia weight
F DQG F  2 two positive constants
U DQG U  0.6 Two random functions
\ 0.46 positive constant

CONCLUSION
The simulation results between standard particle swarm optimization and improved particle swarm optimization
are as follows:
The start is at the bottom left corner,and the destination is at the upper right corner.

100 G

90
80
70
60
50
40
30
20
10
S
0
0 10 20 30 40 50 60 70 80 90 100

FIGURE 2. The standard particle swarm optimization path

As figure 2 shows the standard particle swarm optimization path are too close to obstacles.

020023-3
 100 G

90
80
70
60
50
40
30
20
10
S
0
0 10 20 30 40 50 60 70 80 90 100

FIGURE 3. The improved particle swarm optimization path

As figure 3 shows the improved particle swarm optimization. It is clearly that get the global optimal path length
shorter.

ACKNOWLEDGMENTS
This work was financially supported by Guangdong innovation carrier construction project fund. Project funds:
2013B09030008
Author:
YongWang (1968 - ),male,Ph.D., associate professor, research direction: Internation of Things, Unstructured
Information Processing and Intelligent Computing,Email:[email protected];
FengCai (1990 - ), male, Post-Graduate, research direction: Machine Learning and Intelligent Computing.
YingWang (1970 - ), female, undergraduate, Senior Engineer, research direction: Cloud Computing and
Knowledge Engineering

REFERENCES
1. WAN X F,HU W,FANG W Y,ZHENG B J. Research on path planning of robot based on impr-oved ant colony
algorithm[J]. Computer Engineering and Applications, 2014, 50(18):63-66.
2. DENG G F,ZHANG X P,LIU Y P. Ant colony optimization and particle swarm optimization for robot-path
planning in obstacle environment[J]. Control Theory and Applications,2009,26(8):879-883.
3. ZHANG W X,ZHANG X L,LI Y. Path planning for intelligent robots based on improved particle swarm
optimization algorithm [J]. Journal of Computer Applications,2014, 34(2):510-513.
4. WANG J,WU X X,GUO B L. Robot path planning using improved particle swarm optimization[J]. Computer
Engineering and Applications,2012, 48(15):240-244.
5. LU Y R,ZHAO Y,MENG Y Q,LIU J. Path planning in narrow space by impoved genetic algorithm[J].
Application Research of Computers, 2015, 32(2): 3-418.

020023-4