Application of Dijkstra algorithm in robot

path-planning
Langfang, China
e-mail:[email protected]
Huijuan Wang
North China Institute of Aerospace Engineering
Langfang, China Quanbo Yuan
e-mail: [email protected] North China Institute of Aerospace Engineering
Langfang, China
e-mail: [email protected]
Yuan Yu
Langfang Health School

Abstract—In this paper, the robot is a maze robot. Dijkstra map, but with the following map to make a better description
algorithm is used in the robot path planning. The shortest path is of algorithm.
selected in the process of barrier. Simulation results prove the
model valid; it can effectively solve the maze robot path planning.

Keywords-Robot; path-planning; Dijkstra algorithm;

I. INTRODUCTION
Robot path planning is an important area of research, it is
an important point between artificial intelligence and robotics.
In order to complete the task in a mobile security way, mobile
robot path planning is an important indicator of intelligence.
According to certain evaluation criteria, path planning is a
collision-free path that mobile robot will find and reach the
goal of a state from the initial obstacles environment,[1]. Based
on the situation of environment, path planning can be divided
into global path planning and local path planning. Between
them, the global path planning environment focus on the
information the situation as known, and local path planning
mainly refers to the unknown environmental information on the
overall situation.
It is an important problem that global path planning for
robot path planning in static environment. It has a wide range
of applications. Based on the results of previous studies in this
part, robot path planning applications makes an optimal path to
walk the maze by independence, and it is proved an effective
method with simulation experiments.
II. PROBLEM DESCRIPTION
A. Environment Model
In case, the robot's working space is a two-dimensional
structure of space, we use the same size space by
Figure 1. The map of a robot
two-dimensional grid, it is a standard that the grid size of the
obstacles which the robot can take a free walk in a grid .If the
In order to describe robot motion in the method of
robot can walk without any obstacles, the grid is a free grid,
Cartesian space Left corner of the lower map is as the origin of
and the grid is a barrier grid. In order to sense environmental
coordinates, the horizontal right direction is x-axis positive
information through the sensor, to establish an environmental
direction, the straight up direction is Y-axis positive direction,
model map, the robot is simplified as a point., obstacles is
each axis grid interval corresponds to a unit of length. Any one
simplified as a straight line, the thickness of line is no more
grid can be identified with Cartesian (x, y) uniquely. As is
than a grid of line, to ensure the barrier requirements. To
shown in Figure 2.
simplify the problem, the standard map is not used complex

This paper is supported by Langfang Scientific & Technical Research

Foundation under Grant No.2010011004.

978-1-4244-9439-2/11/$26.00 ©2011 IEEE

Authorized licensed use limited to: University of Leeds. Downloaded on1067
November 26,2023 at 17:34:25 UTC from IEEE Xplore. Restrictions apply.
 Figure 3. Intersect method

III. DIJKSTRA ALGORITHM

A. Dijkstra algorithm
Dijkstra's algorithm is described as follow: firstly, the
introduction of an auxiliary vector D, each of its component D
[i] that the length of the shortest path is found in the current
starting point from v to each end of vi. Its initial state is: if
there are arcs from v to vi, then D [i] is the weight of arc;
otherwise make D [i] is ’. Clearly, the length is:
M in
D[j]= i { D[i] | vięV}
This path is a shortest path from v, This path is (v, vj).
Figure 2. The workspace of a robot Based on the above analysis, the algorithm can be described
as follows:
B. Determining the current obstacles
The robot will bypass the obstacles in above model • Suppose the adjacency matrix arcs with weights to
environment as described above, the robot must determine represent directed graph, arcs [i] [j] represents the
whether to bypass certain obstacles at first. An intersection weight of arc <vi, vj>. If < vi, vj > does not exist, then it
method is presented, the steps are as followings: sets arcs [i] [j] will be ’. S is found the shortest path
collection of the end of point from v, its initial state is
• Drawing a line from the point to the target, the line is l; empty. Then the initial value from the v to map
• Determining whether there are intersections between l remaining vi, may be the shortest possible is:
and obstacles˗ D[i]=arcs[ Locate Vex(G,v)[i] ] vięV
• If so, identifying the obstacles that intersected with l • Select vj, making
first, the obstacle is that the current robot will bypass. M in
Connecting the point with the two endpoints (vi,vi+1)of D[j]= i { D[i] | vięV-S}
the obstacle (vi and vi+1 are points that the robot path vj is the current obtained end point is the shortest path from
planning may go through); the v .Order
• If l has no intersection with the obstacles, the robot will S=S ∪ {j}
directly draw a straight line from the target point and
judgment is end; • Changes the shortest path length from v to a vertex vk.
If
• For vi and vi+1, they have respectively above four
operations. D[j]+ arcs[j][k]< D[k]
Changes D[k]:
D[k]=D[j]+arcs[j][k]
• Repeat two or more times. It is an increasing sequence
by the length of the shortest path from the v to the
remaining vertexes.
B. Using the Dijkstra algorithm for robot path planning
In the last section of the built environment model,
determining the need by using the intersection method to pass

Authorized licensed use limited to: University of Leeds. Downloaded on1068

November 26,2023 at 17:34:25 UTC from IEEE Xplore. Restrictions apply.
the obstacles, drawing the length of the barrier between the end To verify the feasibility of the method described, in this
points on the graph to get multiple graphs of vertex with paper there is a simulation experiment, as shown in Figure 4.
weighted, Figure 4 (the picture shows part of the matrix that v0 The workspace for path planning simulation results is shown in
to v3) is as shown. Adjacency matrix with weight by directed Figure 5. O is the starting point, and G is the target point. The
graph is: coarse straight lines are obstacles, and the fine lines is obtained
path with the simulation of the final path.
∞ 6.3 6.5 ∞
Finally, from the simulation results, the proposed planning
∞ ∞ ∞ 7.9 algorithm can be effectively to solve the robot path planning
problem which is correct and effective.
∞ ∞ ∞ 2.7
V. SUMMARY
∞ ∞ ∞ ∞
This paper presents a solving the shortest path planning
Figure 4. the matrix of digraph methods of maze robot path planning, and algorithms
Simulation results indicate that the application of the proposed
Dijkstra algorithm is implemented in this matrix, it is path planning algorithm , Dijkstra algorithm, can be correct and
obtained the shortest path from the starting point of to the effective in mobile robot path planning.
target.
REFERENCES
IV. SIMULATION RESULTS
[1] ‫ޝ‬Introduction to Robotics‫ޞ‬Jiang Xinsong
[2] ‫ޝ‬Data Structure‫ޞ‬Yan Weimin㧘Wu Weimin㧘 187-190
[3] WANGC 㧘SOHYC㧘WANGH ,et al .A hierarchical enetic algorithm
for path planning ina static environme nt with obstacles[ A].I EEE
Canadian Conference on Electrical and Computer Engineering [C
].Canada :IE EE, 20 02.(3 ):1652 ৻ 1657.
[4] GOLDBERG D E .Genetic Algorithm in Search㧘OptimAssion and
Machine Learning[M].New York: Assion ৻ Wesley Publishing
Company㧘1989.
[5] ‫ޝ‬Robot control system design and simulation matlab‫ ޞ‬Liu Jinkun
[6] NOBORU N,HIDEO T. Path planning of agricultural mobile robot by
neural network and genetic algorithm [J]. Computers and Electronics in
Agriculture,1997,(18):187—204.
[7] WOONGGIE H, SEUNGMIN B, TAEYONG K. Genetic algorithm
based path planning and dynamic obstacle avoidance of mobile
robots[A]. IEEE International Conference on Computational Cybernetics
and Simulation[C].Orlando:IEEE,1997.(3):2747—27

Figure 5. Simulation results of the path planning for Figure 2

Authorized licensed use limited to: University of Leeds. Downloaded on1069

November 26,2023 at 17:34:25 UTC from IEEE Xplore. Restrictions apply.