(9)
The robot controller in the constrained environment can be developed based on the
new virtual force.
4 Simulation Results
To test the effectiveness of the proposed distributed model and the algorithms, simulations
in various situations have been conducted. This section discusses the simulation
results of the proposed algorithms.
76 Jindong Tan
4.1 Autonomous deployment in open space
A group of seven mobile robots are randomly deployed in an open space. Figure
9 (a) shows the initial configuration of the system. Applying the distributed robot
control in eq. (4) to each robot in the group, the system autonomously deploys itself.
It is worth noting that the shape of the Delaunay triangulation changes during the
deployment process. Some nodes gain neighbors and some lose neighbors.
(a) Initial deployment (b) Final deployment
Fig. 9. Self-Deployment of seven randomly deployed mobile robots.
In all these experiments, ki = 0.25, ci j = 40 in the simulation shown in Figure
9. The value of ci j can be changed to congregate or disperse the mobile sensor
network. The convergence of the controllers in eq. (4) is also tested using a large
number of robot nodes. Figure 11 shows the initial deployment and the configuration
after the self-deployment. It is seen that the area of coverage by the system is
greatly enlarged. The performance of the coverage under the distributed control law
is evaluated using a centroidal Voronoi diagram.
Pages:
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149