According to difference of the robot know the
environment information extent, the path planning is
divided two types: the global path planning that the
environment information is known completely and the
local path planning that the environment information is
not known completely or partially. For the local path
planning, the location, shape and size of obstacle in the
working environment of robot can be detected on line by
sensors. We mainly study the problem of mobile robot
global path planning in this paper. Many scholars make
lots of researches on path planning and put forward some
methods, such as artificial potential field approach, the
free space method, neural networks, genetic algorithm,
simulated annealing algorithm, ant colony algorithm and
so on, but these methods exist definite limitation.
Although the ant colony algorithm has been used in
path planning only for few years, many fruit has been
obtained [2]. However, there are some problems on the
application of ant colony algorithm, such as the time for
solution construction in the nonage of algorithm is too
long, and the algorithm easily traps in local optimum, and
so on. Genetic algorithm is also a new optimum algorithm
developed fast recently [3], it has some advantages such as
the parallel search and the searching efficiency is higher.
In addition, it belongs to the random optimize process
essentially, so the local convergence question is not exist.
But the genetic algorithm also exist some shortage, it can
but search out the approximate to excellent solution that
near to global optimal solution in a short time.
Enlightened by the genetic algorithm, based on enhancing
pheromone concentration on optimal path, some more
optimal solutions will be crossed and mutated as genetic
algorithm, and then an improved augment ant colony
algorithm to overcome the defects is researched in this
paper.
According to difference of the robot know the
environment information extent, the path planning is
divided two types: the global path planning that the
environment information is known completely and the
local path planning that the environment information is
not known completely or partially. For the local path
planning, the location, shape and size of obstacle in the
working environment of robot can be detected on line by
sensors. We mainly study the problem of mobile robot
global path planning in this paper. Many scholars make
lots of researches on path planning and put forward some
methods, such as artificial potential field approach, the
free space method, neural networks, genetic algorithm,
simulated annealing algorithm, ant colony algorithm and
so on, but these methods exist definite limitation.
Although the ant colony algorithm has been used in
path planning only for few years, many fruit has been
obtained [2]. However, there are some problems on the
application of ant colony algorithm, such as the time for
solution construction in the nonage of algorithm is too
long, and the algorithm easily traps in local optimum, and
so on. Genetic algorithm is also a new optimum algorithm
developed fast recently [3], it has some advantages such as
the parallel search and the searching efficiency is higher.
In addition, it belongs to the random optimize process
essentially, so the local convergence question is not exist.
But the genetic algorithm also exist some shortage, it can
but search out the approximate to excellent solution that
near to global optimal solution in a short time.
Enlightened by the genetic algorithm, based on enhancing
pheromone concentration on optimal path, some more
optimal solutions will be crossed and mutated as genetic
algorithm, and then an improved augment ant colony
algorithm to overcome the defects is researched in this
paper.
การแปล กรุณารอสักครู่..
