^{1}

^{*}

^{1}

Nowadays, path planning has become an important field of research focus. Considering that the ant colony algorithm has numerous advantages such as the distributed computing and the characteristics of heuristic search, how to combine the algorithm with two-dimension path planning effectively is much important. In this paper, an improved ant colony algorithm is used in resolving this path planning problem, which can improve convergence rate by using this improved algorithm. MAKLINK graph is adopted to establish the two-dimensional space model at first, after that the Dijkstra algorithm is selected as the initial planning algorithm to get an initial path, immediately following, optimizing the select parameters relating on the ant colony algorithm and its improved algorithm. After making the initial parameter, the authors plan out an optimal path from start to finish in a known environment through ant colony algorithm and its improved algorithm. Finally, Matlab is applied as software tool for coding and simulation validation. Numerical experiments show that the improved algorithm can play a more appropriate path planning than the origin algorithm in the completely observable.

The essence of road traffic navigation problem [

Ant colony algorithm [

The basic step in solving the optimization problem applying ant colony algorithm is: the ant walk paths are delegated with feasible solutions of optimization problem, the entire ant group staying on all paths constitutes the solution space of optimization problems. Released pheromone amounts in short path is more than others, with the passage of time, the accumulation of pheromone concentration on the shorter path is increasing gradually, more and more ants will choose these paths. Ultimately, the ants will focus on the best path under the action of the positive feedback.

Ant colony algorithm has played an important role in solving the traveling salesman problem [

Without loss of generality, supposing that the number for a whole ants group are m, network topology for the road node numbers are n, the distance between the road node i and j is

The

In the formula,

This paper will introduce MAKLINK graph theory, a two-dimension path planning model is established to initialize the network topology, two-dimension path planning feasible region is generated by several MAKLINK lines in that MAKLINK graph. Among that, MAKLINK lines show not only that connection lines between two adjacent obstacles vertex where connection lines are not intersecting but also the lines between obstacles vertex and its boundary.

A suboptimal routing road

Of which,

The Formula (4) shows that Dijkstra algorithm will get the initial path through the link line, as long as a new set of parameters

It is required to disperse work space before using ant colony algorithm. Due to the initialization of the freedom of choice link line length is differ, the method of the link line division is named fixed distance classification method. Setting the length is

If the

Ant colony algorithm is used to obtain the path parameter set

i is the whole points in link lines, q is a random number between [0,1],

As for the calculation method of j, firstly calculating the select probability from the link line node

Pheromone update includes real-time pheromone update and path pheromone update. Over here real-time pheromone update demands each ant must update the node pheromone after selecting a node, namely:

It is completed the iterative search when all the ants move from the initial point to the finish line. In order to select a shortest path length when all the ants move to the end, pheromone on each road should be updated, the formula is:

When an ant is searching for the node

To explore the ant number influence on the optimal path, we had analyzed six different sets of ant species, the

The

The

The

The optimum calculation parameters can be chosen in this paper, as a whole, the number of ants is 15, pheromone importance factor is a = 1, importance factor inspired by the function is 2, pheromone volatilization factor is 0.1, the number of iterations is 500.

In this paper, an improved ant colony algorithm is applied for two-dimension path planning; through the algorithm we can plan out optimal paths from start to end. Comparing the improved algorithm to the original ant colony algorithm, comparison of the simulation result shows that improved algorithm can plan a more optimal routing than the original algorithm.

However the simulation is based on the environment such as the known obstacle conditions, path planning in the actual situation maybe just be part of the perception of environmental conditions. How to plan out an optimal path from start to end under the condition of local environment perception is the requirement to be further research content.

RongWang,HongJiang, (2015) Two-Dimension Path Planning Method Based on Improved Ant Colony Algorithm. Advances in Pure Mathematics,05,571-578. doi: 10.4236/apm.2015.59053