Ant Colony Optimization
November 16, 2016
Since I created ant colony optimization algorithm this time, I will introduce it.
Here is my program, please refer this if you want.
Ant Colony Optimization (GitHub)
What is ant colony optimization?
Ant colony optimization is a probabilistic solution to solve like the shortest path problem,
imitates the ant behavior in finding a route to food.
Please see here for details of the algorithm.
Program description
In this program, we can find an approximate solution of the shortest path problem as follows.
- Start from s and go through all nodes
- The number on the path represents the distance between the nodes
In this example, when passing in the order of a, d, b, c, the distance becomes the shortest 8.1.
How to use this
Usage is as follows.
const AntColony = require("./ant-colony")
const NOA = 10
const LIMIT = 100
const Q = 3
const RHO = 0.8
const EPSILON = 0.15
const distance = [
// s a b c d
[0.0, 4.0, 4.0, 4.7, 5.0], // start
[0.0, 0.0, 1.2, 2.5, 1.7], // a
[0.0, 1.2, 0.0, 1.0, 1.4], // b
[0.0, 2.5, 1.0, 0.0, 2.0], // c
[0.0, 1.7, 1.4, 2.0, 0.0], // d
]
const ac = new AntColony(NOA, LIMIT, Q, RHO, EPSILON, distance)
ac.solve()
console.log(ac.result())
Variable | Meaning |
---|---|
NOA |
Number of ants |
LIMIT |
Maximum number of attempts |
Q |
Influence of pheromone |
RHO |
Evaporation rate of pheromone |
EPSILON |
The probability of not considering the pheromone concentration when an ant chooses the next node (epsilon-greedy method) |
distance |
Two-dimensional array of graph |