ACO refactor
This commit is contained in:
parent
a419237226
commit
9758e2ce3c
|
@ -18,8 +18,8 @@ public class AntColonyOptimization {
|
||||||
|
|
||||||
private int maxIterations = 1000;
|
private int maxIterations = 1000;
|
||||||
|
|
||||||
public int numberOfCities;
|
private int numberOfCities;
|
||||||
public int numberOfAnts;
|
private int numberOfAnts;
|
||||||
private double graph[][];
|
private double graph[][];
|
||||||
private double trails[][];
|
private double trails[][];
|
||||||
private List<Ant> ants = new ArrayList<>();
|
private List<Ant> ants = new ArrayList<>();
|
||||||
|
@ -28,8 +28,8 @@ public class AntColonyOptimization {
|
||||||
|
|
||||||
private int currentIndex;
|
private int currentIndex;
|
||||||
|
|
||||||
public int[] bestTourOrder;
|
private int[] bestTourOrder;
|
||||||
public double bestTourLength;
|
private double bestTourLength;
|
||||||
|
|
||||||
public AntColonyOptimization(int noOfCities) {
|
public AntColonyOptimization(int noOfCities) {
|
||||||
graph = generateRandomMatrix(noOfCities);
|
graph = generateRandomMatrix(noOfCities);
|
||||||
|
@ -43,26 +43,17 @@ public class AntColonyOptimization {
|
||||||
|
|
||||||
/**
|
/**
|
||||||
* Generate initial solution
|
* Generate initial solution
|
||||||
*
|
|
||||||
* @param n
|
|
||||||
* @return
|
|
||||||
*/
|
*/
|
||||||
public double[][] generateRandomMatrix(int n) {
|
public double[][] generateRandomMatrix(int n) {
|
||||||
double[][] randomMatrix = new double[n][n];
|
double[][] randomMatrix = new double[n][n];
|
||||||
random.setSeed(System.currentTimeMillis());
|
IntStream.range(0, n)
|
||||||
IntStream.range(0, n).forEach(i -> {
|
.forEach(i -> IntStream.range(0, n)
|
||||||
IntStream.range(0, n).forEach(j -> {
|
.forEach(j -> randomMatrix[i][j] = Math.abs(random.nextInt(100) + 1)));
|
||||||
Integer r = random.nextInt(100) + 1;
|
|
||||||
randomMatrix[i][j] = Math.abs(r);
|
|
||||||
});
|
|
||||||
});
|
|
||||||
return randomMatrix;
|
return randomMatrix;
|
||||||
}
|
}
|
||||||
|
|
||||||
/**
|
/**
|
||||||
* Perform ant optimization
|
* Perform ant optimization
|
||||||
*
|
|
||||||
* @return
|
|
||||||
*/
|
*/
|
||||||
public void startAntOptimization() {
|
public void startAntOptimization() {
|
||||||
IntStream.rangeClosed(1, 3).forEach(i -> {
|
IntStream.rangeClosed(1, 3).forEach(i -> {
|
||||||
|
@ -73,8 +64,6 @@ public class AntColonyOptimization {
|
||||||
|
|
||||||
/**
|
/**
|
||||||
* Use this method to run the main logic
|
* Use this method to run the main logic
|
||||||
*
|
|
||||||
* @return
|
|
||||||
*/
|
*/
|
||||||
public int[] solve() {
|
public int[] solve() {
|
||||||
setupAnts();
|
setupAnts();
|
||||||
|
@ -94,7 +83,7 @@ public class AntColonyOptimization {
|
||||||
*/
|
*/
|
||||||
private void setupAnts() {
|
private void setupAnts() {
|
||||||
IntStream.range(0, numberOfAnts).forEach(i -> {
|
IntStream.range(0, numberOfAnts).forEach(i -> {
|
||||||
ants.stream().forEach(ant -> {
|
ants.forEach(ant -> {
|
||||||
ant.clear();
|
ant.clear();
|
||||||
ant.visitCity(-1, random.nextInt(numberOfCities));
|
ant.visitCity(-1, random.nextInt(numberOfCities));
|
||||||
});
|
});
|
||||||
|
@ -107,23 +96,18 @@ public class AntColonyOptimization {
|
||||||
*/
|
*/
|
||||||
private void moveAnts() {
|
private void moveAnts() {
|
||||||
IntStream.range(currentIndex, numberOfCities - 1).forEach(i -> {
|
IntStream.range(currentIndex, numberOfCities - 1).forEach(i -> {
|
||||||
ants.stream().forEach(ant -> {
|
ants.forEach(ant -> ant.visitCity(currentIndex, selectNextCity(ant)));
|
||||||
ant.visitCity(currentIndex, selectNextCity(ant));
|
|
||||||
});
|
|
||||||
currentIndex++;
|
currentIndex++;
|
||||||
});
|
});
|
||||||
}
|
}
|
||||||
|
|
||||||
/**
|
/**
|
||||||
* Select next city for each ant
|
* Select next city for each ant
|
||||||
*
|
|
||||||
* @param ant
|
|
||||||
* @return
|
|
||||||
*/
|
*/
|
||||||
private int selectNextCity(Ant ant) {
|
private int selectNextCity(Ant ant) {
|
||||||
int t = random.nextInt(numberOfCities - currentIndex);
|
int t = random.nextInt(numberOfCities - currentIndex);
|
||||||
if (random.nextDouble() < randomFactor) {
|
if (random.nextDouble() < randomFactor) {
|
||||||
IntStream.range(0, numberOfCities).filter(i -> i == t && !ant.visited(i)).findFirst();
|
IntStream.range(0, numberOfCities).filter(i -> i == t && !ant.visited(i)).findFirst(); //TODO unused
|
||||||
}
|
}
|
||||||
calculateProbabilities(ant);
|
calculateProbabilities(ant);
|
||||||
double r = random.nextDouble();
|
double r = random.nextDouble();
|
||||||
|
@ -140,8 +124,6 @@ public class AntColonyOptimization {
|
||||||
|
|
||||||
/**
|
/**
|
||||||
* Calculate the next city picks probabilites
|
* Calculate the next city picks probabilites
|
||||||
*
|
|
||||||
* @param ant
|
|
||||||
*/
|
*/
|
||||||
public void calculateProbabilities(Ant ant) {
|
public void calculateProbabilities(Ant ant) {
|
||||||
int i = ant.trail[currentIndex];
|
int i = ant.trail[currentIndex];
|
||||||
|
|
Loading…
Reference in New Issue