用蚁群算法解决旅行商问题c++

蚁群算法是一种模拟蚂蚁觅食行为的启发式算法，可以用于解决旅行商问题（TSP）。下面是一个用C++实现蚁群算法解决旅行商问题的示例代码：

#include <iostream>
#include <vector>
#include <cmath>
#include <cstdlib>
#include <ctime>

using namespace std;

const int MAX_CITIES = 10; // 城市数量
const int MAX_ANTS = 10; // 蚂蚁数量
const int MAX_ITERATIONS = 100; // 最大迭代次数

const double ALPHA = 1.0; // 信息素重要程度因子
const double BETA = 2.0; // 启发函数重要程度因子
const double RHO = 0.5; // 信息素蒸发系数
const double Q = 100.0; // 信息素增加强度系数

struct City {
    double x, y;
};

double distance(const City& city1, const City& city2) {
    double dx = city1.x - city2.x;
    double dy = city1.y - city2.y;
    return sqrt(dx * dx + dy * dy);
}

class Ant {
public:
    Ant() {
        tabu.resize(MAX_CITIES, false);
        path.resize(MAX_CITIES);
    }

    void clear() {
        for (int i = 0; i < MAX_CITIES; ++i) {
            tabu[i] = false;
            path[i] = 0;
        }
    }

    void visitCity(int city) {
        tabu[city] = true;
        path[currentCity] = city;
        currentCity = city;
        tourLength += distance(cities[path[currentCity]], cities[path[currentCity - 1]]);
    }

    int getCurrentCity() const {
        return currentCity;
    }

    double getTourLength() const {
        return tourLength;
    }

    void setCurrentCity(int city) {
        currentCity = city;
    }

private:
    vector<bool> tabu;
    vector<int> path;
    int currentCity = 0;
    double tourLength = 0.0;
};

class ACO {
public:
    ACO() {
        cities.resize(MAX_CITIES);
        ants.resize(MAX_ANTS);
        pheromone.resize(MAX_CITIES, vector<double>(MAX_CITIES, 1.0));

        // 初始化城市坐标
        for (int i = 0; i < MAX_CITIES; ++i) {
            cities[i].x = rand() % 100;
            cities[i].y = rand() % 100;
        }

        // 初始化蚂蚁
        for (int i = 0; i < MAX_ANTS; ++i) {
            ants[i].clear();
            ants[i].setCurrentCity(rand() % MAX_CITIES);
        }
    }

    void updatePheromone() {
        for (int i = 0; i < MAX_CITIES; ++i) {
            for (int j = 0; j < MAX_CITIES; ++j) {
                pheromone[i][j] *= (1.0 - RHO);
            }
        }

        for (int i = 0; i < MAX_ANTS; ++i) {
            for (int j = 0; j < MAX_CITIES; ++j) {
                int city1 = ants[i].path[j];
                int city2 = ants[i].path[(j + 1) % MAX_CITIES];
                pheromone[city1][city2] += Q / ants[i].getTourLength();
                pheromone[city2][city1] += Q / ants[i].getTourLength();
            }
        }
    }

    void antColonyOptimization() {
        for (int iteration = 0; iteration < MAX_ITERATIONS; ++iteration) {
            for (int i = 0; i < MAX_ANTS; ++i) {
                while (ants[i].getCurrentCity() != -1) {
                    int nextCity = selectNextCity(ants[i]);
                    ants[i].visitCity(nextCity);
                }

                if (ants[i].getTourLength() < bestTourLength) {
                    bestTourLength = ants[i].getTourLength();
                    bestTour = ants[i].path;
                }

                ants[i].clear();
                ants[i].setCurrentCity(rand() % MAX_CITIES);
            }

            updatePheromone();
        }
    }

    int selectNextCity(const Ant& ant) {
        int currentCity = ant.getCurrentCity();
        double sum = 0.0;

        for (int i = 0; i < MAX_CITIES; ++i) {
            if (!ant.tabu[i]) {
                sum += pow(pheromone[currentCity][i], ALPHA) * pow(1.0 / distance(cities[currentCity], cities[i]), BETA);
            }
        }

        double r = (double)rand() / RAND_MAX;
        double probability = 0.0;

        for (int i = 0; i < MAX_CITIES; ++i) {
            if (!ant.tabu[i]) {
                probability += pow(pheromone[currentCity][i], ALPHA) * pow(1.0 / distance(cities[currentCity], cities[i]), BETA) / sum;
                if (r <= probability) {
                    return i;
                }
            }
        }

        return -1;
    }

    void printBestTour() {
        cout << "Best tour: ";
        for (int i = 0; i < MAX_CITIES; ++i) {
            cout << bestTour[i] << " ";
        }
        cout << endl;
        cout << "Best tour length: " << bestTourLength << endl;
    }

private:
    vector<City> cities;
    vector<Ant> ants;
    vector<vector<double>> pheromone;
    vector<int> bestTour;
    double bestTourLength = numeric_limits<double>::max();
};

int main() {
    srand(time(nullptr));

    ACO aco;
    aco.antColonyOptimization();
    aco.printBestTour();

    return 0;
}

这段代码实现了蚁群算法解决旅行商问题。它使用了随机生成的城市坐标作为输入，通过迭代更新信息素矩阵和蚂蚁的路径来寻找最优的旅行路径。最终输出最优路径和路径长度。