结合邻域连接法的蚁群优化（NACO）求解TSP问题（Matlab代码实现）

%% The Parameters
Iteration=1;                          % No. of Iterations
no_ant=1;                             % Population Size (No. of Ants)
Q=100;
%t0=1;
t0=10*Q/(no_Var*mean(D_Set.D(:)));      % Initial Phromone
alpha=1;                                % Phromone Exponential Weight
beta=5;                                 % Heuristic Exponential Weight
rate=0.7;                              % Evaporation Rate
%% Initialization
eta=1./D_Set.D;                         % Heuristic Information Matrix
t=t0*ones(no_Var,no_Var);               % Phromone Matrix
BC=zeros(Iteration,1);                  % Array to Hold Best Cost Values
% Empty Ant
empty_ant.Tour=[];
empty_ant.Cost=[];
% Ant Colony Matrix
ant=repmat(empty_ant,no_ant,1);
% Best Ant
BS.Cost=inf;
% the Level_one of NJ
Leaf=Leaf_1(D_Set.D);
% Runtime
T_1 = cputime;
%% NEW_ACO Main Program
for it=1:Iteration    
    for k=1:no_ant                      % Move Ants
        Leaf1=Leaf;
        ant(k).Tour=randi([1 no_Var]);  %% {choose a random non visited city}
        city=ant(k).Tour;
        for l =2:no_Var
            ant_tour=ant(k).Tour(end);          
            [row,col] = find(Leaf1==city);
            if isempty(row)
                P=t(ant_tour,:).^alpha.*eta(ant_tour,:).^beta;           
                P(ant(k).Tour)=0;
                P=P/sum(P);
                city=RouletteSelection(P);
                ant(k).Tour=[ant(k).Tour city];
            else
                l=l+1;                  % Pass the joining cities 
                if col==2
                    city=Leaf1(row,1);
               else
                    city=Leaf1(row,2);
                end
                ant(k).Tour=[ant(k).Tour city];
                Leaf1(row,1)=0;
                Leaf1(row,2)=0;
            end
          %  ant(k).Tour=[ant(k).Tour city];
        end
        ant(k).Cost=CostFunction(ant(k).Tour);
        if ant(k).Cost<BS.Cost
            BS=ant(k);
        end
    end
    % Update Phromones
    for k=1:no_ant
        tour=ant(k).Tour;
        tour=[tour tour(1)]; 
        for ll=1:no_Var
            ant_tour=tour(ll);
            city=tour(ll+1);
            t(ant_tour,city)=t(ant_tour,city)+Q/ant(k).Cost;
        end
    end
    % Evaporation
    t=(1-rate)*t;
    % Store Best Cost
    BC(it)=BS.Cost;