基于蚁群算法的TSP问题matlab实现

function [bestsol,bestcost] = ACO(parameters) %Parameters: nDim = parameters.nDim; MaxIt = parameters.MaxIt; % Maximum Number of Iterations nAnt = parameters.nAnt; % Number of Ants (Population Size) Q = parameters.Q; alpha = parameters.alpha; % Phromone Exponential Weight beta = parameters.beta; % Heuristic Exponential Weight rho = parameters.rho; % Evaporation Rate dispflag = parameters.dispflag; showflag = parameters.showflag; Objunchange = parameters.Objunchange; % Problem definition model = CreateTspModel(nDim); loss = @(path) PathLength(path, model); % Initialization tau0 = 1/(nDim*nDim); % Initial Phromone tau = tau0*ones(nDim,nDim); % Phromone Matrix eta = 1./(model.d+eps); % Heuristic Information Matrix bestcost = zeros(MaxIt,1); % Array to Hold Best Cost Values empty_ant.path = []; % Empty Ant empty_ant.cost = []; ant = repmat(empty_ant,nAnt,1);% Ant Colony Matrix bestsol.cost = inf; iter = 1; unchange = 0; while iter<=MaxIt && unchange < Objunchange % Update Ants objUpdate = 0; for s = 1:nAnt ant(s).path = randi([1 nDim]); for t = 2:nDim i = ant(s).path(end); P = (tau(i,:).^alpha).*(eta(i,:).^beta); P(ant(s).path) = 0; P = cumsum(P/sum(P)); j = find(rand<=P,1,'first'); ant(s).path = [ant(s).path j]; end ant(s).cost=loss(ant(s).path); if ant(s).cost < bestsol.cost bestsol = ant(s); objUpdate = objUpdate + 1; end end % Update Phromones for s=1:nAnt path = ant(s).path; path = [path path(1)]; %#ok for t = 1:nDim i = path(t); j = path(t+1); tau(i,j) = tau(i,j) + Q/ant(s).cost; % Enhancement end end % Evaporation tau = (1-rho)*tau; % Store Best Cost bestcost(iter)=bestsol.cost; % Show Iteration Information if dispflag disp(['Iteration ' num2str(iter) ': Best Cost = ' num2str(bestcost(iter))]); end % Plot Solution if showflag figure(1); PlotSolution(bestsol,model,iter,bestsol.cost); pause(0.01); end if objUpdate==0 unchange = unchange + 1; else unchange = 0; end iter = iter + 1; end