混合蛙跳算法SFLA的matlab实现

function pop = RunFLA(pop, params) %% FLA Parameters q = params.q; % Number of Parents alpha = params.alpha; % Number of Offsprings beta = params.beta; % Maximum Number of Iterations sigma = params.sigma; CostFunction = params.CostFunction; VarMin = params.VarMin; VarMax = params.VarMax; VarSize = size(pop(1).Position); BestSol = params.BestSol; nPop = numel(pop); % Population Size P = 2*(nPop+1-(1:nPop))/(nPop*(nPop+1)); % Selection Probabilities % Calculate Population Range (Smallest Hypercube) LowerBound = pop(1).Position; UpperBound = pop(1).Position; for i = 2:nPop LowerBound = min(LowerBound, pop(i).Position); UpperBound = max(UpperBound, pop(i).Position); end %% FLA Main Loop for it = 1:beta % Select Parents L = RandSample(P,q); B = pop(L); % Generate Offsprings for k=1:alpha % Sort Population [B, SortOrder] = SortPopulation(B); L = L(SortOrder); % Flags ImprovementStep2 = false; Censorship = false; % Improvement Step 1 NewSol1 = B(end); Step = sigma*rand(VarSize).*(B(1).Position-B(end).Position); NewSol1.Position = B(end).Position + Step; if IsInRange(NewSol1.Position, VarMin, VarMax) NewSol1.Cost = CostFunction(NewSol1.Position); if NewSol1.Cost<B(end).Cost B(end) = NewSol1; else ImprovementStep2 = true; end else ImprovementStep2 = true; end % Improvement Step 2 if ImprovementStep2 NewSol2 = B(end); Step = sigma*rand(VarSize).*(BestSol.Position-B(end).Position); NewSol2.Position = B(end).Position + Step; if IsInRange(NewSol2.Position, VarMin, VarMax) NewSol2.Cost = CostFunction(NewSol2.Position); if NewSol2.Cost<B(end).Cost B(end) = NewSol2; else Censorship = true; end else Censorship = true; end end % Censorship if Censorship B(end).Position = unifrnd(LowerBound, UpperBound); B(end).Cost = CostFunction(B(end).Position); end end % Return Back Subcomplex to Main Complex pop(L) = B; end end