【BP预测】基于人工蜂群算法优化BP神经网络实现数据预测附matlab代码

1 简介

针对BP网络水质评价模型的不足,引入人工蜂群(ABC)算法,将求解BP神经网络各层权值、阀值的过程转化为蜜蜂寻找最佳蜜源的过程,提出了一种新的结合人工蜂群算法的BP网络水质评价方法(ABC-BP)。并以2000—2006年渭河监测断面的10组实测数据作为测试样本对其水质进行了评价,实验结果表明该方法得到的水质评价结果准确,并具有很强的稳定性和鲁棒性。

2 部分代码

登录后复制

          
          

           
           function [bestsol,yy] = ABC(prob,lb,ub,Np,T,limit)
           
           
%% Starting of ABC
           
           
f = NaN(Np,1);                      % Vector to store the objective function value of the population members
           
           
fit = NaN(Np,1);                    % Vector to store the fitness function value of the population members
           
           
trial = NaN(Np,1);                  % Initialization of the trial vector
           
           
D = length(lb);                     % Determining the number of decision variables in the problem
           
           
P = repmat(lb,Np,1) + repmat((ub-lb),Np,1).*rand(Np,D);   % Generation of the initial population
           
           
for p = 1:Np
           
           
    f(p) = prob(P(p,:));            % Evaluating the objective function value
           
           
    fit(p) = CalFit(f(p));          % Evaluating the fitness function value
           
           
end
           
           
[bestobj, ind] = min(f);            % Determine and memorize the best objective value
           
           
bestsol = P(ind,:);                 % Determine and memorize the best solution
           
           
bestobj1=1000;
           
           
for t = 1:T
           
           
    %% Employed Bee Phase
           
           
    for i = 1:Np
           
           
        [trial,P,fit,f] = GenNewSol(prob, lb, ub, Np, i, P, fit, trial, f, D);
           
           
    end
           
           
    %% Onlooker Bee Phase
           
           
    % as per the code of the inventors available at https://abc.erciyes.edu.tr/
           
           
    % prob=(0.9.*Fitness./max(Fitness))+0.1;
           
           
    % MATLAB Code of the ABC algorithm version 2 has been released (14.12.2009) (more optimized coding)
           
           
    probability = 0.9 * (fit/max(fit)) + 0.1;
           
           
    m = 0; n = 1;
           
           
    while(m < Np)
           
           
        if(rand < probability(n))
           
           
            [trial,P,fit,f] = GenNewSol(prob, lb, ub, Np, n, P, fit, trial, f, D);
           
           
            m = m + 1;
           
           
        end
           
           
        n = mod(n,Np) + 1;
           
           
    end
           
           
    [bestobj,ind] = min([f;bestobj]);
           
           
    CombinedSol = [P;bestsol];
           
           
    bestsol = CombinedSol(ind,:);
           
           
    if bestobj<bestobj1
           
           
        yy(t)=bestobj;
           
           
    else
           
           
          yy(t)=bestobj1;
           
           
    end
           
           
    %% Scout Bee Phase
           
           
    [val,ind] = max(trial);
           
           
    if (val > limit)
           
           
        trial(ind) = 0;                     % Reset the trial value to zero
           
           
        P(ind,:) = lb + (ub-lb).*rand(1,D); % Generate a random solution
           
           
        f(ind) = prob(P(ind,:));            % Determine the objective function value of the newly generated solution
           
           
        fit(ind) = CalFit(f(ind));          % Determine the fitness function value of the newly generated solution
           
           
    end
           
           
end
           
           
[bestfitness,ind] = min([f;bestobj]);
           
           
CombinedSol = [P;bestsol];
           
           
bestsol = CombinedSol(ind,:);
          
          

• 1.
• 2.
• 3.
• 4.
• 5.
• 6.
• 7.
• 8.
• 9.
• 10.
• 11.
• 12.
• 13.
• 14.
• 15.
• 16.
• 17.
• 18.
• 19.
• 20.
• 21.
• 22.
• 23.
• 24.
• 25.
• 26.
• 27.
• 28.
• 29.
• 30.
• 31.
• 32.
• 33.
• 34.
• 35.
• 36.
• 37.
• 38.
• 39.
• 40.
• 41.
• 42.
• 43.
• 44.
• 45.
• 46.
• 47.
• 48.
• 49.
• 50.
• 51.
• 52.

3 仿真结果

​4 参考文献

[1]苏彩红, 向娜, 陈广义,等. 基于人工蜂群算法与BP神经网络的水质评价模型[J]. 环境工程学报, 2012.

博主简介：擅长智能优化算法、神经网络预测、信号处理、元胞自动机、图像处理、路径规划、无人机等多种领域的Matlab仿真，相关matlab代码问题可私信交流。

部分理论引用网络文献，若有侵权联系博主删除。