Scilab平台粒子群算法实现：优化函数搜索

function [f_opt,x_opt]=PSO(fun,dim,itval_min,itval_max,num,typ,time,v_step,c0_strth,c1_strth,c2_strth,figon)
//***********************************************
//PSO algorithm for finding the maximal/minimal value of a function defined on n dimention space
//input parameter
//           fun: <function>  optimization function which should be in the form of [y]=fun(x)
//                            x=[x(1) x(2) x(3) ... ect.] where x(n) is the nth variation
//           dim: <integer>   dimention of the assemble of definition of the function
//     itval_min:  <array>    the minimum boundaries of each dimention
//     itval_max:  <array>    the maximum boundaries of each dimention
//         [num]: <integer>   the number of particals
//         [typ]:    <1>      find the maximal value of the function
//                 <other>    find the minimal value of the function
//      [v_step]:  <float>    v_step times defaut reference velociy depend on the boundary of each dimention
//                 <array>    reference velocity for each dimention
//    [c0_strth]:  <[0,1]>    the strength parameter for the present velocity
//    [c1_strth]:  <[0,1]>    the strength parameter for the local attractors
//    [c2_strth]:  <[0,1]>    the strength parameter for the global attractor
//       [figon]:    <1>      show figure 
//output parameter
//         x_opt:  <array>    the coordinates of the max/min value of fun
//         f_opt:  <float>    the value of fun(x)
//
//exemple
//  deff('[y]=fun(x)','y=100*(x(:,1)-x(:,2).^2)^2+(x(:,2)-1)^2');
//  dim=2;
//  ivmax=[5,5];
//  ivmin=[-5,-5];
//  [x]=PSO(fun,dim,ivmin,ivmax,20,0,500,1,0.99,0.8,0.8,1)
//  [x]=PSO(fun,dim,ivmin,ivmax,20,0,500) //for fast
//