test_cost_func2.m

模拟退化算法在C环境下的实现

function [cost_value, cost_flag, user_acceptance_flag] = ...
  test_cost_func2 (x, critical_cost_value, no_test_flag);
%
% $Id: test_cost_func2.m,v 1.24 2002/12/25 20:15:06 ssakata Exp ssakata $
%
% a_i = parameter_upper_bound[i] 
s_i = 0.2;
t_i = 0.05;
c_r = 0.15;

q_n = 0.0;
parameter_dimension = 4;

for i = 1 : parameter_dimension,
  j = mod(i-1,4);
  switch (j)
     case 0, d_i = 1.0;
     case 1, d_i = 1000.0;
     case 2, d_i = 10.0;
     otherwise, d_i = 100.0;
  end
  if x(i) > 0.0
    k_i = fix(x(i) / s_i + 0.5);
  elseif x(i) < 0.0
    k_i = fix(x(i) / s_i - 0.5);
  else
    k_i = 0;
  end

  if abs (k_i * s_i - x(i)) < t_i
    if k_i < 0
      z_i = k_i * s_i + t_i;
    elseif k_i > 0
      z_i = k_i * s_i - t_i;
    else
      z_i = 0.0;
    end
    q_n = q_n + c_r * d_i * z_i * z_i;
  else
    q_n = q_n + d_i * x(i) * x(i);
  end
end;  
cost_value = q_n;
cost_flag=1;
if no_test_flag == 0
  if cost_value <= critical_cost_value
    user_acceptance_flag = 1;
  else
    user_acceptance_flag = 0;
  end
end