你的问题究竟是什么?你看到是什么错误?我们不能测试有完整的代码,在这种情况下,函数initSwarm……
我发现最低的生产成本矩阵,利用PSO,下面的代码。现在我想做同样的算法。我非常感激如果有人帮我这样做
4视图(30天)
显示旧的评论
清晰的所有clc;密切所有;
一个= 360;
h = 10;
l = 40;
o = 20;
p = 1;
% %生产矩阵
D = [60 60 80 60 60 100 90 90 50 50
60 60 60 70 80 80 70 80 70 80
50 60 80 80 70 90 80 90 80
80 90 100 70 80 80 70 100 70 60
100 60 50 50 90 100 90 90 50 70];
(p t) = (D)大小;
迭代= 100;
人口= 50;
xmax = 1.25 * D;
xmin = 0.75 * D;
N =人口;
N_GER =迭代;
PHI1 = 1.5;
PHI2 = 1.5;
W = 1;
v = 0 (p, N);
X_MAX = xmax;
X_MIN = xmin;
vmin = ((xmax) - (xmin)) / (N * 5);
vmax = ((xmax) - (xmin)) / (N * 5);
gb = 0 (p、t);
gbestvalue = 10000000000 + 0 (p、t);
gaux = 1 (p、t、N);
xb = 0 (p, N);
fitBest = 0 (N, 1);
适合= 0 (N, 1);
ng = 1;
x = initSwarm (D, N, p, t X_MIN X_MAX);
为j = 1: N
(成本(j), Icost (j), Scost (j),
Ocost (j)] = Obj_func (D, x (:,:, j), A, h, l, o, p);
fitBest (j) =成本(j);
结束
[a, b] = min(成本);
gb = x (:,:, b);
gbestvalue =成本(b);
嘲笑= Icost (b);
’= Scost (b);
护镜= Ocost (b);
xb = x;
而(ng < = N_GER)
i = 1;
为k = 1: N
randnum1 =兰德([p、t]);
randnum2 =兰德([p、t]);
v (:,:, k) = w * v (:,:, k) + randnum1。* (PHI1。* (xb (:,: k) - x (:,:, k)))
+ randnum2。* (PHI2。* (gBest-x (:,:, k)));
为i = 1: p
为j = 1: t
v (i, j, k) = ((v (i, j, k) < = vmin (i, j))。* vmin) + (i, j)
((v (i, j, k) > vmin (i, j)。* v (i, j, k));
v (i, j, k) = ((v (i, j, k) > = vmax (i, j)。* vmax (i, j)) +
((v (i, j, k) < vmax (i, j))。* v (i, j, k));
结束
结束
x (:,:, k) =装天花板(x (:,:, k) + abs (v (:,:, k)));
为i = 1: p
为j = 1: t
如果x (i, j, k) < X_MIN (i, j)
x (i, j, k) =装天花板(X_MIN (i, j));
elseifx (i, j, k) > X_MAX (i, j)
x (i, j, k) =地板(X_MAX (i, j));
结束
结束
结束
x (:,:, k) = Repair_Strategy (D, x (:,:, k));
% (:,:,k) = x (:,:, k)
结束
而(我< = N)
如果(我= = N)
为j = 1: N
(成本(j), Icost (j), Scost (j),
Ocost (j)] = Obj_func (D, x (:,:, j), A, h, l, o, p);
适合(j) =成本(j);
如果适合(j) < fitBest (j)
fitBest (j) =适合(j);
xb (:,:, j) = x (:,:, j);
ib (j) = Icost (j);
某人(j) = Scost (j);
ob (j) = Ocost (j);
结束
结束
[a, b] = min(配合);
如果(适合(b) < gbestvalue)
gb = x (:,:, b);
gbestvalue =适合(b);
嘲笑= ib (b);
’=某人(b);
护镜= ob (b);
结束
结束
我=我+ 1;
结束
GenBest (ng) = gbestvalue;
ng = ng + 1;
结束
图
情节(GenBest)、宋惠乔网格
包含(“\ bfIteration #”)
ylabel (“\ bfF价值”)
gb = gb
gfunc = gbestvalue
gIcost =嘲笑
gScost =’
gOcost =采空区
函数(群)= initSwarm (D, N, p, t V_MIN V_MAX)
群= 0 (p, N);
为k = 1: N
为i = 1: p
为j = 1: t
临时(i, j, k) = abs(地板(兰德(1,1)* (V_MAX (i, j) -V_MIN (j)) + V_MIN (i, j)));
结束
结束
群(:,:,k) = Repair_Strategy (D,临时(:,:,k));
结束
函数(成本、Icost Scost Ocost] = Obj_func (D, Qr, A, h、l o, p)
(p t) = (D)大小;
为i = 1: p
我(1)= 0;
为j = 2: t
我(I, j) = (I, j - 1) + Qr (I, j) - d (I, j);
发票(i, j) = max (i (i, j), 0);
Shrt (i, j) = abs(最低(i (i, j), 0));
结束
结束
为j = 1: t
问(j) =总和(Qr (:, j));
SInv (j) =总和(发票(:,j));
Sshrt (j) =总和(Shrt (:, j));
Ocst (j) = max (q (j) *便士);
结束
Icost = h。*总和(SInv);
Scost = l。*总和(Sshrt);
Ocost = o。*总和(Ocst);
成本= Icost + Scost + Ocost;
函数城市= Repair_Strategy (D,问)
(p t) = (D)大小;
为i = 1: p
SumD (i) = (D(我,:))总和;
SumQ (i) = (Q(我,:))总和;
结束
DiffQD = SumQ-SumD;
德尔地板= (DiffQD / t);
为i = 1: p
为j = 1: t - 1
Qd (i, j) = Q (i, j)德尔(我);
结束
sumN (i) =总和(Qd(我,:));
Qd (t) = SumD (i) -sumN(我);
结束
城市= Qd;
