在用matlab的优化工具箱中的linprog求解器或者其他求解器求最优值时，怎么设置变量约束条件为整数

可以用分支定界法求解整数规划问题,给你源码：
function [x,fm] = IntProgFZ(f,A,b,Aeq,beq,lb,ub)
x = NaN;
fm = NaN;
NF_lb = zeros(size(lb));
NF_ub = zeros(size(ub));
NF_lb(:,1) = lb;
NF_ub(:,1) = ub;
F = inf;

while 1
sz = size(NF_lb);
k = sz(2);
opt = optimset('TolX',1e-9);
[xm,fv,exitflag] = linprog(f,A,b,Aeq,beq,NF_lb(:,1),NF_ub(:,1),[],opt);
if exitflag == -2
xm = NaN;
fv = NaN;
end
if xm == NaN
fv = inf;
end
if fv ~= inf
if fv < F
if max(abs(round(xm) - xm))<1.0e-7
F = fv;
x = xm;
tmpNF_lb = NF_lb(:,2:k);
tmpNF_ub = NF_ub(:,2:k);
NF_lb = tmpNF_lb;
NF_ub = tmpNF_ub;
if isempty(NF_lb) == 0
continue;
else
if x ~= NaN
fm = F;
return;
else
disp('不存在最优解!');
x = NaN;
fm = NaN;
return;
end
end
else
lb1 = NF_lb(:,1);
ub1 = NF_ub(:,1);
tmpNF_lb = NF_lb(:,2:k);
tmpNF_ub = NF_ub(:,2:k);
NF_lb = tmpNF_lb;
NF_ub = tmpNF_ub;
[bArr,index] = find(abs((xm - round(xm)))>=1.0e-7);
p = bArr(1);
new_lb = lb1;
new_ub = ub1;
new_lb(p) = max(floor(xm(p)) + 1,lb1(p));
new_ub(p) = min(floor(xm(p)),ub1(p));
NF_lb = [NF_lb new_lb lb1];
NF_ub = [NF_ub ub1 new_ub];
continue;
end
else
tmpNF_lb = NF_lb(:,2:k);
tmpNF_ub = NF_ub(:,2:k);
NF_lb = tmpNF_lb;
NF_ub = tmpNF_ub;
if isempty(NF_lb) == 0
continue;
else
if x ~= NaN
fm = F;
return;
else
disp('不存在最优解!');
x = NaN;
fm = NaN;
return;
end
end
end
else
tmpNF_lb = NF_lb(:,2:k);
tmpNF_ub = NF_ub(:,2:k);
NF_lb = tmpNF_lb;
NF_ub = tmpNF_ub;
if isempty(NF_lb) == 0
continue;
else
if x ~= NaN
fm = F;
return;
else
disp('不存在最优解!');
x = NaN;
fm = NaN;
return;
end
end
end
end