Dynamic-Calibration/utils/YALMIP-master/modules/global/propagatequadratics.m

144 lines
5.9 KiB
Matlab
Executable File

function pout = propagatequadratics(p)
pout = p;
if p.bilinears~=0
F_struc = p.F_struc;
p.F_struc = [-p.F_struc(1:p.K.f,:);p.F_struc];
p.K.f=2*p.K.f;
InequalityConstraintState = [p.EqualityConstraintState;p.EqualityConstraintState;p.InequalityConstraintState];
if p.K.l+p.K.f>0
quadratic_variables = find(p.bilinears(:,2) == p.bilinears(:,3));
if ~isempty(quadratic_variables)
quadratic_variables = p.bilinears(quadratic_variables,1);
for i = 1:length(quadratic_variables)
k = quadratic_variables(i);
if (p.lb(k) < p.ub(k)-1e-4)
x = p.bilinears(p.bilinears(:,1)==k,2);% x^2
candidates = find((InequalityConstraintState==1) & p.F_struc(1:p.K.f+p.K.l,1+k))';
for j = candidates
a = p.F_struc(j,2:end);
b = p.F_struc(j,1);
aij = a(k);
% rest + cij*x + (+-)x^2 >= 0
cij = a(x);
cij = cij/abs(aij);
b = b/abs(aij);
a = a/abs(aij);
% aij = sign(aij);
a(k) = 0;
a(x) = 0;
indNEG = find(a < 0);
indPOS = find(a > 0);
LB = p.lb;
UB = p.ub;
LB(k) = 0;
UB(k) = 0;
LB(x) = 0;
UB(x) = 0;
a(k) = 0;
a(x) = 0;
if aij < 0
% rest + cij*x - x^2 >= 0
% Derive upper bound on rest
rest = (b+a([indPOS(:);indNEG(:)])*[UB(indPOS);LB(indNEG)]);
% Write as rest >= (x - center)^2-(center)^2
center = cij/2;
radii2 = rest + (cij/2)^2;
newUB = center + sqrt(radii2);
newLB = center - sqrt(radii2);
p.lb(x) = max(p.lb(x),newLB);
p.ub(x) = min(p.ub(x),newUB);
else
% rest + cij*x + x^2 >= 0
% (x+cij/2)^2 - (cij/2)^2 + rest >= 0
% (x+cij/2)^2 >= radii
% Lower bound on rest
rest = (b+a([indPOS(:);indNEG(:)])*[UB(indPOS);LB(indNEG)]);
center = -cij/2;
radii2 = -rest + (cij/2)^2;
if radii2 > 0
left = center-sqrt(radii2);
right = center+sqrt(radii2);
if p.ub(x) < right
p.ub(x) = min(p.ub(x),left);
end
if p.lb(x) > left
p.lb(x) = max(p.lb(x),right);
end
end
end
end
elseif p.lb(k)<0
end
end
end
end
if p.K.l+p.K.f>0
bilinear_variables = find(p.bilinears(:,2) ~= p.bilinears(:,3));
if ~isempty(bilinear_variables)
bilinear_variables = p.bilinears(bilinear_variables,1);
for i = 1:length(bilinear_variables)
k = bilinear_variables(i);
if p.lb(k) >= -5000000000 & (p.lb(k) < p.ub(k)-1e-4)
x = p.bilinears(p.bilinears(:,1)==k,2);% x^2
y = p.bilinears(p.bilinears(:,1)==k,3);% x^2
candidates = find((InequalityConstraintState==1) & p.F_struc(1:p.K.f+p.K.l,1+k))';
for j = candidates
a = p.F_struc(j,2:end);
aij = a(k);
if aij > 0
indNEG = find(a < 0);
indPOS = find(a > 0);
LB = p.lb;
UB = p.ub;
LB(k) = 0;
UB(k) = 0;
a(k) = 0;
newLB = (-p.F_struc(j,1)-a([indPOS(:);indNEG(:)])*[UB(indPOS);LB(indNEG)])/aij;
p.lb(k) = max(p.lb(k),newLB);
elseif aij < 0
indNEG = find(a < 0);
indPOS = find(a > 0);
LB = p.lb;
UB = p.ub;
LB(k) = 0;
UB(k) = 0;
a(k) = 0;
newUB = (p.F_struc(j,1)+a([indPOS(:);indNEG(:)])*[UB(indPOS);LB(indNEG)])/(-aij);
p.ub(k) = min(p.ub(k),newUB);
end
end
if p.lb(k)>0 & p.lb(x)>0 & p.lb(y)>0
p.ub(x) = min(p.ub(x), p.ub(k)/p.lb(y));
p.ub(y) = min(p.ub(y), p.ub(k)/p.lb(x));
p.lb(x) = max(p.lb(x), p.lb(k)/p.ub(y));
p.lb(y) = max(p.lb(y), p.lb(k)/p.ub(x));
end
elseif p.lb(k)<0
end
end
end
end
end
if ~isequal([p.lb p.ub],[pout.lb pout.ub])
pout.changedbounds = 1;
end
pout.lb = p.lb;
pout.ub = p.ub;