Change Value during optimization process

Hey,

is it possible to change a value during the optimization loop?

Example: I want to maximize x and y

lb = lowerbound ub = upperbound
    a = lb(0) ub(100) 
    b = lb(0) ub(50)
  x = a+b
  y = a+b

To archieve the solution I need the solution to be x = 50 + 25 and y = 50 + 25 how would i express this in code? Thanks!

example code I thought of but doesnt work

xyprob = optimproblem;

a = optimvar('a','LowerBound',0,'UpperBound',500);
b = optimvar('b','LowerBound',0,'UpperBound',100);

y = optimvar('y');
x = optimvar('x');

xyprob.ObjectiveSense = 'maximize';

xyprob.Objective = x
xyprob.Objective = y

xyprob.Constraints.econs1 = x == a+b
xyprob.Constraints.econs2 = y == a+b

[xysol,fval] = solve(xyprob);
tbl = struct2table(xysol)

[Matt J]

Example: I want to maximize x and y

Since x=y why don't we just say that we are maximizing x?

If x~=y, what does it mean to maximize two objectives at once in a linear program?

Why are x and y optimvars at all? Why wouldn't you just have

 a = optimvar('a','LowerBound',0,'UpperBound',500);
 b = optimvar('b','LowerBound',0,'UpperBound',100);

 xyprob.Objective = a+b ;

[Tim]

Hey thank you for your quick answer.

Yes thats right I need to maximize 2 or more variables at once. I am new to optimization with matlab so I dont know which solver I should choose. Maybe I need to go for fgoalattain ?

Here is a litte explanation for the problem.

a = storage filled with 100 kwh b = storage filled with 50 kwh x = house y = house

Explanation: x and y are houses, who need energy. They draw energy from the storages a and b. So if x takes 30 from a and 20 from b there is still 70 in a and 30 in b, but it should be equal.

The solution I want it to be is x = 50 + 25 and y = 50 + 25

Answer 1

[Matt J]

How about this?

xyprob = optimproblem;
xa = optimvar('xa','LowerBound',0,'UpperBound',100);
xb = optimvar('xb','LowerBound',0,'UpperBound',50);
ya = optimvar('ya','LowerBound',0,'UpperBound',100);
yb = optimvar('yb','LowerBound',0,'UpperBound',50);

xyprob.ObjectiveSense = 'maximize';
xyprob.Objective = xa+xb;

xyprob.Constraints.econs1 = xa+ya <= 100;
xyprob.Constraints.econs2 = xb+yb <= 50;
xyprob.Constraints.econs3 = xa+xb == ya+yb;
xyprob.Constraints.econs4 = xa+xb + ya+yb == 150;
[xysol,fval] = solve(xyprob);
tbl = struct2table(xysol)

[Tim]

wow you are a genius! Thx so much