Solver stops, but should not

Hey,

in the following code I represent 3 houses with 3 energy storages and they can share energy with another. The goal is to maximize the energy in each house, but they should avoid going in to the negative direction. When I give the houses lowerbounds of 0 and the solution cant avoid being 0 for one house I dont receive a solution. How should I tell Matlab that I want to maximize each house seperatly ?

Edit: I would like to use "energyprob.Objective = H3 * H2 * H1;" ,but I read that matlab cant do this. Is that true?

Thank you so much!

energyprob = optimproblem;

% 3 Houses 3 Storages one day

% H1 = optimvar('H1','LowerBound',0,'UpperBound',10);
% H2 = optimvar('H2','LowerBound',0,'UpperBound',10);
% H3 = optimvar('H3','LowerBound',0,'UpperBound',10);

H1 = optimvar('H1','UpperBound',10);
H2 = optimvar('H2','UpperBound',10);
H3 = optimvar('H3','UpperBound',10);

E1 = optimvar('E1','LowerBound',0,'UpperBound',10);
E2 = optimvar('E2','LowerBound',0,'UpperBound',10);
E3 = optimvar('E3','LowerBound',0,'UpperBound',10);

E_in_1 = optimvar('E_in_1','LowerBound',0,'UpperBound',10);
E_in_2 = optimvar('E_in_2','LowerBound',0,'UpperBound',10);
E_in_3 = optimvar('E_in_3','LowerBound',0,'UpperBound',10);

E_out_1 = optimvar('E_out_1','LowerBound',0,'UpperBound',10);
E_out_2 = optimvar('E_out_2','LowerBound',0,'UpperBound',10);
E_out_3 = optimvar('E_out_3','LowerBound',0,'UpperBound',10);

Anz_E = 3;
PV = 1; % energy from the sun
E_start = 4;
E_start_ges = Anz_E*(PV + E_start); % 15

Bedarf1 = 3; % Energy Demand
Bedarf2 = 3; % Energy Demand
Bedarf3 = 6; % Energy Demand

energyprob.Constraints.start1 = E1 == E_in_1 - E_out_1 + E_out_2 + E_out_3;
energyprob.Constraints.start2 = E2 == E_in_2 - E_out_2 + E_out_1 + E_out_3;
energyprob.Constraints.start3 = E3 == E_in_3 - E_out_3 + E_out_2 + E_out_1;

energyprob.Constraints.start4 = E3 + E1 + E2 == E_start_ges;
energyprob.Constraints.start6 = E_out_1 + E_out_2 + E_out_3 == E_in_1 + E_in_2 + E_in_3;

energyprob.Constraints.HE11 = H1 >= 0;
energyprob.Constraints.HE21 = H2 >= 0;
energyprob.Constraints.HE31 = H3 >= 0;

energyprob.Constraints.HE1 = H1 == E1 - Bedarf1;
energyprob.Constraints.HE2 = H2 == E2 - Bedarf2;
energyprob.Constraints.HE3 = H3 == E3 - Bedarf3;

energyprob.Objective = H3 + H2 + H1;
% energyprob.Objective = H3

energyprob.ObjectiveSense = 'maximize';

[sol,fval] = solve(energyprob)

[Walter Roberson]

multiplying variables is nonlinear and Problem Based Optimization can only handle linear and some quadratic .

[Tim]

thank you! I will try another solver then. Should I use Fmincon ?

[Walter Roberson]

fmincon is one possibility. Note that it is a local minimizer, not a global minimizer.