Hi Antonio,
Just to clarify, the objective function value is not a price. Price is a
marginal, not total, value. The objective function is a total value/cost. In
MOST, the objective is essentially the sum of the costs (of energy, reserves,
etc) minus the value of dispatchable loads. In the case where all loads are
dispatchable, it is essentially the negative of the net benefits.
There should be *no* constraint violations if the optimization solves
successfully (i.e. mdo.results.success == 1).
Ray
> On Apr 24, 2019, at 10:20 AM, Antonio L'Abbate
> <[email protected]> wrote:
>
> Dear Professor Zimmerman,
> I would like to ask one more thing on this topic. I set the problem as you
> suggested, now I am analyzing the results: computed price is always negative
> (f = -13290302.0242). In order to compare what I found out, I’ve run the
> given example “most_ex6_uc”, getting a negative price again as objective of
> the simulation (f = -4987987.512). My query is therefore: is price supposed
> to be negative as a consequence of the problem formulation or the reason lies
> within the violation of a constraint?
> Ps I reported the prices just to give an idea of the order of magnitude I am
> obtaining.
> Thanks in advance,
> Best wishes,
> Antonio L’Abbate.
>
> Inviato da Posta <https://go.microsoft.com/fwlink/?LinkId=550986> per Windows
> 10
>
> Da: Antonio L'Abbate <mailto:[email protected]>
> Inviato: mercoledì 10 aprile 2019 15:43
> A: MATPOWER discussion forum <mailto:[email protected]>
> Cc: Andrea Mazza <mailto:[email protected]>
> Oggetto: R: Adding RES and LOAD profiles in MOST.
>
> Dear Professor Zimmerman,
> thank you for your reply. Even though I had already figured out how to upload
> profiles your hints gave me confirmation that my procedure is correct and
> that potential issues that may occur in the bigger network I am working on
> wouldn’t be due to the formulation, but to the network parameters themselves.
> Best wishes,
> Antonio L’Abbate.
>
> Inviato da Posta <https://go.microsoft.com/fwlink/?LinkId=550986> per Windows
> 10
>
> Da: Ray Zimmerman <mailto:[email protected]>
> Inviato: martedì 9 aprile 2019 17:04
> A: MATPOWER discussion forum <mailto:[email protected]>
> Oggetto: Re: Adding RES and LOAD profiles in MOST.
>
> Hi Antonio,
>
> First, about CT_REP vs. CT_REL, see Table 9-3 in the MATPOWER User’s Manual.
>
> Then, you should be able to start with the code from most_ex6.uc.m, namely …
>
> mpc = loadcase(casefile);
> xgd = loadxgendata('ex_xgd_uc', mpc);
> [iwind, mpc, xgd] = addwind('ex_wind_uc', mpc, xgd);
> profiles = getprofiles('ex_wind_profile_d', iwind);
> profiles = getprofiles('ex_load_profile', profiles);
> nt = size(profiles(1).values, 1); % number of periods
> mdi = loadmd(mpc, nt, xgd, [], [], profiles);
> mdo = most(mdi, mpopt);
>
> … and simply used modified versions of each of the input files. It looks like
> this may be how you started.
>
> Your version of ex_wind_uc.m will have 3 rows for the gen and xgd_table.data
> fields, rather than one.
>
> Your version of ex_load_profile.m should be very similar to the original,
> assuming you want to scale all loads together. Simply set
> loadprofile.values(:, 1, 1) to your own 24 x 1 vector of values representing
> the total load in each period. If you want to scale the loads separately,
> then you need to set loadprofile.values(:, 1, :) to a 24 x 3 vector of
> values, and set loadprofile.rows to [3;4;5], meaning that the 3 columns of
> loadprofile.values(:, 1, :) should be applied to loads at buses 3, 4 and 5,
> respectively.
>
> Your version of ex_wind_profile_d.m should have windprofile.values(:, 1, :)
> set to a 3 x 24 matrix of either scale factors (to scale the PMAX values set
> in your ex_wind_uc) or absolute MW values (for chg_type = CT_REP). And
> windprofile.rows should be set to [1;2;3], meaning that the 3 columns of
> windprofile.values(:, 1, :) should be applied to wind generators 1, 2, and 3
> respectively.
>
> Hope this helps,
>
> Ray
>
>
>
> On Apr 2, 2019, at 5:01 AM, Antonio L'Abbate
> <[email protected]
> <mailto:[email protected]>> wrote:
>
> Hello Community,
> I am trying to run a multiperiod deterministic UC in a network which is based
> on the example “most_ex6_uc”, but I haven’t been able to successfully
> implement loads and res profiles so far. I have been following the manual,
> but it is not clear to me which is the proper way to deal with profiles,
> since they won’t vary with time unit (1:1:24 hrs).
> My simplified network is composed of 5 buses, 3 traditional generators, 3
> loads modelled as negative gens (as showed in the abovementioned example) and
> 3 RES generators: the traditional gens and the loads are described in the mpc
> case, while the RES have been added in the xgd through the related
> loadxgendata command.
> I am attaching the relevant parts of the code in order to ease comprehension.
>
> mpc.gen = [
> 1 1250 0 25 -25 1 100
> 1 200 60 0 0 0 0
> 0 0 0 250 250 0 0;
> 1 125 0 25 -25 1
> 100 1 200 65 0 0 0 0
> 0 0 0 250 250 0 0;
> 2 200 0 50 -50 1
> 100 1 500 60 0 0 0 0
> 0 0 0 600 600 0 0;
> 3 -400 0 0 0 1
> 100 1 0 -450 0 0 0 0
> 0 0 0 500 500 0 0;
> 4 -350 0 0 0 1 100 1 0 -350 0 0 0 0 0
> 0 0 500 500 0 0;
> 5 -300 0 0 0 1 100 1 0 -300 0 0 0 0 0
> 0 0 500 500 0 0;];
>
> xgd_table.data = [
> 1 1 1 1 5 250 10 250 1e-9 1e-9 1e-6
> 250 1e-6 250;
> 1 1 3 1 1e-8 250 2e-8 250 1e-9 1e-9 1e-6
> 250 1e-6 250;
> 1 1 1 1 1.5 600 3 600 1e-9 1e-9 10
> 100 10 250;
> 2 1 1 1 1e-8 800 2e-8 800 1e-9 1e-9 1e-6
> 800 1e-6 800;
> 2 1 1 1 1e-8 800 2e-8 800 1e-9 1e-9 1e-6
> 800 1e-6 800;
> 2 1 1 1 1e-8 800 2e-8 800 1e-9 1e-9 1e-6
> 800 1e-6 800;];
>
> %(Main script)
> mpc = loadcase(casefile);
> xgd = loadxgendata('ex_xgd_ucR', mpc);
> [iwind, mpc, xgd] = addwind('ex_wind_ucR', mpc, xgd);
> % create load profiles
> aux1=ex_load_profileR; %extracting profiles through an auxiliary
> variable
> load_values=aux1.values;
> load_values = load_values./repmat(max(load_values,[],2),1,24); %conversion
> to pu
> iload = 4:6;
> for i = 1:size(load_values, 1)
> wp = load_values(i,:)';
> if i==1
> profiles = getprofiles('ex_load_profile_d_R',iload(i));
> else
> profiles = getprofiles('ex_load_profile_d_R',profiles,iload(i));
> end
> end
> % create res profiles
> aux2=ex_wind_profileR;
> RES_values=aux2.values;
> for i = 1:1:size(RES_values,2)
> wp = RES_values(:,i);
> profiles = getprofiles('ex_wind_profile_d_R',profiles,iwind(i)); %
> RES_values profiles
> end
> nt = 24;
> mpc_full = mpc;
> xgd_full = xgd;
> mdi = loadmd(mpc, nt, xgd, [], [], profiles);
> mdo = most(mdi, mpopt);
> if verbose
> ms = most_summary(mdo);
> end
>
> %ex_load_profile (idx_gen & idx_ct defined)
> loadprofile = struct( ...
> 'type', 'mpcData', ...
> 'table', CT_TLOAD, ...
> 'rows', 0, ...
> 'col', CT_LOAD_ALL_PQ, ...
> 'chgtype', CT_REP, ...
> 'values', [] );
> tmp2=load('tload_profiles.mat');
> tload_profiles=tmp2.three_load_profiles;
> loadprofile.values = tload_profiles; %(3x24 matrix)
>
> %ex_load_profile_d
> global wp;
> loadprofile = struct( ...
> 'type', 'mpcData', ...
> 'table', CT_TGEN, ...
> 'rows', 0, ...
> 'col', PMAX, ...
> 'chgtype', CT_REP, ...
> 'values', [] );
> loadprofile.values(:, 1, :) = -wp;
>
> %wind_profile (idx_gen & idx_ct defined)
> windprofile = struct( ...
> 'type', 'mpcData', ...
> 'table', CT_TGEN, ...
> 'rows', 0, ...
> 'col', PMAX, ...
> 'chgtype', CT_REP, ...
> 'values', [] );
> tmp1=load('res_profiles.mat');
> res_profiles=tmp1.three_res_profiles;
> windprofile.values = res_profiles; %(24x3 matrix)
>
> %ex_wind_profile_d
> global wp;
> windprofile = struct( ...
> 'type', 'mpcData', ...
> 'table', CT_TGEN, ...
> 'rows', 0, ...
> 'col', PMAX, ...
> 'chgtype', CT_REP, ...
> 'values', [] );
> windprofile.values(:, 1, :) = wp;
>
> Only the first res profile is recognized, while the load ones are not
> uploaded and only the Pmin values are given in the resulting optimased
> dispatch. I would like to ask for advice and for further clarification about
> the use of CT_REP and REL, since I cannot understand where and how to set
> these parameters and for what extent.
> Ps I set the Gurobi method as solver.
> Thanks in advance,
> Best Regards.
> Antonio L’Abbate.
> Inviato da Posta <https://go.microsoft.com/fwlink/?LinkId=550986> per Windows
> 10
