Optimization Problem. I need find a local minimun instead of global minimum

function [JPF, P_sim] = builder_pdf_est(t,rr,lambda) %% %--------------------------Estimated_PDF----------------------------------%

% Parameters estimated by the Global Minimum Searching Algorithm Patternsearch

t1min = t(1); tp1 = t(3); t2min = t(2); tp2 = t(4); alfa = t(5);

%---Application of the model using RR series created by the Simulation----%

for k=1:length(rr)

    if rr(k) < t1min

        p1(k) = 0;

    elseif (rr(k) >= t1min && rr(k) < (t1min+tp1))

        p1(k) = (lambda*((rr(k)-t1min)/tp1)*exp(lambda*(((rr(k)-t1min)^2/(2*tp1)))));

    else

        p1(k) = (lambda*exp(lambda*(tp1/2-(rr(k)-t1min-tp1))));

    end
end
for k=1:length(rr)
    if rr(k) < t2min

        p2(k) = 0;

    elseif (rr(k) >= t2min && rr(k) < (t2min+tp2))

        p2(k) = (lambda*((rr(k)-t2min)/tp2)*exp(lambda*(((rr(k)-t2min)^2/(2*tp2)))));

    else

        p2(k) = (lambda*exp(lambda*(tp2/2-(rr(k)-t2min-tp2))));

    end

end

P = (alfa)*p1+(1-alfa)*p2; % Estimated PDF P_sim = (P/sum(P)); % Normalization useful for the graph JPF = -sum(log10(P')); % Calculation of Joint Probability function

---------------------------------------------------------------------------------------------

function [a, fval, exitflag, OUTPUT] = Estimator(rr, data) %%

% Estimation of parameters teta=[t1min t2min tp1 tp2] by jointly maximizing the

% log-likehood function with respect to teta

% Algorithm used: ______________

t = [ rand(1) rand(1) rand(1) rand(1) rand(1)]; %Initial Guess options = psoptimset('MaxFunEvals', 1e10000, 'MaxIter', Inf); lambda = data(6); % As lambda we are using the set parameter

Lb = zeros(1,5); Ub = [1 ,.5, 1, .5, 1]; A = [0 1 0 1 0; 1,1,0,0,0; 0,0,1,1,0]; b = [.5, 1, 1]; JPF = @(t) builder_pdf_est(t,rr, lambda); [a, fval, exitflag, OUTPUT] = patternsearch(JPF, t, A, b, [], [], Lb, Ub, [], options);

[Alan Weiss]

What is your question? I do not understand what you are asking.

Alan Weiss

MATLAB mathematical toolbox documentation