ajaykumarsampath

%% In this scipt we calcualte the maximum and 
% average flop counts required with different
% algorithms
clear all;
close all;
clc;
%% Reevaluating the Average gradient and Hessian counts 

% LBFGS
AverGrad(1,:)=[84.0426 68.7755 129.750 80.4286...

ajaykumarsampath

%% plot figures 

result_data=cell(3,2);
gurobi_ip=cell(3,1);
%prediction hoizon 
result_data{1,1}=[8 10 30 45;0.1954 0.2042 0.6179 0.8864;0.0419 0.0590 0.1318 0.2052]';
result_data{1,2}=[8 10 30 45;1.1501 1.572 3.6162 5.2541;0.1887 0.221 0.4452 0.7612]';
gurobi_ip{1,1}=[8 10;1.9436 2.7031;2.2394 3.1234]';

figure(1)

ajaykumarsampath

/*
 * Effinet_data_generation.cuh
 *  Created on: Jul 26, 2015
 *      Author: ajay
 */

/*
 * This header file has the implementation of the functions that allocate the memory in the
 * GPUs.
 */

ajaykumarsampath

clear all
syms a b w l r k Q x;
f1=a^2+4*(w*l+r*k-r*k*x)*(b-a)/((1-x)*Q/x);
f=symfun(0.5*(a+...
    (a^2+4*(w*l+r*k-r*k*x)*(b-a)...
    /((1-x)*Q/x))^0.5),[a b w l r k Q x]);
g=diff(f,x);

%{ 
result 

ajaykumarsampath

%%
% This function generates a system with different terminal functions and constraints but
% with same size. The constraint are preconditioned accodingly.
% We solve the method using different methods. First formulated using
% 1) Gurobi-IP 2) Gurobi-AS directly

clear all;
close all;
clear model;
clc;

ajaykumarsampath

function [ S ] = system_masses(N, options )
%masses_model function is used to find the dynamic equations of N masses
%system. The no of masses is specified by N and options specify the constraints
%on the system.
% The syntax of the function is
% S=masses_model(N,options)

default_options = struct('M', 1*ones(N,1), 'b', 0.1*ones(N+1,1),...
    'k',1*ones(N+1,1),'xmin',-5*ones(2*N,1), 'xmax', 5*ones(2*N,1), 'umin', ...
    -5*ones(N,1),'umax',5*ones(N,1), 'Ts', 0.1);

ajaykumarsampath

%% MPC

clear all
clear classes
clc

% User-defined parameters:
load('dwn_big.mat');
f0=0.1;
k=3000;

ajaykumarsampath

function [ Tree details] = Tree_formation( W,ops)
%This function generates the scenario tree.
%The no of scenatios are reduced by forward selection.
%The input is scenario of random variables

%syntax: [Tree]=Tree_formation(W,ops)
%Input:     W: The stochastic varible.
%              W=[scenarios,no of varibles,future steps]
%              The stocastic variable is considered independent
%         ops: This contains the option to the tree

ajaykumarsampath

clear all;
close all;
clc;
Nm=10; % Number of masses
T_sampling=0.25;
sys_no_precond=system_generation(Nm,struct('Ts',T_sampling,'xmin', ...
    -5*ones(2*Nm,1), 'xmax', 5*ones(2*Nm,1), 'umin', -1*ones(Nm,1),'umax',...
    1*ones(Nm,1)));