PCCI method for hybrid uncertainty analysis

This code is used to analyze the hybrid uncertainty with random variables % and interval variables by using PCCI method proposed by Jinglai

% function [int_mean,int_std]=PCCI_regression(fun_name,a,b,k1,K1,Expansion,c,d,k2,K2,Distribution)
%Input
% fun_name the called function name
% a the lower bound vector of interval input
% b the upper bound vector of interval input
% k1 the order of CI expansion
% K1 the scanning (validation) points of each interval variable, larger value makes the result more accurate
% Expansion expansion type of Chebyshev polynomials-'full' or 'partial'
% c the mean vector of random variable
% d the standard deviation vector of random variable
% k2 the order of PC expansion
% K2 the level of LHS sampling points (not used currently)
% Distribution currently only contains 'gaussian' and 'uniform' distribution

%Output
% int_mean a structure contains the interval mean of response
% int_std a structure contains the interval standard deviation of response

%Example
%[int_mean,int_std]=PCCI_regression(@fun_test,[-1;-1],[1;1],2,20,'partial',[0;1],[0.1;0.2],2,100,'gaussian')