计算lpcc的matlab程序

function[y]=lpcc(input) p=12; NJ=256; for m=0:1:p r(m+1)=0; for i=1:1:(NJ-m) r(m+1)=r(m+1)+input(i)*input(i+m); end end R=zeros(1,p); R(1,(1:p))=r(1,(2:p+1)); R=R'; E=zeros(1,p); k=E; a=zeros(p,p); e=r(1); %Durbin算法初始化 k(1)=r(2)/r(1); a(1,1)=k(1); E(1)=(1-k(1)^2)*e; l=2; for i=2:1:12; %算法开始 m_t=sum(a((i-1:-1:1),i-1).*R((1:1:i-1),1)); k(i)=(R(i)-m_t)/E(i-1); a(i,i)=k(i); a((1:1:i-1),l)=a((1:1:i-1),i-1)-k(i)*a((i-1:-1:1),i-1); E(i)=(1-k(i)^2)*E(i-1); l=l+1; end y=a(:,12)'; %预测系数 A=zeros(1,12); %计算LPCC参数，使用12阶 A=y(1,end:-1:1); c=zeros(1,12); c(1)=A(1); for n=2:1:12 c(n)=A(n)+sum((1:1:n-1)/12.*c(1:1:n-1).*A(n-(1:1:n-1))); end y=c;