重复M次模拟计算自协方差估计的.doc

重复M次模拟计算自协方差估计的： （注模拟一次产生的随机数据为列向量) y=zeros(600,1000);epsilon=randn(600,1000); x=zeros(600,1000); for j=1:1000 for i=3:600 y(i,j)=2/1.8*cos(2.13)*y(i-1,j)-1/1.8/1.8*y(i-2,j)+epsilon(i,j); end t=1:500; x(t,j)=y(t+100,j); end (1)取前N=50个数据 meanx=mean(x(1:50,:)); stdx=std(x(1:50,:)); gamma(1,:)=stdx.*stdx*49/50; for j=1:1000 s=0; for k=1:5 for i=1:50-k s=s+(x(i,j)-meanx(1,j))*(x(i+k,j)-meanx(1,j)); end gamma(k+1,j)=s/50; s=0; end end AveI =mean(gamma) 1.3789 -0.6190 -0.0585 0.2049 -0.1079 -0.0121 StdI=std(gamma) 0.3429 0.2379 0.2028 0.2359 0.2363 0.2147 (2) 取前N=100个数据 meanx=mean(x(1:100,:)); stdx=std(x(1:100,:)); gamma(1,:)=stdx.*stdx*99/100; for j=1:1000 s=0; for k=1:5 for i=1:100-k s=s+(x(i,j)-meanx(1,j))*(x(i+k,j)-meanx(1,j)); end gamma(k+1,j)=s/100; s=0; end end AveI=mean(gamma) 1.3774 -0.6203 -0.0616 0.2250 -0.1217 -0.0047 StdI=std(gamma) 0.2403 0.1662 0.1475 0.1678 0.1623 0.1583 (3) 取前N=300个数据 meanx=mean(x(1:300,:)); stdx=std(x(1:300,:)); gamma(1,:)=stdx.*stdx*299/300; for j=1:1000 s=0; for k=1:5 for i=1:300-k s=s+(x(i,j)-meanx(1,j))*(x(i+k,j)-meanx(1,j)); end gamma(k+1,j)=s/300; s=0; end end AveI=mean(gamma) 1.3856 -0.6235 -0.0640 0.2320 -0.1211 -0.0011 StdI=std(gamma) 0.1330 0.0947 0.0877 0.0971 0.0938 0.0925 结果如下: N=50 0 1 2 3 4 5 4．4386 -2.3115 -1.3397 3.1031 -1.8850 -0.6850 1.3866 -2.3098 -1.3877 3.2474 -1.9855 -0.7687 4.3749 1.1047 0.8470 1.8610 1.2850 0.7873 N=100 0 1 2 3 4 5 4.4003 -2.3058 -1.3624 3.1688 -1.9254 -0.7371 4.3749 -2.3098 -1.3877 3.2474 -1.9855 -0.7687 1.4046 0.7648 0.6089 1.3244 0.9174 0.5943 N=300 0 1 2 3 4 5 4.3832 -2.3120 -1.3717 3.2158 -1.9697 -0.7448 4.3749 -2.3098 -1.3877 3.2474 -1.9855 -0.7687 0.8130 0.4418 0.3570 0.7769 0.5318 0.3495 重复M次模拟计算自协方差估计的： （注模拟一次产生的随机数据为列向量) y=zeros(600,1000);epsilon=randn(600,1000); x=zeros(600,1000); for j=1:1000 for i=3:600 y(i,j)=2/1.1*cos(2.13)*y(i-1,j)-1/1.1/1.1*y(i-2,j)+epsilon(i,j); end