资源简介
分析研究了用CV模型,CA模型,Singer模型对单机动目标进行跟踪算法的matlab仿真。比较了各模型的滤波性能,得出了一些有意义的结论。
代码片段和文件信息
clear;
clc;
T=2;
num=50;
N0=400/T;
N1=600/T;
% N3=610/T;N4=660/T;N5=900/T;
x=zeros(N11);
y=zeros(N11);
vx=zeros(N11);
vy=zeros(N11);
x(1)=-10000;y(1)=2000;
vx(1)=15;vy(1)=0;
ax=0;ay=0;
var=100;
for i=1:N1-1
if(i>N0-1&i<=N1-1)
ax=-0.075;ay=0.075;
vx(i+1)=vx(i)+ax*T;
vy(i+1)=vy(i)+ax*T;
else
ax=0;ay=0;
vx(i+1)=vx(i);
vy(i+1)=vy(i);
end
x(i+1)=x(i)+vx(1)*T;
y(i+1)=y(i)+vy(i)*T;
end
rex(numN1)=0;
rey(numN1)=0;
%
for m=1:num
nx=100*randn(N11);
ny=100*randn(N11);
zx=x+nx;
zy=y+ny;
rex(m1)=-10000;
rey(m1)=2000;
rex(m2)=-9960;
rey(m2)=2000;
ki=0;
low=1;high=0;
u=0;ua=0;
e=0.8;
xks(1)=zx(1);
yks(1)=zy(1);
xks(2)=zx(2);
yks(2)=zy(2);
o=[1T000.5*T^20;0100T0;001T00.5*T^2;00010T;000010;000001];
h=[100000;001000];
g=[T^2/40;T/20;0T^2/4;0T/2;10;01];
q=[100000;010000];
% o=[1T00;0100;001T;0001];
% h=[1000;0010];
% g=[T^2/20;T0;0T^2/2;0T];
% q=[100000;010000];
% perr=[var^2var^2/T00;var*var/T2*var^2/(T^2)00;00var^2var^2/T;00var^2/T2*var^2/(T^2)];
perr=diag([100001000010000100001000010000]);
vx=(zx(2)-zx(1))/2;vy=(zy(2)-zy(1))/2;
xk=[zx(1);vx;zy(1);vy;ax;ay];
%
for r=3:N1
z=[zx(r);zy(r)];
xk1=o*xk;
perr1=o*perr*o‘;
k1=perr1*h‘*inv(h*perr1*h‘+q);
xk=xk1+k1*(z-h*xk1);
perr=(eye(6)-k1*h)*perr1;
xks(r)=xk(11);
yks(r)=xk(31);
vxks(r)=xk(21);
vyks(r)=xk(41);
xk1s(r)=xk1(11);
yk1s(r)=xk1(31);
vxk1s(r)=xk1(21);
vyks1(r)=xk1(41);
perr11(r)=perr(11);
perr12(r)=perr(12);
perr22(r)=perr(22);
rex(mr)=xks(r);
rey(mr)=yks(r);
end
end
ex=0;ey=0;
eqx=0;eqy=0;
ex1(N11)=0;ey1(N11)=0;
qx(N11)=0;qy(N11)=0;
for i=1:N1
for j=1:num
ex=ex+x(i)-rex(ji);
ey=ey+y(i)-rey(ji);
%eqx=eqx+(x(i)-rex(ji))^2;
eqx=eqx+(x(i)-rex(ji))^2;
eqy=eqy+(y(i)-rey(ji))^2;
end
% eqx/num-(ex1(i)^2)
%sqrt((double(eqx)/num-(ex1(i)^2)));
ex1(i)=ex/num;
ey1(i)=ey/num;
qx(i)=sqrt((double(eqx)/num-(ex1(i)^2)));
qy(i)=(eqy/num-(ey1(i)^2))^0.5;
ex=0;eqx=0;ey=0;eqy=0;
end
figure(1);
plot(xy‘k-‘zxzy‘g:‘xksyks‘r-.‘);
legend(‘true trace‘‘observation samples‘‘estimated samples‘);
figure(2);
plot(zxzy);
legend(‘observation samples‘);
figure(3)
plot(xksyks);legend(‘estimated trace‘);
figure(4);plot(xex1);legend(‘the error at x anix‘);
figure(5);plot(xqx);
legend(‘the square error at x anix‘);
属性 大小 日期 时间 名称
----------- --------- ---------- ----- ----
文件 2898 2010-06-18 22:09 仿真大作业\ca.m
文件 3378 2010-06-18 19:39 仿真大作业\cv2.m
文件 4441 2010-06-18 21:59 仿真大作业\singer3.m
文件 446305 2010-06-18 22:25 仿真大作业\单模型机动目标跟踪算法的仿真研究.doc
目录 0 2010-06-18 22:25 仿真大作业
----------- --------- ---------- ----- ----
457022 5
评论
共有 条评论