我知道问题出现在30行后，前28行没问题，前28行的目的是为了求t，然后代入后面的运算。总体运算思路是：y1*y2，然后求出半高宽度，需要u和半高宽度的关系，

怎么修改？？？

可能是x、y2，矩阵维度不一致，我不会修改，

x=linspace(0.88,0.92,100);

A=2.5488;B=0.01800964;C=1.1557;D=0.06959044;

E=2.8525;F=0.01800946;G=1.5141;H=0.06922161;

a=(1+(A.*(x.^2))./((x.^2)-B)+(C.*(x.^2))./((x.^2)-D)).^(1/2);%no

b=(1+(E.*(x.^2))./((x.^2)-F)+(G.*(x.^2))./((x.^2)-H)).^(1/2);%ne

aa=6.63*10^(-34);%普朗克常量

bb=3*10^8;%真空中的光速

cc=1.6*10^(-19);%电子电荷

dd=aa*bb./(cc.*x*10^(-6));

ee=(0.2309.*(dd.^2)./(((5.29^2)-dd.^2).^2));%旋光率

ff=x.*ee./(2*pi.*a);

c=0.4154;%入射角

d=((((cos(c))^2)./((a.^2).*(1+ff).^2))+((sin(c)).^2)./(b.^2)).^(-1/2);%考虑旋光率以后的ni

e=atan(((a./b).^2).*((1+ff).^2)./((1-ff).^2)).*tan(c);%衍射角

f=((((cos(e)).^2)./((a.^2).*((1-ff).^2)))+(((sin(e)).^2)./(a.^2))).^(-1/2);%考虑旋光率以后的nd

g=((d.^2)+(f.^2)-2.*d.*f.*cos(c-e))./x;

h=A*B./(((x.^2)-B).^2);

i=C*D./(((x.^2)-D).^2);

j=(-x./a).*(h+i);%dno/d.o

k=E*F./(((x.^2)-F).^2);

l=G*H./(((x.^2)-H).^2);

m=(-x./b).*(k+l);%dne/d.o

n=(((2.*a)./(b.^2)).*j)-(((2.*(a.^2))./(b.^3)).*m);

o=(((d.^3).*((cos(c))^2))./(((1+ff).^2).*a.^3));

p=((d.^3).*((sin(c))^2))./(b.^3);

q=o.*j+p.*m;%dni/d.o

ss=(((1-ff).^(-2)).*((cos(e)).^2)+((sin(e)).^2)).*((f./a).^3).*j;%dnd/d.o

t=2*pi.*((d-f)-x.*(q-ss));%b'

y1=(sin((-t.*4000.*(x-0.9)./(2.*(0.9^2)))).^2)./((-t.*4000.*(x-0.9)./(2.*(0.9^2))).^2);

u=linspace(0,10,100);v=u*10^-3;

y2=(sin((-t*4000.*(x-v-0.9)./(2.*((0.9+v).^2)))).^2)./((-t.*4000.*(x-v-0.9)./(2.*((0.9+v).^2))).^2);

y3=y1.*y2;

W=max(y3)/2;

T=abs(x(find(y3>=W,1,'first')) - x(find(y3>=W,1,'last')));

Q=T.*1000;

plot(v,Q)