求助

422258511

p=[724.570000000000,739.774223126620,771.489090673942,804.563606600594,839.056060404793,875.027240514377,912.540541418428,951.662075391767,992.460789009143,1035.00858465455]
t=[724.570000000000,746.620000000000,778.270000000000,800.800000000000,827.750000000000,871.100000000000,912.370000000000,954.280000000000,995.010000000000,1037.20000000000]
[pn,minp,maxp,tn,mint,maxt]=premnmx(p,t);          %原始数据归一化
3 N) H. E  m4 Z% |" [
net=newff(minmax(pn),[9,1],{'tansig','purelin'},'traingdx'); %设置网络,建立相应的BP网络
net.trainParam.show=2000; % 训练网络
net.trainParam.lr=0.01;
& b0 s2 A- o$ u9 A) o$ {# lnet.trainParam.epochs=10000;
( t+ z! f' \2 L" Fnet.trainParam.goal=1e-5;
/ }' V! U7 p: U1 B0 {% O. h[net,tr]=train(net ,pn,tn);                              %调用TRAINGDM算法训练BP网络

an=sim(net,pn);                               %对BP网络进行仿真
# G6 h! V% t' m1 F) S. Y& ra=postmnmx(an,mint,maxt);                     %还原数据
x=2001:2010;
4 e+ l, L1 M) z$ ~- V  l4 lnewk=a(1,;
0 O6 |  m# m* v2 `: E2 ]  x7 b! kplot(x,newk,'r-o',x,t,'b--+')
+ f5 _1 ?* M# Z2 h" F7 T4 P" G1 plegend('网络输出值/万人','实际人口数/万人');
- T$ X- c- N- Mxlabel('年份');
% q/ x! n0 G! g, h5 i0 ]ylabel('人口数');
5 S6 J) T* K" P# Stitle('预测人口数对比表');
) M$ [- y' c/ B9 |# `9 {( z
pnew1=[1079.38044723977,1125.65457635540,1173.91252408743,1224.23933874207,1276.72371473226,1331.45814888992,1388.53910347951,1448.06717620014,1510.14727747579,1574.88881534617]
pnew=pnew1
pnewn=tramnmx(pnew,minp,maxp);
anewn=sim(net,pnewn);                               %对BP网络进行仿真
anew=postmnmx(anewn,mint,maxt);                     %还原数据
* P9 X1 u. x- h, \( ry=anew'
3 [9 e% n5 h) S# s9 |/ O$ L

. b" ?! X0 o. h预测的结果怎么都一样呀？》？？求助

wssl103050