如:

 y=[0.004727321 -0.000131427 -0.016868652 -0.02206931 0.000565095 -0.015403815 0.011055909 -0.002626233 0.013615303 -0.008617022 0.004355091 0.000554525 0.008631263 -0.001801558 -0.017912737 0.001736653 -0.001412598 -0.001047169 0.007705676 -0.00052925 0.019802613 0.00263205 -0.019001018 0.006453316 0.00689125 0.00139217 -0.003677394 -0.012459264 0.002078434 -0.000463948 -0.004059158 -0.013789763 0.0079596 0.010866068 -0.001173143 0.013802899 0.002595466 0.001316768 0.006754097 -0.001626103 0.003515611 0.003760499 0.004055666 -0.004910502 0.003961937 0.008482788 0.010437866 -0.00218703 0.002373754 0.010717095 -0.007157333 0.01713951 0.013810798 0.010148087 0.005167766 0.004071829 -0.007840278 0.01292267 -0.004271623 -0.015219491 0.004630592 0.020613353 0.014780881 0.002980233 0.000366443 -0.002954363 0.004784116 0.023512579 -0.004302478 0.006216412 -0.015976001 0.010406081 -0.002354482 0.005228033 0.009981886 -0.021967872 -0.002534176 0.000102571 0.016472853 -0.003259807 0.003605301 0.006214843 0.000254461 0.001418451 0.005819727 -0.016011144 0.005942427 -0.003371195 -0.022522208 -0.007571909 -0.004166733 0.00038549 0.011255168 -0.000500131 0.012525416 -0.000486347 -0.003712162 0.014935484 0.001529005 0.004897744 0.004434299 -0.005957956 0.000942303 0.002214643 0.005220243 -0.008310546 0.002765077 0.016305939 0.007828901 0.007790645 -0.000313395 0.002403451 0.012003336 0.002317744 -0.00154159 -0.020439223 0.020422868 0.014026803 0.004714894 0.008367566 -0.007732512 0.022288765 -0.004940746 -0.007107442 0.012652079 -0.000529108 0.016580435 0.039493966 0.022750651 0.009489537 -0.00537624 0.042572428 0.038219899 -0.030551232 0.007376414 -0.00484263 0.026136476 -0.001120786 -0.032056657 -0.008494593 0.000320569 0.014113093 0.021469865 0.005307727 -0.009648281 0.026131725 -0.00878167 0.009117048 -0.003258895 -0.053349056 0.001220471 -0.025202797 0.000915314 -0.002775631 -0.011081764 0.001730522 0.02639539 0.007507288 0.003807627 0.003629901 -0.001276508 0.006039835 0.017275847 0.003734427 -0.00015858 0.019721953 0.000352951 0.0152134 -0.009483698 0.022009456 0.013330928 -0.006477286 0.002392801 0.006619271 -0.003327968 -0.048408046 0.005701275 0.01061728 0.000808134 -0.023209144 0.006054122 0.006168737 0.000366294 0.011717109 0.000706039 -0.006806391 -0.00784398 -0.029662521 -0.007515207 0.000162438 0.000118685 -0.01930596 -0.014463586 0.021416016 -0.001278012 0.017420941 -8.09434E-05 -0.021912537 0.015865134 0.013004023 -0.008092006 -0.003311857 0.001958674 0.007619523 -0.000588837 0.006586455 6.16135E-06 0.016888166 0.000351421 0.002525711 0.002084353 -0.013233734 0.012726906 0.00395772 0.004831498 -0.006536612 0.004087431 0.004190698 0.012471865 -0.003815173 0.000177579 0.018553699 0.006943476 0.001431053 -0.001740163 0.004877486 -0.008926417 -0.001738768 -0.005643419 0.007240284 0.006909985 0.008900608]';
z=iddata(y)

Time domain data set with 232 samples.
Sampling interval: 1                  
                                      
Outputs      Unit (if specified)      
   y1                                 
                                      
>> armax(z,'na',5,'nc',3)
Discrete-time IDPOLY model: A(q)y(t) = C(q)e(t)                      
A(q) = 1 + 0.03509 q^-1 - 0.1458 q^-2 + 0.565 q^-3 - 0.08006 q^-4    
                                                         - 0.1644 q^-5
                                                                     
                                                                     
C(q) = 1 + 0.03411 q^-1 - 0.15 q^-2 + 0.6385 q^-3                    
                                                                     
Estimated using ARMAX from data set z                                
Loss function 0.000147109 and FPE 0.000158212                        
Sampling interval: 1  

则模型为:

Y=-0.03509Yt-1 + 0.1458Yt-2 - 0.565Yt-3+ 0.08006Yt-4+0.1644Yt-5+ε+ 0.03411εt-1 - 0.15εt-2 + 0.6385εt-3

下一步如何作预测?

我也不知道怎么做了，

`` `

用SCA做预测就好了，何必自找苦吃?

比如ARMA(2,2)

y(2) = a1*y(1) + e(2)

y(3) = a1*y(2)+a2*y(1)+e(3)+b1*e(2)

y(4) = a1*y(3)+a2*y(2)+e(4)+b1*e(3)+b2*e(2)

y(5) = a1*y(4)+a2*y(3)+e(5)+b1*e(4)+b2*e(3)

.

.

.

具体的,你可以把前面没有的比如当y(1)时候没有y(0)和y(-1)

你可以把它们设做NaN然后转化NaN为0
for i=1:length(y)
       if isnan(y(i))==1
          y(i)=0;
       end
end

非常的感谢！

系统辨识工具箱里有一个predict函数，直接用就可以了。

我想用MATLAB的statistic工具箱和System Identification工具箱进行时间序列的预测，不知道怎么用，不知哪位大虾能不吝赐教？谢谢了哦！

把时间T往前取一步成T+1就可以预测了啊。