灰色預(yù)測模型的matlab運(yùn)行代碼(講解)灰色預(yù)測模型GM(1,1)的matlab運(yùn)行代碼例由1990—2001年中國蔬菜產(chǎn)量，建立模型預(yù)測2002年中國蔬菜產(chǎn)量，并對預(yù)測結(jié)果作檢驗(yàn)。分析建模：給定原始時(shí)間1990—2001年資料序列X(k),對X(k)生成1-AGO(累加)序列X(k)及Y。見下表K123456789101112X19519,19578,19637,19695,16602,25723,30379,34473,38485,40514,42400,48337X19519,39097,58734,…………264605,307005,355342Y-----1957819637…………405144240048337其中X(k)=;Y=對上述X(k)的GM(1,1),得到對值）error2(i)=error1(i)/x0(i);（計(jì)算相對誤差）end>>C=std(error1)/std(x0);（計(jì)算后驗(yàn)差檢驗(yàn)數(shù),std表示標(biāo)準(zhǔn)差）>>k=1;（k表示預(yù)測長度，這里每次預(yù)測下一年）a=-0.106210475032772>>bb=1.399996741173038e+04BB=1.0e+05*-0.2930800000000000.000010000000000-0.4891550000000000.000010000000000-0.6858150000000000.000010000000000-0.8673000000000000.000010000000000-1.0789250000000000.000010000000000-1.3594350000000000.000010000000000-1.6836950000000000.000010000000000-2.0484850000000000.000010000000000-2.4434800000000000.000010000000000-2.8580500000000000.000010000000000-3.3117350000000000.000010000000000>>C（求后檢驗(yàn)數(shù)）C=0.163969348419772>>x（原始數(shù)據(jù)的對應(yīng)的預(yù)測數(shù)據(jù)，這里也是12個(gè)）x=1.0e+04*Columns1through31.9519000000000001.6957693858307821.885790370699694Columns4through62.0971043303046062.3320972683461912.593422554345592Columns7through92.8840308835651903.2072035941156563.566589717441855Columns10through123.9662471805347304.4106886250944284.904932361001964>>eroor1（求殘差）eroor1=1.0e+03*Columns1through402.6203061416921850.7790962930030651.276043303046055Columns5through86.7189726834619070.2112255434559191.5386911643481002.400964058843441Columns9through122.8191028255814450.8515281946526961.7068862509442840.712323610019637>>error2（求相對誤差）error2=Columns1through400.1338393166662670.0396749143455250.064790215945471Columns5through80.4047086304940310.0082115438889680.0506498293014290.069647667996503Columns9through120.0732519897513690.0210181219986350.0402567512015160.014736611912606a=-0.106210475032772b=1.399996741173038e+04>>BB=1.0e+05*-0.2930800000000000.000010000000000-0.4891550000000000.000010000000000-0.6858150000000000.000010000000000-0.8673000000000000.000010000000000-1.0789250000000000.000010000000000-1.3594350000000000.000010000000000-1.6836950000000000.000010000000000-2.0484850000000000.000010000000000-2.4434800000000000.000010000000000-2.8580500000000000.000010000000000-3.3117350000000000.000010000000000方法二程序（1）一次累加生成序列的matlab命令>>x0=[19519,19578,19637,19695,16602,25723,30379,34473,38485,40514,42400,48337];>>x1(1)=x0(1);>>x1(1)x1(1)=19519>>fort=2:12;x1(t)=x1(t-1)+x0(t);endx1回車x1=Columns1through81951939097587347842995031120754151133185606Columns9through12224091264605307005355342（2）由一次累加生成序列緊鄰均值生成Z的matlab命令：x0=[19519,19578,19637,19695,16602,25723,30379,34473,38485,40514,42400,48337];>>x1(1)=x0(1);>>fort=2:12;x1(t)=x1(t-1)+x0(t);z1(t)=(1/2)*(x1(t)+x1(t-1));endz1=1.0e+05*Columns1through700.29310.48920.68580.86731.07891.3594Columns8through121.68372.04852.44352.85813.31