1、第31卷第9期 西南大学学报 (自然科学版) 2009年9月Vo l1 31 No1 9 Journal of So uthw est University ( Natural Science Edition) Sep1 2009cI|: 1673- 9868( 2009) 09-0138- 04 i g g 3 3 卜长明1, 冯国建1, 彭 义2, 王 华211 ivr !, i400045; 21 2v/ , i400710K1: 提出了用于预测垃圾产生量的变权重组合预测模型, 该模型通过把单项预测模型进行组合分析, 以误差平方和最小为准则, 求出最优加权组合系数, 建立组合预测模型.
2、通过实例分析, 表明其预测精度高于各单项模型预测值, 在城市生活垃圾产生量的预测工作中有一定的应用价值.1 o M: 城市生活垃圾; 产生量; 变权重组合预测ms |: X33DS M : A“SE g 3hn 4, g 9 Es .t 3 v ib,/Y. “ - ) Z T1 3A agv ) 1 , 3A ( v E1/ 15 1/ 8, g)E1/ 5 1/ 3)a aS,g 6y/ A A ,“ n 3 .yNy/A - u EsA1. E E 4aLB E 5a * E 6.tZE1+, B E:F E g .1 M F g 3 3 5 7-9 , |n , N HW,iL !:
3、y( t)t L= 3 , t = 1, 2, , N ; yCi( t)i t 3 ; w i(t)i t F , O Eni= 1w i(t) = 1 i = 1, 2, , n; t = 1, 2, , N5M F VV U:yC( t) = Eni= 1w i(t)yCi(t)yC(t)M F t . !eiti t :eit = y( t) - yCi (t) l : 2008-11-05 “: i g g$ “( 200511) .Te: ( 1982-) , 3, ,p V 3,1V Yr !Z .KDM “ w i(t) V Kl=E p, |m inS = ENt= 1 yC
4、(t) - y(t) 2 = minw1 , , wN 0 e21 + e22 + ,+ e2N et = yC( t) - y( t) = Eni= 1w i(t) yi(t) - y(t)= w 1(t), , w n(t) e1t, , ent T= e1t, , ent w1 (t) , , w n(t) T#e2t = w 1(t), , w n(t) #e1te2tsent# e1t, , ent # w 1(t), , w n( t) T = W Tt A tW tW t = w 1(t), , w n(t) TA t = e1t, , ent T # e1t, , ent =e
5、1te1t , e1tente2te1t , e2tents s sente1t , ententSt = ENt= 1e2t = ENt = 1W Tt A tW tN, “ (d ,#5?ZE pd F ,minS Hq:RT W t = 1W t 0m1 i g1996- 2005 M 3 : R = 1, , 1 T ,n_ . VW t = A- 1t RR TA tR = w 1t, w 2t, , w ntTt = 1, 2, , N2 L s i g1996- 2005 M 3 3 (nV1)s, | wL aLB wL T s. g ,#S ( M 3 440 kg 2 ,V
6、9 i g1996 - 2005 M 3 ,nm1.Vm V A, i g M 3 v9F.V1 i g1996- 2005 M g 10M z1996 1997 1998 1999 2000 g / 3 0221 77 3 0421 92 3 0591 69 3 0721 34 3 0911 09M z2001 2002 2003 2004 2005 g / 3 0971 91 3 1131 83 3 1301 10 3 1441 23 3 1691 16139第9期 卜长明, 等: 重庆市城市生活垃圾产生量预测21 1 sY wL aLB wL ,T# nV2.VV V A, (M?Z t
7、 , L= W i v.V2 t Mz 3 wL LB wL 1996 1 3301 02 1 3311 68 - 11 66 1 3311 42 - 11 40 1 3311 62 - 11 601997 1 3381 89 1 3381 38 01 51 1 3381 11 01 78 1 3381 18 01 711998 1 3461 26 1 3451 09 11 17 1 3441 80 11 46 1 3441 77 11 491999 1 3511 83 1 3511 79 01 04 1 3511 49 01 34 1 3511 39 01 442000 1 3601 08
8、1 3581 48 11 60 1 3581 19 11 89 1 3581 06 21 022001 1 3631 08 1 3651 18 - 21 10 1 3641 88 - 11 80 1 3641 75 - 11 672002 1 3701 08 1 3711 87 - 11 79 1 3711 58 - 11 50 1 3711 48 - 11 402003 1 3771 24 1 3781 55 - 11 31 1 3781 27 - 11 03 1 3781 24 - 11 002004 1 3831 46 1 3851 23 - 11 77 1 3841 96 - 11 5
9、0 1 3851 03 - 11 572005 1 3941 43 1 3911 91 21 52 1 3911 66 21 77 1 3911 85 21 5821 2 F | 3 F,y M F .V3M F “ aF# L= W. nV4.V3 M “ aF# MzM “ LB M F L= 1996 01 146 01 664 01 190 1 3311 45 1 3301 02 11 431997 01 077 01 722 01 201 1 3381 12 1 3381 89 - 01 771998 01 514 01 007 01 479 1 3441 78 1 3461 26
10、- 11 481999 01 011 01 885 01 104 1 3511 49 1 3511 83 - 01 342000 01 832 01 091 01 077 1 3581 17 1 3601 08 - 11 912001 01 124 01 703 01 173 1 3641 86 1 3631 08 11 782002 01 817 01 169 01 014 1 3711 56 1 3701 08 11 482003 01 024 01 877 01 099 1 3781 27 1 3771 24 11 032004 01 192 01 102 01 706 1 3851 0
11、1 1 3831 46 11 552005 01 018 01 009 01 974 1 3931 24 1 3941 43 - 11 19V4 #M F ( Z) LB M F251 75 241 86 241 43 181 83m2 M F L= W1 m2M F L= W1 .Vm V A, L=W1 ,.3 4 M FZE,V7 V8 4.TV 1 B ( A. V / :1)4M FZE ZEF L=,4 ;140 西南大学学报(自然科学版) 投稿网址 http: / / xbgjx t1 sw u1 cn 第 31卷2)M FZE 1M L $,5Y . ID: 1 ,R +. g
12、 3 3A !9 2 M .:S S, 2005. 2 . 3Q A “d_ D .s:v, 2003. 3 o, .A F M FZE J . 2 S/v(1 S) , 2005,20( 4) : 53- 57. 9 u , ,L. r M FS wLZE J .r , 2005, 26( 9) : 1443-1447. 10 i gd9 . id9 Mk M .:Sd9, 2006: 61.Forecastingof the Output of Municipal SolidWastein ChongqingBU Chang-ming1, FENG Guo- jian1, PENG Yi2
13、, WANG H ua211 Co lleg e o f Civil Eng inee ring , Chong qing Unive rsity , Chong qing 400045 , China ;21 Schoo l of Te chno log y and Engineering , Southwe st Univ ersity , Cho ngqing 400710 , ChinaAbstract: This paper introduces a variable-w eight combined fo recasting m odel applied to M SW ( m u
14、nicipalsolid w aste) output forecasting. The m ethod com bines the fo recasting models available one by one and usesthe practical g uide line based on the minim al sum of er ror square values, then g ets o ptimal w eight coeff-icients of com bining forecasting and sets up the optim al combining for ecasting mo del. A case study indica-ted that the model is more accur ate than any sing le fo recasting models and is applicable in forecasting o fM SW output.Key words: municipal solid w aste; output; v ariable-w eig ht com bined forecasting3 I 汤振金 141第9期 卜长明, 等: 重庆市城市生活垃圾产生量预测
