1、null null c I | :1002- 1566(2002)01- 0041- 06Logistic w L E Z E null殷祚云( ; C ,o . F (1984) B “ ,V V 4 9 V A , H W 9 = “ 9 9 w L M ,. v !N/ !t K v (40,2null4) (50,4null6) W , N ) N + (44null null null null null null null null null null null null null null null d 9 5 null null null null 21 null 1 null
2、 2002 M 1 9 . ) N Nm,5 T (7) :K= 2null 2null4null 4null6= 6null6453bv = “ 9 K 9 Z E L ad L B T V 5bV 5null B Z E v = “ 9 Logistic w L E TK 9 EK estimatemethod B Z Eregressionmethod K a r% “ R2 Q iterationtimes E L 5null6845 5null0466 0null1210 0null9900three- point lineard L 5null6326 6null0507 0nul
3、l1471 0null9975 12nonlinear E L 5null6305 5null3369 0null1320 0null9971four- point lineard L 5null6326 6null0507 0null1471 0null9975 12nonlinear. E L 6null6453 4null4015 0null0873 0null9304yielding point lineard L 5null6326 6null0507 0null1471 0null9975 17nonlinearnull null E | t= 10a40a70d; E | t=
4、10a40a50a80dbV V , K 9 Z E L ad L B 9 d 9 vl ( M t :(1) 3 K 9 Z E ( Y V L B p 6 ,/ ) Kaaar 9 S d L B T # d 9 ( % “ ) B , . E Q K , L 1 M 9 | !t| v V 7 K 9 V v # ;(2)3 K 9 Z E Kaaar 9 d L B T v l : E E . E ,y N 9 K E K ;(3) E Z E % “ R2 v l :d L B E E . E ,V d L B E K , E Q , E a. E Q b 5 N , % “ v (
5、 L 1 ( 0null97 , L 2 ( 0null93 ),y 7 4 E Z E Logistic B Z M L E z b3a) (1) 9 K H , E . E 1 p 1 4 L , E 1 p 13 L b(2) 9 K H , E 1 E I n 1 ,y 7 | L b E a L f , E 5 v a f , E eL b E a E 1 a 1 M ( S a ) ( ) f , Y 9 V Y V L w L 2, 4 E ;7 . E 5 N K , 1 p L 1 M 9 | !t| l b P | !t|i d l H ,. E g V v K | S ,
6、7 | !t| l H ,K 9 45Logistic w L E Z E E b3 K 9 Z E , V j L = f 7 , B f / E D b(3) L B Z E p Logistic e L L + , O d 9 B * L B Z # B “ A _ Z E , Logistic L M 9 3 Y , 13b7 d L B V V Logistic K l = 9 , 1 L B E , E , A 5 S ,7 O S K T H y b(4)K z Z E ,5 L B p Logistic 3 9 i O A _ , N 9 S d L B , “ V K Q 3
7、 K DE b I D 1 null ? , ; C ,o . F .0null618E 3 w L E J. j ,1984,3:59 63.2 null , U .C 3 d 9 M. g : ,1985,445 449.3 null Z 7 , y .d 9 s M. : S ,1987,267 273.4 null l , .Logistic w L # 3 s J. ,1993,8(3):81 86.5 null c , * ,f B . a r 0 o Y # s J. 3 ,1995,15(1):66 71.6 null m ,f 3 . X 3 + Y J. 3 ,1995,1
8、5(2):214 219.7 null Z . C 3 ? Z J. 3 ,1995.14(2):70 75.8 null f N k ; , X V . 3 $ E % ) h J. S ,1998,34(4):123 128.11null f L , W 6 . l Logistic w # J. 2 ,1999,19(1):48- 51.12 null SwedaT, KoideT.Applicabilityofgrowthequationsofthegrowthoftreesinstemradius(I):Applicationtowhitespruce. J. Jap. For. S
9、ci., 1981,61:321 329.13null . Z E M. q : v ,1996,206 226.14null . L B s Z E M. Z : Z S / ,1990,48 50.Study on the fitting methods of logistic curveYIN Zuonullyun(GuangdongForest Research Institute, Guangzhounull 510520,China)Abstract:Logisticcurvehas far- rangingpracticability. Theformulaestimatingp
10、arameterK ofthemodelwith3-point methodwas deduced. In this paper, theother two new methods estimating K valueof the Logistic curveequation, 4- point method and yielding point method were put forward. After K valuewas evaluated, theothertwo coefficients a, r were estimated with linear regression. Wit
11、h two quoted examples discussed, the resultsindicatedthat threemethods estimatingK valueall makegoodfitting precision, especially 4- point method best.Moreover, Nonlinear regression applying above- mentioned estimate values of K, a, r as starting points, easilygivesthesecoefficients best evaluation.Key word:logisticcurve; 3- point method; 4- point method; yielding point method; regression46null null null null null null null null null null null null null null null d 9 5 null null null null 21 null 1 null 2002 M 1
