请dingd主任帮忙
本帖最后由 fhn 于 2010-11-13 10:19 编辑主任,你好,我用的是1stopt1.5,但是sharedmodel好像不好用,请问是版本的问题吗?谢谢。如果是版本问题,您能帮我算下下面的程序吗?谢谢。
Title "eps";
Parameter P1,P2,P3,P4,P5,P6,P7,P8,P9,P10,P11;
Variable x,y1,y2 ;
SharedModel ;
Function y1=9.4-P1*10000*1.6*1.6/(0.22*9.11*8.854*(39.4784*x*x+P2*P2))-P3*(39.4784*x*x-P4*P4)/((39.4784*x*x-P4*P4)*(39.4784*x*x-P4*P4)+39.4784*x*x*P5*P5)-P6*(39.4784*x*x-P7*P7)/((39.4784*x*x-P7*P7)*(39.4784*x*x-P7*P7)+39.4784*x*x*P8*P8)-P9*(39.4784*x*x-P10*P10)/((39.4784*x*x-P10*P10)*(39.4784*x*x-P10*P10)+39.4784*x*x*P11*P11);
y2=P1*10000*1.6*1.6*P2/(0.22*9.11*8.854*(248.05*x*x*x+6.283*x*P2*P2))+P3*6.283*x*P5/((39.4784*x*x-P4*P4)*(39.4784*x*x-P4*P4)+39.4784*x*x*P5*P5)+P6*6.283*x*P8/((39.4784*x*x-P7*P7)*(39.4784*x*x-P7*P7)+39.4784*x*x*P8*P8)+P9*6.283*x*P11/((39.4784*x*x-P10*P10)*(39.4784*x*x-P10*P10)+39.4784*x*x*P11*P11);
Data;
0.30732 -44.811 7.0532
0.32195 -45.63 8.1728
0.33658 -45.925 9.9178
0.35122 -45.6 11.737
0.36585 -44.456 13.553
0.38049 -42.839 15.406
0.39512 -41.166 16.912
0.40975 -39.403 17.786
0.42439 -37.577 18.25
0.43902 -36.099 18.436
0.45366 -35.11 18.094
0.46829 -34.275 17.236
0.48292 -33.44 16.313
0.49756 -32.838 15.579
0.51219 -32.544 14.928
0.52683 -32.384 14.326
0.54146 -32.278 13.842
0.55609 -32.165 13.403
0.57073 -31.836 12.999
0.58536 -31.252 12.833
0.6 -30.673 12.896
0.61463 -30.16 12.839
0.62927 -29.412 12.548
0.6439 -28.377 12.309
0.65853 -27.414 12.179
0.67317 -26.653 11.829
0.6878 -25.837 11.174
0.70244 -24.947 10.512
0.71707 -24.299 9.9076
0.7317 -23.945 9.1251
0.74634 -23.647 8.1958
0.76097 -23.427 7.3994
0.77561 -23.513 6.7239
0.79024 -23.815 5.9606
0.80487 -24.07 5.2
0.81951 -24.308 4.6718
0.83414 -24.681 4.316
0.84878 -25.092 3.9772
0.86341 -25.399 3.7151
0.87804 -25.663 3.5835
0.89268 -25.891 3.4735
0.90731 -25.979 3.3843
0.92195 -25.993 3.4116
0.93658 -26.076 3.42
0.95121 -26.074 3.2324
0.96585 -25.805 3.0472
0.98048 -25.523 3.1081
0.99512 -25.549 3.1731
1.0098 -25.668 2.8971
杯具,沦为做题的了? 回复 2 # dingd 的帖子
dingd主任,不是这样的,我想知道是否是因为版本问题,如果是的话,我正想办法购买4.0版本,让您帮着算算主要是我想看看1stopt和origin算的差别。 在拟合方面,不论是易用性还是拟合效果方面,Origin都远不如1stOpt。
网上的1.5 PJ版似乎无法运行这种模式。
下面两张图式不考虑参数范围限制的拟合结果,整体均方差为RMSE: 0.401694。用任何其它软件可以说都难以达到这一效果。当然,你的参数范围是有限制的,自己再试试吧,计算量不小。 回复 4 # dingd 的帖子
感谢dingd主任,可否告诉我拟合出来的参数数值,我好和origin有个比较。 "eps"
迭代数: 41
计算用时(时:分:秒:微秒): 00:03:55:96
优化算法: 麦夸特法(Levenberg-Marquardt) + 通用全局优化法
计算结束原因: 达到收敛判断标准
均方差(RMSE): 0.146220094596482
残差平方和(SSE): 1.0476354871264
相关系数(R): 0.999775060161714
相关系数之平方(R^2): 0.999550170921358
决定系数(DC): 0.999550168454576
卡方系数(Chi-Square): -0.0188600726713256
F统计(F-Statistic): 104437.13015431
参数 最佳估算
---------- -------------
P1 0.182162725931055
P2 0.000231106083485798
P3 199.100630560448
P4 3.38571700571544
P5 2.15215415935754
P6 461.10843277154
P7 5.40874270283985
P8 3.28572383389455
P9 562.236402878342
P10 1.74631891174537
P11 2.19802342782235
====== 结果输出 =====
No 实测值y1 计算值y1
1 -44.811 -44.7550437
2 -45.63 -45.7663633
3 -45.925 -45.9568601
4 -45.6 -45.4367139
5 -44.456 -44.3596515
6 -42.839 -42.8928111
7 -41.166 -41.2047689
8 -39.403 -39.4483770
9 -37.577 -37.7562817
10 -36.099 -36.2387942
11 -35.11 -34.9705970
12 -34.275 -33.9938573
13 -33.44 -33.3088290
14 -32.838 -32.8784290
15 -32.544 -32.6351151
16 -32.384 -32.4916616
17 -32.278 -32.3586545
18 -32.165 -32.1588538
19 -31.836 -31.8381059
20 -31.252 -31.3706123
21 -30.673 -30.7556379
22 -30.16 -30.0145929
23 -29.412 -29.1807300
24 -28.377 -28.2964223
25 -27.414 -27.4039832
26 -26.653 -26.5435411
27 -25.837 -25.7523563
28 -24.947 -25.0593190
29 -24.299 -24.4882925
30 -23.945 -24.0541346
31 -23.647 -23.7642735
32 -23.427 -23.6188634
33 -23.513 -23.6098977
34 -23.815 -23.7226527
35 -24.07 -23.9364559
36 -24.308 -24.2264578
37 -24.681 -24.5645024
38 -25.092 -24.9225255
39 -25.399 -25.2726974
40 -25.663 -25.5907467
41 -25.891 -25.8563748
42 -25.979 -26.0535485
43 -25.993 -26.1718886
44 -26.076 -26.2054142
45 -26.074 -26.1526920
46 -25.805 -26.0155871
47 -25.523 -25.7989811
48 -25.549 -25.5091773
49 -25.668 -25.1530234
页:
[1]