2楼: Originally posted by dingd at 2014-08-12 22:44:06
不强调Matlab的话，1stOpt倒是方程简单方便：

VarConstant T=;
Function y=(a+b*T)*(x/(1-x))^c;
Data;
X=;
Y=;
Data;
X=;
Y=;
Data;
X=;
Y=;


均方差(RMSE):0.814290243914772
残差平方和(SSE): ...

VarConstant T=[278,293,313];
Function y=(a+b*T)*(x/(1-x))^c;
Data;
X=[0.113 0.245 0.336 0.431 0.589 0.635 0.757 0.771 0.851 0.877];
Y=[4.94 7.20 8.16 9.22 10.89 12.74 17.80 18.38 24.72 28.70];
Data;
X=[0.113 0.231 0.331 0.432 0.544 0.591 0.725 0.755 0.817 0.851];
Y=[4.58 6.67 7.62 8.88 10.15 10.87 14.55 17.39 21.83 24.65];
Data;
X=[0.112 0.201 0.316 0.432 0.484 0.532 0.658 0.747 0.794 0.823];
Y=[3.27 4.99 5.57 7.48 7.90 9.02 12.20 15.33 17.11 19.34];

4楼: Originally posted by 1413965491 at 2014-08-13 09:51:48
使用matlab中的三维拟合曲线就行，自变量看成x和t，因变量看成y