这个问题看起来相当微不足道--数据中的错误如此之少，以至于它看起来像是人为构造的，这是可疑的。这可能是出于演示的目的，也可能是在应用程序中“左手不与右手交谈”，参数没有正确地从一个地方传递到另一个地方。
假设你是真诚的，你的拟合基本上是“完美的”，你的曲线基本上是彼此精确的垂直反射;并且估计I显示基本上等于拟合参数：

import matplotlib.pyplot as plt
import numpy as np
from scipy.optimize import curve_fit

time, decaying, raising = np.array((
    0.0000, 1.0000, 0.0000, 0.1000, 0.9900, 0.0100, 0.2000, 0.9802, 0.0198, 0.3000, 0.9704, 0.0296, 0.4000, 0.9608,
    0.0392, 0.5000, 0.9512, 0.0488, 0.6000, 0.9418, 0.0582, 0.7000, 0.9324, 0.0676, 0.8000, 0.9231, 0.0769, 0.9000,
    0.9139, 0.0861, 1.0000, 0.9048, 0.0952, 1.1000, 0.8958, 0.1042, 1.2000, 0.8869, 0.1131, 1.3000, 0.8781, 0.1219,
    1.4000, 0.8694, 0.1306, 1.5000, 0.8607, 0.1393, 1.6000, 0.8521, 0.1479, 1.7000, 0.8437, 0.1563, 1.8000, 0.8353,
    0.1647, 1.9000, 0.8270, 0.1730, 2.0000, 0.8187, 0.1813, 2.1000, 0.8106, 0.1894, 2.2000, 0.8025, 0.1975, 2.3000,
    0.7945, 0.2055, 2.4000, 0.7866, 0.2134, 2.5000, 0.7788, 0.2212, 2.6000, 0.7711, 0.2289, 2.7000, 0.7634, 0.2366,
    2.8000, 0.7558, 0.2442, 2.9000, 0.7483, 0.2517, 3.0000, 0.7408, 0.2592, 3.1000, 0.7334, 0.2666, 3.2000, 0.7261,
    0.2739, 3.3000, 0.7189, 0.2811, 3.4000, 0.7118, 0.2882, 3.5000, 0.7047, 0.2953, 3.6000, 0.6977, 0.3023, 3.7000,
    0.6907, 0.3093, 3.8000, 0.6839, 0.3161, 3.9000, 0.6771, 0.3229, 4.0000, 0.6703, 0.3297, 4.1000, 0.6637, 0.3363,
    4.2000, 0.6570, 0.3430, 4.3000, 0.6505, 0.3495, 4.4000, 0.6440, 0.3560, 4.5000, 0.6376, 0.3624, 4.6000, 0.6313,
    0.3687, 4.7000, 0.6250, 0.3750, 4.8000, 0.6188, 0.3812, 4.9000, 0.6126, 0.3874, 5.0000, 0.6065, 0.3935, 5.1000,
    0.6005, 0.3995, 5.2000, 0.5945, 0.4055, 5.3000, 0.5886, 0.4114, 5.4000, 0.5827, 0.4173, 5.5000, 0.5769, 0.4231,
    5.6000, 0.5712, 0.4288, 5.7000, 0.5655, 0.4345, 5.8000, 0.5599, 0.4401, 5.9000, 0.5543, 0.4457, 6.0000, 0.5488,
    0.4512, 6.1000, 0.5434, 0.4566, 6.2000, 0.5379, 0.4621, 6.3000, 0.5326, 0.4674, 6.4000, 0.5273, 0.4727, 6.5000,
    0.5220, 0.4780, 6.6000, 0.5169, 0.4831, 6.7000, 0.5117, 0.4883, 6.8000, 0.5066, 0.4934, 6.9000, 0.5016, 0.4984,
    7.0000, 0.4966, 0.5034, 7.1000, 0.4916, 0.5084, 7.2000, 0.4868, 0.5132, 7.3000, 0.4819, 0.5181, 7.4000, 0.4771,
    0.5229, 7.5000, 0.4724, 0.5276, 7.6000, 0.4677, 0.5323, 7.7000, 0.4630, 0.5370, 7.8000, 0.4584, 0.5416, 7.9000,
    0.4538, 0.5462, 8.0000, 0.4493, 0.5507, 8.1000, 0.4449, 0.5551, 8.2000, 0.4404, 0.5596, 8.3000, 0.4360, 0.5640,
    8.4000, 0.4317, 0.5683, 8.5000, 0.4274, 0.5726, 8.6000, 0.4232, 0.5768, 8.7000, 0.4190, 0.5810, 8.8000, 0.4148,
    0.5852, 8.9000, 0.4107, 0.5893, 9.0000, 0.4066, 0.5934, 9.1000, 0.4025, 0.5975, 9.2000, 0.3985, 0.6015, 9.3000,
    0.3946, 0.6054, 9.4000, 0.3906, 0.6094, 9.5000, 0.3867, 0.6133, 9.6000, 0.3829, 0.6171, 9.7000, 0.3791, 0.6209,
    9.8000, 0.3753, 0.6247, 9.9000, 0.3716, 0.6284, 10.0000, 0.3679, 0.6321
)).reshape((-1, 3)).T

def decay_model(x: np.ndarray, b: float) -> np.ndarray:
    return np.exp(-x/b)

def raise_model(x: np.ndarray, b: float) -> np.ndarray:
    return 1 - np.exp(-x/b)

b_est_decay = -time[-1] / np.log(decaying[-1])
b_est_raise = -time[-1] / np.log(1 - raising[-1])
(b_decay,), _ = curve_fit(f=decay_model, xdata=time, ydata=decaying, p0=b_est_decay)
(b_raise,), _ = curve_fit(f=raise_model, xdata=time, ydata=raising, p0=b_est_raise)
print('b (decay) =', b_decay)
print('b (raise) =', b_raise)

fig, ax = plt.subplots()
ax.plot(time, decaying, label='decay (experiment)')
ax.plot(time, raising, label='raise (experiment)')
ax.plot(time, decay_model(time, b_decay), label='decay (fit)', linestyle='--')
ax.plot(time, raise_model(time, b_raise), label='raise (fit)', linestyle='--')
ax.legend()
plt.show()

个字符
x1c 0d1x的数据