根据情况可以选择分段拟合或者多项式拟合，使用多项式拟合的话，结果如下：
a0 = -88.616
a1 = 0.199288
a2 = 0.000149994
x0 = 435  counter y = 26.4568  true y = 25  error = 5.82737%
x1 = 462  counter y = 35.4703  true y = 35  error = 1.34375%
x2 = 476  counter y = 40.2301  true y = 40  error = 0.575153%
x3 = 489  counter y = 44.7025  true y = 45  error = -0.661166%
x4 = 502  counter y = 49.2256  true y = 50  error = -1.54883%
x5 = 607  counter y = 87.6168  true y = 90  error = -2.64796%
x6 = 648  counter y = 103.506  true y = 105  error = -1.42327%
x7 = 690  counter y = 120.305  true y = 120  error = 0.253921%
x8 = 702  counter y = 125.202  true y = 125  error = 0.161324%
x9 = 763  counter y = 150.762  true y = 150  error = 0.50827%
x10 = 810  counter y = 171.218  true y = 170  error = 0.716542%
x11 = 853  counter y = 190.513  true y = 190  error = 0.270211%
x12 = 874  counter y = 200.138  true y = 200  error = 0.0691344%
x13 = 897  counter y = 210.832  true y = 210  error = 0.395992%
x14 = 917  counter y = 220.259  true y = 220  error = 0.117779%
x15 = 936  counter y = 229.326  true y = 230  error = -0.292864%
x16 = 964  counter y = 242.886  true y = 240  error = 1.20256%
x17 = 992  counter y = 256.681  true y = 260  error = -1.27652%
x18 = 1008  counter y = 264.669  true y = 265  error = -0.124732%

  total error = 1.6584

看来你的情况使用分段（线性或者曲线）拟合会更合适。

用的什么传感器?  标准传感器都有数据表格或者曲线方程了