| 摘要 |
PURPOSE:To detect peak positions stably in a short time even when the property of an object characteristic curve is unknown by selecting plural measurement points, and approximating measurement data regressively and calculating the peak positions. CONSTITUTION:Plural measurement points are selected according to reference points Xo and Yo and the measurement data Xi and Yi are approximated regressively to calculate the peak positions Xp and Yp. Then, the measurement points are changed in order according to the obtained peak positions Xp and Yp to repeat measurement and regressive approximation. Then the measurement is finished to convergence condition that the deviations of the obtained peak positions Xp and Yp become smaller that constant values Kx and Ky. Further, when convergence is not obtained even after the measurement is repeated specific times, an alarm is generated and the measurement is finished. Consequently, the peak positions are detected stably in a short time unless the property of the object characteristic curve is known.
|
