Abstract: Abstract：To improve the precision in calculation and state prediction of realtime data near the endpoints of measurement sequence, the cubicpolynomial firstorderderivativetheorem (CPFODT) is put forward and proved, and a new algorithm based on CPFODT is researched for calculating the first order derivative of realtime data. How the sampling length, measure error bounds and sampling interval affect the results is shown by simulation calculation. The errors are analyzed, and some basic principles of parameter selection are given. Verification of the algorithm is made by measurement, and the results indicate that it is very robust and reliable. Not only can this algorithm improve the precision in calculation and state prediction of realtime data near the endpoints of measurement sequence, but also some interpolation algorithms can derive from it. So, it can provide high accurate data.

Key words: Keywords： cubic polynomial, firstorder derivative, realtime data processing, robustness