Nonlinear approximation method of vehicle velocity V t and statistical population of experimental cases.

In car crash analysis three calculations methods can be distinguished: analytical (Campbell, McHenry, Strother, Prasad, Crash3 etc.), comparative and graphical (Lindquist et al., 2003; Prasad, 1990; Sharma et al., 2007; Wach and Unarski, 2006; Żuchowski, 2015) [15,23,25,29,33]. The number and reliability of these methods, in reference to modern vehicles and their structure, may lead to unclear conclusions. This issue proves to be significant, especially due to slight modifications of both input parameters and input data for the analysis of the method. This may give substantially different answers to the questions asked in court. After a thorough analysis of this problem, a new analytical method was devised, based on a new input database - NHTSA, which shows great improvement in the accuracy of obtained results. The aim of this paper is to prove that the nonlinear method is a very effective tool for processing experimental data in a large enough number of cases.