Reproduction of kinematics of cars involved in crash events using nonlinear autoregressive models

Vehicle crashworthiness can be assessed by the variety of methods - the most common and direct one is a vehicle crash test. Visual inspection and obtained measurements, such as car acceleration, are used to examine impact severity of an occupant and overall car safety. However, those experiments are complex, time-consuming, and expensive. We propose a method to reproduce car kinematics during a collision using a feedforward neural network to estimate the system by use of nonlinear autoregressive (NAR) models. Specifically, feasibility of applying neural networks with an NAR model to the analysis of experimental data is explored by application to measurements of a vehicle crash test. This model allows us to predict the kinematic responses (acceleration, velocity, and displacement) of a given car during a collision. The major advantage of this approach is that those plots can be obtained without additional teaching of a network.