Filtering Impulse Data
DOI link for Filtering Impulse Data
Filtering Impulse Data book
A vehicle crash is an impulsive event, lasting perhaps ⅛ sec for automobiles. As such, the time histories of physical phenomena like sound, pressure, stress, acceleration, and so on are filled with what would often be called “noise.” These peaks and valleys are not random occurrences, as the term “noise” might imply, but on the other hand, noise is not the phenomenon that causes injury, structural damage, or other important consequences that might be the focus of interest. Rather, noise can be the trees that obscure the forest we are trying to understand. Our understanding depends on our ability to separate out the noise from the underlying data and trends in the data. This is where filters come in.
Reconstructionists encounter filtering when they analyze crash test data, which form the technical underpinning of much of the work they do. It is also the case that the vehicles themselves, in this era of digital electronics, are constantly analyzing data streams in an attempt to discern what is happening and what actions to take. Here, the need to see through the noise and get to the important information is vital to the safety and performance of vehicles. In this case, however, the filtering must be done in real time. Increasingly, reconstructionists are expected to understand how the filters performed in the crashes they are investigating.
Topics in Chapter 11 include analog filters, filter order, Bode plots, filter types, digital filters, Finite Impulse Response (FIR) filters, Infinite Impulse Response (IIR) filters, use of the Z-transform, an example of finding the difference equation from the transfer function, bilinear transforms, digital filters for airbag applications, and an example of a digital filter in an airbag sensor.