Generation of ride comfort index

Abstract

The main objective of this project is to obtain a ride comfort index of a vehicle. This parameter shall be representative of the vibration received by the human body during the ride. The comfort, as we understand, is the sensation which comes from vehicle dynamics and not from the luxury cabin. Initially the comfort evaluation of the cars was done subjectively. The driver-tester exposed his own feelings when he tested the car on a random circuit and compared with other vehicles that he tested before. This method presents several problems: Mainly the indeterminacy caused by the different sensations showed by diverse pilots who tested the same vehicle; and also the fail maintain of a database, the information of the models tested is sometimes lost and distorts, and finally the perception can change by the time. Therefore the method that we built was to resolve these deficiencies. With this objective we designed a work routine consisting of a first phase of data collection. The vehicle was instrumented with triaxial accelerometers at different locations of the vehicle, but we were focused on the steering wheel and the seat: the contact points between the machine and the human body. The process continues with two types of test: a random test and a deterministic. Both are representative of the full spectrum of driving, from driving on the open road to the impacts. The IDIADA [16] tracks were used for this purpose. In order to generate the comfort index a computer code was created and implemented in Diadem 2011 v11.3 using a similar Visual Basic programming language. That program generates four values to index: the comfort of the steering wheel in the random and deterministic tests, and seating comfort in the random and deterministic tests. A similar algorithm is used in these four studies. The first step is a pre-processing (average, cuts, …) based on the original signal recorded on the track. The second step will use a specially designed filter for each study, in order to obtain a signal of perceived comfort. The third one is obtaining a representative value of the filtered signal. The fourth is getting initial indexes from regressions adjusted by a comparative. Finally, the whole indexes obtained were mixed to generate new ones that are easily to understand. The objective results obtained in the operational simulation, contrasted with the subjective results generated by SEAT professional drivers, were fully accepted. The main conclusion is that it is possible to generate objective measures to rate the comfort of the vehicles. We have also fulfilled all the requirements of the automotive industry sector.