β Scribed by JOHN G. CHERNG; ALI GOKTAN; MARK FRENCH; YI GU; ANIL JACOB
No coin nor oath required. For personal study only.
Shaker tables are commonly used in laboratories for automotive vehicle component testing to study durability and acoustics performance. An example is development testing of car seats. However, it is di$cult to repeat the measured road data perfectly with the response of a shaker table as there are basic di!erences in dynamic characteristics between a #exible vehicle and substantially rigid shaker table. In addition, there are performance limits in the shaker table drive systems that can limit correlation. In practice, an optimal drive signal for the actuators is created iteratively. During each iteration, the error between the road data and the response data is minimised by an optimising algorithm which is generally a part of the feed back loop of the shake table controller.
This study presents a systematic investigation to the errors in time and frequency domains as well as joint time}frequency domain and an evaluation of di!erent digital signal processing techniques that have been used in previous work. In addition, we present an innovative approach that integrates the dynamic characteristics of car seats and the human body into the error-minimising iteration process. We found that the iteration process can be shortened and the error reduced by using a weighting function created by normalising the frequency response function of the car seat. Two road data test sets were used in the study.
π SIMILAR VOLUMES