Bayesian Filtering and Smoothing to Measure Damper Characteristics
Abstract: During the development of high performance damper products for vehicle applications, it is essential to measure the dampers’ characteristics to ensure their performance. Some of the measurements are performed in dynamometers. The dynamometer actuates the damper with a predefined position signal and measures the actual position and the resulting force. These measurements suffer from noise; the problem is especially bad if the signals are studied with respect to velocity. This is due to that the velocity is not directly measurable; it must be calculated from position by differentiation, which amplifies the noise. The purpose of this work was to improve these measurements by doing acomparison of different Bayesian filters and smoothers. The comparison included the Kalman filter, the extended Kalman filter, the unscented Kalman filter, the Rauch-Tung-Striebel smoother, the extended Rauch-Tung-Striebel smoother and the unscented Rauch-Tung-Striebel smoother. The smoothers are only applicable in offline applications, while the filters could be used in real time. The filters reduced the noise in the position signal and greatly reduced the noise in the velocity signal. The smoothers showed the same behaviors as the filters, but with much less noise. Only small improvements were visible in the force signal.
AT THIS PAGE YOU CAN DOWNLOAD THE WHOLE ESSAY. (follow the link to the next page)