In this article an algorithm is introduced as a method to detect and prevent erratic clamping conditions during the electric seat adjustment based only on rev signals of the adjusting motors. Features involving a clamping condition are characterized by the wavelet transform of the adjustment motor rev signal sent. Mapping of the symptoms to support the decision whether a clamping condition has occured or not is performed by a neural network.
Print ISSN: 0178-2312
Volume: 55, 12/2007
Pages: 608 - 614