Dear guest, welcome to this publication database. As an anonymous user, you will probably not have edit rights. Also, the collapse status of the topic tree will not be persistent. If you like to have these and other options enabled, you might ask Pascal Held for a login account.
| This site is powered by Aigaion - A PHP/Web based management system for shared and annotated bibliographies. For more information visit www.aigaion.nl. |
|
Application of support vector machines to discriminate vehicle crash events
| Type of publication: | Mastersthesis |
| Citation: | moewes_application_2007 |
| Year: | 2007 |
| Month: | October |
| School: | School of Computer Science, University of Magdeburg |
| Address: | Magdeburg, Germany |
| Abstract: | This diploma thesis was written within the scope of my work on a project in the Department of Restraint Systems at Siemens VDO Automotive AG in Regensburg, Germany from March 26 until September 28, 2007. The security requirements of motor vehicles in the road traffic rises up gradually nearly every year due to the massively increasing number of cars, the continuously becoming faster automobiles and the steadily growing amount of safety regulations. With the objective of fulfill these requirements, automotive suppliers like Siemens VDO work with the utmost care on the implementation of restraint systems together with motor vehicle manufacturers to protect the driver, passengers, pedestrians and other road users. Nowadays, restraint systems like multistage drivers’ and passengers’ airbags, belt tensioner and side airbags can be found in almost every modern car. Many steps in the prototypical vehicle development are necessary to come up with an absolutely reliable restraint system. The embedded systems have to be adopted and calibrated to every type of vehicle and many different crash scenarios. The crash data usually contain conflicts which are equivalent to some requirements that are hard to fulfill. The discussions with the customers about changes of demands to solve these conflicts commonly take several weeks. A conflict found at the end of a project very often entails the project’s deathblow. In my thesis, I will concentrate on conflict analysis in order to improve and accelerate the calibration process. I will present machine learning algorithms based on support vector machines that try to find conflicts in crash data. |
| Keywords: | Conflict Analysis, Safety-Related Systems, Support Vector Machine |
| Authors | |
| Added by: | [ADM] |
| Total mark: | 0 |
|
Attachments
|
|
|
|
|
Notes
|
|
|
|
|
|
Topics
|
|
