Active Suspension: Future Lessons from the Past
Dr. Dan Williams
Director, ADAS and Autonomy
Commercial Vehicle Technologies ZF Friedrichshafen
Active suspension was a topic of great research interest near the end of the last century.
Ultimately broad bandwidth active systems were found to be too expensive in terms of both energy and financial cost. This past work, developing the ultimate vehicle suspension, has relevance for today’s vehicle designers working on more efficient and effective suspension systems for practical vehicles. From a control theorist’s perspective, it provides an interesting case study in the use of “practical” knowledge to allow “better” performance than predicted by theoretically optimal linear controllers.
A brief history of active suspension and the specific founding contributions of William F. Milliken will be introduced. Milliken collaborated with Peter Wright, David Williams, and others at Lotus to develop Lotus Modal Control. In a parallel effort, Dean Karnopp and Don Margolis presented the notion of inertial (Skyhook) damping. These notions will be compared, the combination of these efforts will be discussed, and eventual vehicle results presented.
Most of the contemporary literature treated active suspension as a theoretical vibration isolation problem, but handling improvements from active suspension were even more impressive. Handling and actual hardware considerations motivated the confluence of both primary approaches. This innovative implementation of a control algorithm preserving features of both Lotus modal control and inertial damping is discussed, and compared with theoretical optimal controllers. Finally, a surprising fundamental performance limit of the modal inertial damping algorithm is discussed, and a solution presented.
Date(s) - 03/01/2018
4:00 pm - 5:00 pm
1062 Bainer Classroom
Categories No Categories