ON-BOARD PREDICTION OF POWER CONSUMPTION IN AUTOMOBILE ACTIVE SUSPENSION SYSTEMS—I: PREDICTOR DESIGN ISSUES

✍ Scribed by R. Ben Mrad; S.D. Fassois; J.A. Levitt; B.I. Bachrach

• Publisher: Elsevier Science
• Year: 1996
• Tongue: English
• Weight: 615 KB
• Volume: 10
• Category: Article
• ISSN: 0888-3270
• DOI: 10.1006/mssp.1996.0010

✦ Synopsis

On-board prediction of automobile active suspension power consumption is considered important in securing fast engine response and avoiding potential vehicle hesitation or surge. In this study the problem is considered within the context of a broad-bandwidth hydraulic active suspension design, with the following main objectives: (a) power consumption prediction feasibility assessment, (b) examination of predictor design options, and (c) achievable prediction accuracy assessment.

In this part I the first two objectives are pursued based on a detailed, non-linear, mathematical system model. The power consumption signal is found to be non-stationary, but stationarity is shown to be possible to impose through a moving-average transformation. Predictor design options are then considered: first, the wheel vertical acceleration is shown to be ineffective as a prediction leading indicator; second, the relative merits of time/step-recursive and time-recursive predictor realizations are examined, and the latter are shown to be presently preferable from the point of view of computational complexity, error propagation, and predictor stability; third, indirect (standard) vs. direct predictor types are considered, and a direct predictor is developed based on a novel estimator characterised by features important for on-board implementation.