Comparison between a model-free and model-based controller of an automotive semi-active suspension system

Juan C. Tudón-Martínez, Ruben Morales-Menendez, Ricardo Ramírez-Mendoza, Olivier Sename, Luc Dugard

Research output: Contribution to conferencePaper

4 Citations (Scopus)


A comparison between two control strategies of automotive semi-active suspension systems is presented by using a pick-up truck model in CarSim™; the controllers are based on different frameworks. The Linear Parameter Varying (LPV) controller, considered as model-based controller, includes the constraints of the semi-active damper by using two scheduling parameters; while, the Frequency Estimation-Based (FEB) controller only requires measurements of a wheel-station for defining the damping force according to the objective controls, i.e. is free of a vehicle and actuator model. Each wheel-station has an independent controller and, experimental data are used to model an MR damper in each quarter of vehicle. A running test on the road based on ISO 8606 is used to compare the performance in comfort and road holding of both controllers. Simulation results in CarSim™ show that the LPV approach has the best comfort performance; in comparison with the passive suspension system, the pitch angle is reduced 36% and the vertical chassis displacement 24%. However, the FEB controller is better for road holding: a lower suspension deflection with a tire compression reduced 30%.

Original languageEnglish
Number of pages6
Publication statusPublished - 1 Jan 2013
Externally publishedYes
EventIFAC Proceedings Volumes (IFAC-PapersOnline) -
Duration: 1 Jan 2010 → …


ConferenceIFAC Proceedings Volumes (IFAC-PapersOnline)
Period1/1/10 → …

Bibliographical note

Copyright 2018 Elsevier B.V., All rights reserved.

All Science Journal Classification (ASJC) codes

  • Control and Systems Engineering


Dive into the research topics of 'Comparison between a model-free and model-based controller of an automotive semi-active suspension system'. Together they form a unique fingerprint.

Cite this