A comprehensive experimental validation of a scaled car-like vehicle: Lateral dynamics identification, stability analysis, and control application. (November 2021)

• Record Type: Journal Article
• Authors: Ribeiro, A.M.; Koyama, M.F.; Moutinho, A.; de Paiva, E.C.; Fioravanti, A.R.
• Is Part Of: Control engineering practice. Volume 116(2021)
• Publication Date: 2021-11
• Subjects: Scaled vehicle -- Parameter identification -- Stability analysis -- Vehicle lateral control -- Car-like robots; Automatic control -- Periodicals; 629.89
• DOI: 10.1016/j.conengprac.2021.104924
• Languages: English
• ISSNs: 0967-0661

Abstract

Abstract: The use of scaled vehicle prototyping is driven through costs, safety, and time effectiveness in evaluating state-of-the-art algorithms, particularly to demonstrate and facilitate application in advanced vehicle control and sensing systems. This paper illustrates the use of a car-like 1:5-scale vehicle as a valid and reliable test-bed platform whose lateral dynamics are similar to those of a full-sized vehicle. The validation scheme is proposed by way of three model-based essential concepts, comprising an identification process, a stability analysis step, and a control design. The identification process is proposed to determine the unknown model parameters necessary to correctly describe the vehicle behavior. The mathematical model, validated through a series of matching experiments, is then used for further analysis and control purposes. The analysis design is intended to identify the vehicle's operating conditions, according to its stability properties. Finally, a lateral control strategy that merges knowledge acquired from these analyses is introduced for a path-following system. For each presented characterization, a wide set of experiments were carried out using the scaled vehicle platform. A persistent agreement between expected (simulated) and achieved (measured) lateral response was obtained. The results demonstrate a systematic consistency regarding the platform test-bed and the behavior described by the mathematical model.