Non-linear State-space Model Identification from Video Data using Deep Encoders

I., Beintema, Gerben; Tóth, Roland; Schoukens, Maarten

doi:10.1016/j.ifacol.2021.08.442

Cited by 6 publications

(1 citation statement)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…with τ 0 ≥ 0 burn in parameter, the estimator can be seen as a nonlinear extension linear subspace identification Beintema et al (2021) or as a generalization of the multiple shooting method (Ribeiro et al, 2019). Furthermore, due to the independence of the loss function on each section, the proposed method allows for i) computational speedup by utilizing modern parallelization methods and ii) the utilization of batch optimization methods (e.g.…”

Section: Model Estimationmentioning

confidence: 99%

Identification of the nonlinear steering dynamics of an autonomous vehicle

Rödönyi¹,

I.²,

Tóth³

et al. 2021

Preprint

View full text Add to dashboard Cite

Automated driving applications require accurate vehicle specific models to precisely predict and control the motion dynamics. However, modern vehicles have a wide array of digital and mechatronic components that are difficult to model, manufactures do not disclose all details required for modelling and even existing models of subcomponents require coefficient estimation to match the specific characteristics of each vehicle and their change over time. Hence, it is attractive to use data-driven modelling to capture the relevant vehicle dynamics and synthesise model-based control solutions. In this paper, we address identification of the steering system of an autonomous car based on measured data. We show that the underlying dynamics are highly nonlinear and challenging to be captured, necessitating the use of data-driven methods that fuse the approximation capabilities of learning and the efficiency of dynamic system identification. We demonstrate that such a neural network based subspace-encoder method can successfully capture the underlying dynamics while other methods fall short to provide reliable results.

show abstract

Section: Model Estimationmentioning

confidence: 99%