Weighted expectation-maximisation algorithm for parameter estimation of misspecified dynamical models: application to vehicle dynamics. Issue 20 (2021)

• Record Type: Journal Article
• Authors: Matz, Jules; Birouche, Abderazik; Mourllion, Benjamin; Bouziani, Fethi; Basset, Michel
• Is Part Of: IFAC-PapersOnLine. Volume 54:Issue 20(2021)
• Page Start: 8
• Page End: 13
• Publication Date: 2021
• Subjects: Identification -- nonlinear misspecified model -- EM algorithm -- vehicle dynamics; Automatic control -- Periodicals; 629.805
• DOI: 10.1016/j.ifacol.2021.11.145
• Languages: English
• ISSNs: 2405-8963

Abstract

Abstract: The problem of parameter estimation for nonlinear state-space models is addressed using the expectation-maximisation algorithm. Model states and parameters are iteratively estimated using cubature Kalman smoothing and maximum a posteriori estimation. A modification to this technique is proposed by weighting measurement samples so the algorithm equally tries to approximate all system dynamics, even those poorly represented in the measurements. The method is applied to parameter estimation of a vehicle dynamics model.