Identification of cascade water tanks using a PWARX model. (June 2018)

• Record Type: Journal Article
• Authors: Mattsson, Per; Zachariah, Dave; Stoica, Petre
• Is Part Of: Mechanical systems and signal processing. Volume 106(2018)
• Page Start: 40
• Page End: 48
• Publication Date: 2018-06
• Subjects: System identification -- Nonlinear systems -- Piecewise ARX models; Structural dynamics -- Periodicals; Vibration -- Periodicals; Constructions -- Dynamique -- Périodiques; Vibration -- Périodiques; Structural dynamics; Vibration; Periodicals; 621
• DOI: 10.1016/j.ymssp.2017.12.027
• Languages: English
• ISSNs: 0888-3270

Abstract

Highlights: Develops a method for PWARX identification. Applies the method to a cascade tank system. Incorporate saturation in model structure. Abstract: In this paper we consider the identification of a discrete-time nonlinear dynamical model for a cascade water tank process. The proposed method starts with a nominal linear dynamical model of the system, and proceeds to model its prediction errors using a model that is piecewise affine in the data. As data is observed, the nominal model is refined into a piecewise ARX model which can capture a wide range of nonlinearities, such as the saturation in the cascade tanks. The proposed method uses a likelihood-based methodology which adaptively penalizes model complexity and directly leads to a computationally efficient implementation.