The paper deals with the identification of linear systems when they are in a pathway in series with a saturation nonlinearity. The objective is to estimate the parameters of the linear system and to obtain some characterisation of the nonlinearity, using only the input and output signals of the pathway. Both Wiener structures and Hammerstein structures are considered, and it is found that pertu...

Pseudorandom signals have been used in system identification for many years, usually in their binary form, usually for the identification of linear systems and usually because of their correlation propemes.' Pseudorandom binary signals are unsuitable for the identification of nonlinear systems but pseudorandom ternary signals have been used for the purpose, exploiting their higher-order correla...

This paper considers a method for optimal input design in system identification for control. The approach addresses model predictive control (MPC). The objective of the framework is to provide the user with a model which guarantees that a specified control performance is achieved, with a given probability. We see that, even though the system is nonlinear, using linear theory in the input design...

System identification is concerned with obtaining good models from data, i.e. with data driven modeling. In this contribution the aim is to explain and discuss ideas, general approaches and theories underlying identification of linear systems. Identification of linear systems is a nonlinear problem and is prototypical also for many parts of identification of nonlinear systems.

In this paper we develop a method for identifying SISO Wiener-type nonlinear systems, that is, systems consisting of a linear dynamic system followed by a static nonlinearity. Unlike previous techniques developed for Wiener system identification, our approach allows the identification of systems with nonlinearities that are known but not necessarily invertible, continuous, differentiable, or an...

In this paper a grey-box model of a nonlinear dynamical system is constructed. This involves using a Gaussian process to emulate model error the error that arises as a result of flaws in one’s physical-law based model of the system. The work shows how such an approach can be extended towards dynamical systems. Specifically, it is applied to experimental data, obtained from a dynamical system wh...

Abstract: Identification of continuous-time non-linear systems characterised by fractional order dynamics is studied. The Riemann-Liouville definition of fractional differentiation is used. A new identification method is proposed through the extension of Hammerstein-type models by allowing their linear part to belong to the class of fractional models. Fractional models are compact and so are us...

Many different discrete-time recurrent neural network architectures have been proposed. However, there has been virtually no effort to compare these arch:tectures experimentally. In this paper we review and categorize many of these architectures and compare how they perform on various classes of simple problems including grammatical inference and nonlinear system identification.

The investigation of robot-environment interaction is the main aim of the RobotMODIC project at the Universities of Essex and Sheffield. The methods developed under this project model and characterise all aspects relevant to the robot’s operation: modelling of sensor perception (“environment identification” or simulation), sensor modelling, and task modelling. In this paper we describe a new pr...

Developing robust and reliable control code for autonomous mobile robots is difficult, because the interaction between a physical robot and the environment is highly complex, subject to noise and variation, and therefore partly unpredictable. This means that to date it is not possible to predict robot behaviour based on theoretical models. Instead, current methods to develop robot control code ...