Input design for Bayesian identification of non-linear state-space models

TL;DR

This paper introduces an algorithm for designing optimal inputs to improve Bayesian identification of non-linear state-space models, using a tractable optimization approach based on Markov chain parametrization.

Contribution

It presents a novel input design method that minimizes the posterior Cramér-Rao bound for non-linear state-space models using Markov chain parametrization.

Findings

Effective input design improves model identification accuracy.

The method is demonstrated through a simulation example.

Optimization reduces computational complexity.

Abstract

We propose an algorithm for designing optimal inputs for on-line Bayesian identification of stochastic non-linear state-space models. The proposed method relies on minimization of the posterior Cram\'er Rao lower bound derived for the model parameters, with respect to the input sequence. To render the optimization problem computationally tractable, the inputs are parametrized as a multi-dimensional Markov chain in the input space. The proposed approach is illustrated through a simulation example.