Set-values filtering for discrete time-delay genetic regulatory networks with time-varying parameters

We propose a new discrete-time model for genetic regulatory networks (GRNs) in this paper. The new model includes time-varying degradation rates, translation rates of mRNA, and protein as well as time-delay and bounded external noise, which better approximates the practical GRNs. The aim is to estimate the concentrations of mRNA and protein in the proposed model. Due to the unknown external noise and time-varying parameters in this model, the set-values filtering is adopted. The so-called set-values filtering problem is to find an upper bound for the quadratic estimation error of the filtering dynamics and minimize this bound at each time step provided that the external noise is bounded. A recursive linear matrix inequality (RLMI) based optimization approach is developed to compute the filter gains. The effectiveness of the design method is illustrated via two numerical examples.