“…While system characterisation is concerned primarily with setting up mathematical models to represent input–output relationships, system identification deals with the choice of a specific model from a class of models which is mathematically equivalent to a given physical system. System identification is an important approach to model systems and has already been widely researched in practical and theoretical fields, for example, networked systems with unknown parameters and randomly missing outputs [6], modelling of the Duffing oscillator [7], multi‐degree‐of‐freedom (MDOF) systems [8], industrial multistage compressed air system [9], multiscale spatio‐temporal dynamical systems [10], multi‐step ahead predictions [11], tracking control [12], modelling [13] etc. In practice, when linear models fail, non‐linear models, because of their better approximation capabilities, appear to be powerful tools for modelling practical situations.…”