PEMFC Fractional-order Subspace Identification Model

A proton exchange membrane fuel cell (PEMFC) is a new type of hydrogen fuel cell that plays an indispensable role in an energy network. However, the multivariable and fractional-order characteristics of PEMFC make it difficult to establish a practical model. Herein, a fractional-order subspace identification model based on the adaptive monarch butterfly optimization algorithm with opposition-based learning (ALMBO) algorithm is proposed for PEMFC. Introducing the fractional-order theory into the subspace identification method by adopting a Poisson filter for with input and output data, a weight matrix is proposed to improve the identification accuracy. Additionally, the ALMBO algorithm is employed to optimize the parameters of the Poisson filter and fractional order, which introduces an opposition-based learning strategy into the migration operator and incorporates adaptive weights to improve the optimization accuracy and prevent falling into a locally optimal solution. Finally, the PEMFC fractional-order subspace identification model is established, which can accurately describe the dynamic process of PEMFC.