Multiobjective Optimization and Data-Driven Constraint Adaptive Predictive Control for Efficient and Stable Operation of PEMFC System

Efficient and stable operation (ESO) of the proton exchange membrane fuel cell (PEMFC) system is challenging because the nonlinearity of system not only aggravates the difficulty of performance analysis but also puts forward higher requirements for controller design. In this article, a hierarchical performance enhancement control strategy (HPECS) is proposed, which can deal with optimal trajectory seeking under multiobjective constraints and nonlinear problem of system separately. In upper optimization level, the multiobjective optimization based on dynamic adjustment of weighting factor is proposed to seek optimal trajectory for ESO of PEMFC system considering the harm of oxygen starvation, oxygen saturation to system, and contradiction characteristic between net power and efficiency; in lower control level, a data-driven constraint adaptive predictive control is proposed for optimal trajectory tracking with the nonlinearity of system. The proposed HPECS is implemented and validated on the experimental PEMFC system. The comparative studies show that the proposed HPECS is superior in performance enhancement, disturbance rejection, and tracking ability. The normalized multiobjective optimization function can be basically guaranteed to above 0.9, and a nearly 7.2% enhancement in average can be also observed, which shows that ESO of PEMFC system can be achieved in changeable working conditions under proposed HPECS.

Automated summary

This article proposes a hierarchical performance enhancement control strategy (HPECS) to effectively address the challenges of efficient and stable operation of proton exchange membrane fuel cell (PEMFC) system.