Addressing the challenges in accurately gauging the anode’s internal state variables and the delay in the hydrogen supply system’s reaction under dynamic load in proton exchange membrane fuel cells (PEMFC), this document proposes a nonlinear PEMFC system model. The efficacy of the model is confirmed by comparing experimental data. Subsequently, the unscented Kalman filter (UKF) and sliding mode observer (SMO) models are developed to assess the anode’s internal condition as separate observers, with a comparative analysis of both models’ effectiveness in estimating the anode’s internal state. A multiobserver (MO) design approach is suggested to overcome the shortcomings of the UKF and SMO models in estimating the internal states of the anode, merging the UKF and SMO models. Enhancing the precision of anode hydrogen pressure, anode nitrogen pressure, and anode pressure is achieved through the dynamic adjustment of each observer’s weight via particle swarm optimization (PSO). Findings indicate the MO model’s distinct benefits in estimating anode states, particularly in noisy environments. Relative to UKF and SMO, the MO estimate’s root mean square error (RMSE) for anode hydrogen pressure has decreased to 54.25 Pa, showing reductions of 62.55% and 88.27%, respectively. The RMSE for the estimated anode nitrogen pressure stands at 155.14 Pa, accompanied by decreases of 44.59% and 82.88%, correspondingly. The RMSE for the MO’s anode pressure estimation stands at 212 Pa, showing 44.98% and 42.16% decreases relative to UKF and SMO, respectively, confirming the MO model’s efficacy in enhancing anode state estimation precision.
Jia et al. (Mon,) studied this question.