Modeling solid oxide fuel cells using continuous-time recurrent fuzzy systems

Continuous-time recurrent fuzzy systems allow modeling continuous-time processes whose dynamic behavior can be stated in linguistic if–then rules. If the dynamics are known at certain mesh points, the interpolating character of the fuzzy system between the mesh points yields a complete dynamical model of the process. In this contribution, continuous-time recurrent fuzzy systems are employed to model the electrical behavior of a solid oxide fuel cell, which both can be described qualitatively and is known quantitatively from measurements. Due to the transparency of the model an easy estimation of its parameters is possible. Additionally, the model parameters are optimized numerically to enhance the model performance.