Due to the inherent degradation of materials and the influence of various operating conditions such as load, temperature, and fuel supply, the health state of the solid oxide fuel cell (SOFC) will be inevitably degraded, resulting in system fault or even failure. The state of health (SOH) estimation and remaining useful life (RUL) predicition are beneficial to the maintenance of the SOFC systems, such as preventing unplanned shutdown, which is of great significance for ensuring the safety, reliability, and economy. Therefore, based on the framework of the data-driven and degradation model, this paper develops a method for SOH estimation and RUL prediction for SOFC. Finally, the framework was validated using longterm experimental data from a 1-cell stack.
Keywords SOFC, data-driven, degraded model, SOH estimation, RUL prediction