Model correction and updating of a stochastic degradation model for failure prognostics of miter gates

Understanding the degradation of quoin block is vital for failure prognostics in miter gates. Due to complicated mechanism, models based on simplifications and assumptions cannot accurately describe damage evolution. It observed that small errors a simplified model can lead large discrepancy remaining useful life estimation attributed error accumulation over time. Aiming address this issue prognostics, paper presents dynamic correction framework using strain measurements. In proposed framework, polynomial chaos expansion (PCE) employed compensate missing physics stochastic model. A maximum likelihood method developed estimate uncertain parameters physics-based along with unknown PCE measurements as observables. The updated then applied gate. Results case study show approach effectively improve accuracy