Efficient on-line parameter estimation in TRANSCEND for nonlinear systems

Prognosis and Health Management methodologies require efficient parameter estimation approaches to enable systematic system reconfiguration and adaptive control to accommodate faulty behaviors, and to predict future system states. However, accurate and timely on-line parameter estimation of complex, nonlinear systems is difficult and can be computationally expensive. In this work, we propose a more efficient technique for on-line parameter estimation in TRANSCEND. This new approach is based on previous works on model decomposition and dependency compilation. We generate a set of smaller estimation tasks from the global estimation problem to reduce the computational burden. We tested the approach in a nonlinear three-tank system. Current results demonstrate that our method is more efficient and it does not compromise on the accuracy in the estimation.