Estimating health parameters or symptoms of a degrading system

摘要: The present invention is concerned with the estimation of health parameters p(k) representing symptoms a slowly degrading system, in particular industrial gas turbines. According to invention, an true or independent parameter vector at time step k uses previous k−1 as starting value for production predicted k. Based on latter and set measured values input variables extended model prediction output produced. This compared yield error. From this error, estimator turn produces estimate revision parameters. Preferably, includes generalised process noise stochastic uncertainty and/or a-priori knowledge evolution