Inferential process modeling, quality prediction and error detection using multi-level data segregation

摘要

A process modeling technique uses a single statistical model, such as a PLS, PRC, MLR, etc. model developed from historical data for a typical process, and uses that model to perform quality prediction or error detection for various different process stages of a process. The modeling technique determines averages (and possibly standard deviations) of the process parameters for each of a set of product classes, throughputs, etc., compares process parameter measurements online with those agents, and uses those comparisons in a single process model to perform quality prediction or error detection across the various stages of the process , Because only the mean and standard deviations of the process model process parameters are updated, a single process model can be used to perform quality prediction or error detection while the process is operating in any of the defined process stages or stages. The sensitivity (robustness) of the process model can also be set manually or automatically for each process parameter to tune or adapt the model over time.