Modeling technology for (T,p)-/spl rho/ table in mass flow-meter

A method based on the training technology of a fuzzy inference adaptive artificial neural network and nonlinear least-square (linear in structure) system identification technology for modeling the (T,P)-/spl rho/ table for a mass flow-meter is introduced. The model has several advantages such as saving calculation workload and storage space, having essential filterability. Thus the method is an effective help for the current development of high-degree integration technology of measuring and instrumentation.