One parameterized model of indirect measurement based on neural network and its sensitivity coefficient computing

In indirect measurement, the prerequisite of evaluating the measurand and its measurement uncertainty is to necessarily know the parametric model of the indirect measuring process. A new method is developed in this paper to solve the problem of establishing the parameterized model of indirect measurement based on Radial Basis Function neural network. The analytic expression of an indirect measurement model and the computing formula of its sensitivity coefficient are derived. The simulation results prove that the method of parameterized measurement model building and the method of sensitivity coefficient computing enjoy high precision, thus effectively guarantees the estimation precision of measurement result and its uncertainty in indirect measurement, which is a useful supplement to Guide to the Expression of Uncertainty in Measurement.