Predictive neural networks control in the high accuracy DC voltage reference source

In the paper we compare the predictive abilities of several types of neural networks in order to use them for voltage correction in a high precision solid state DC voltage reference source (DCVRS). The observed time series are characterised as outputs of nonlinear dynamical system. Networks are trained until the correlation dimension and the leading Lyapunov exponent of the predicted signals reaches the values of the same invariant measures of the observed system.