Regularized Differentiation of Measurement Data Using A Priori Information on Signal
Andrzej Miękina, Roman Z. Morawski · 1990
This paper is concerned with an authors' original algo- rithm for real-time differentiation of discrete measurement data. The effectiveness of this algorithm depends on a regularization parameter whose value should be fitted to the level of disturbances that the data are subject to. A simple method for choosing this value has been pro- posed which requires only scanty apriori information on the data, viz., an estimate of the signal bandwidth and an estimate of the signal-to- noise ratio. The effectiveness of this method has been demonstrated using synthetic data and computer experimentation methodology. It is shown that the attainable accuracy of differentiation is very close to the optimum which may be reached via empirical optimization of the parameter of regularization.