A Hybrid Computer Optimal Filter,
Luis Basañez, Brunet,P., Bucy,R. S., Rafael M. Huber, D. S. Miller · Defense Technical Information Center (DTIC) · 1975
In 1971 a hybrid computer algorithm for implementation of an optimal nonlinear one-step predictor by applying Bayes' Rule to sequentially update the conditional probability density function from the latest data was presented in a paper. Such a filter has been constructed at the Laboratorio de Automatica at the Universidad Politecnica de Barcelona. This paper contains a description of the hybrid optimal filter as it was actually implemented for the cubed sensor problem. A comparison of the hardware and software architectural features with those proposed in the 1971 paper presented and reasons for differences given. Accuracy of estimates and run times are compared against those predicted in and those of all digital filters performing the same estimation.