Innovation approach to distributed-parameter detection and estimation†
Spyros G. Tzafestaş · International Journal of Systems Science · 1978
The innovation approach to detection and estimation is applied to the case of distributed-parameter (DP) systems and signal processes. The concept of the DP innovation process is studied for both gaussian and poissonian observation models, and formulae are derived for the likelihood ratio of some signal detection problems associated with these models. Then state-estimation problems are considered and solved for several classes of non-linear DP systems, with time delays or not, coupled with both observation models, as well as for Poisson driven DP systems. A set of examples is provided to indicate some areas where DP detection/estimation theory can be used.