On hierarchical estimation and system identification using the MAP criterion

This paper presents a general treatment of hierarchically structured maximum-a posteriori (MAP) estimation and system parameter identification algorithms for large-scale systems. Continuous-time and discrete-time algorithms for hierarchical estimation and system identification in large-scale systems are derived for several supremal infimal coordination methods. Consolidation and unification of previous results in hierarchical systems theory as applied to system identification are accomplished. Numerical examples are considered to demonstrate the effectiveness of the algorithms which are potentially of considerable utility in computer modeling of large systems.