State Estimation of a Hybrid Markov Process with Application to Multitarget Tracking.
Abstract
Many estimation problems, including failure detection and multitarget tracking, rely on the state estimation of a hybrid Markov process. An algorithm for estimating the state sequence of such a process must keep the amount of computation at a reasonable level. Techniques are derived for reducing the computational burden, that do not increase the error probability. In a multiple system case, the concept of clustering, which greatly diminished the computational complexity, can be applied without loss of optimality under precise conditions. For the application to multitarget tracking, the computation of an optimal grating is done. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 1982
- Accession Number
- ADA111728
Entities
People
- Franck E. Bruneau
Organizations
- Massachusetts Institute of Technology