A Parallel Estimation Algorithm
A parallel estimation algorithm is proposed for large scale dynamic systems. The algorithm is bi-modal in that the local estimates are computed in parallel at high sampling rates considering the subsystems as decoupled from each other. At the same time, the outputs of the local estimators are used as initial data for the overall estimator which takes the interaction effects into account in a block-iterative fashion. Each individual locally optimal estimator can have a different sampling rate to match the subsystem dynamics.
KeywordsGain Matrix Optimal Estimator High Sampling Rate Individual Subsystem Subsystem Dynamic
Unable to display preview. Download preview PDF.
- M. Hodzic and H. AlKhatib (1983). Distributed architecture for implementation of an algorithm for estimation of sparse large scale dynamic systems, ISMM Conference MIMI’83, San Antonio, Texas.Google Scholar