.png){kind=icon}
.png){kind=icon}
Return
Multiple model estimation: A convex model formulation
.png){kind=icon}
PRE
AI
DOI：10.1002/acs.1034.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The multiple model (MM) framework provides an elegant solution to adaptive filtering problems. An important issue in the MM framework is how the estimation is performed. In this paper, a brief overview is given of the mainstream methods for MM estimation and a new method is proposed. Contrary to existing methods that mostly adopt a hybrid model structure, the newly proposed method uses a more general MM framework that allows for weighted combinations of the local models. The main advantage of this framework is that it has better model interpolation properties. These improved properties allow for smaller model sets, which are very useful in, for example, fault detection and identification (FDI) of partial faults. The improved interpolation properties are demonstrated by means of two simulation examples, one in which an FDI problem is addressed, and one in which a target tracking problem is addressed. Monte Carlo simulation results of these two examples are given. In these simulations, the well-known interacting IMM filter is compared with two estimation algorithms based on the proposed model structure. Copyright (C) 2008 John Wiley & Sons, Ltd.
Keywords:
multiple model systems
model interpolation
joint parameter and state estimation
jump Markov linear systems
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
3.8
Papers:
2.6K
Citations:
3.6K