A Missing Data Approach to Data-Driven Filtering and Control
- 13 July 2016
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Automatic Control
- Vol. 62 (4) , 1972-1978
Abstract
In filtering, control, and other mathematical engineering areas, it is common to use a model-based approach, which splits the problem into two steps: 1) model identification and 2) model-based design. Despite its success, the model-based approach has the shortcoming that the design objective is not taken into account at the identification step, i.e., the model is not optimized for its intended use. This technical note proposes an approach for data-driven filtering and control that combines the identification and the model-based design into one joint problem. The signal of interest is modeled as a missing part of a trajectory of the data generating system. Subsequently, the missing data estimation problem is reformulated as a mosaic-Hankel structured matrix low-rank approximation/completion problem. A local optimization method, based on the variable projections principle, is then used for its numerical solution. The missing data estimation approach and the solution method proposed are illustrated on filtering and smoothing examples.Keywords
This publication has 25 references indexed in Scilit:
- Software for weighted structured low-rank approximationJournal of Computational and Applied Mathematics, 2014
- Data-driven simulation and controlInternational Journal of Control, 2008
- Linear dynamic filtering with noisy input and outputAutomatica, 2005
- Prefiltering in Iterative Feedback Tuning: Optimization of the Prefilter for AccuracyIEEE Transactions on Automatic Control, 2004
- Separable nonlinear least squares: the variable projection method and its applicationsInverse Problems, 2003
- Markov Data-Based LQG Control1Journal of Dynamic Systems, Measurement, and Control, 1998
- Iterative feedback tuning: theory and applicationsIEEE Control Systems, 1998
- The unfalsified control concept and learningIEEE Transactions on Automatic Control, 1997
- Data-based synthesis of a multivariable linear-quadratic regulatorAutomatica, 1996
- Total least squares for affinely structured matrices and the noisy realization problemIEEE Transactions on Signal Processing, 1994