Robust particle filter—anti‐zero bias modification. (15th March 2016)

• Record Type: Journal Article
• Authors: Kozierski, P.; Lis, M.; Horla, D.
• Is Part Of: International journal of robust and nonlinear control. Volume 26:Number 16(2016)
• Page Start: 3645
• Page End: 3661
• Publication Date: 2016-03-15
• Subjects: robust state estimation -- large measurement errors -- power system -- bad data; Automatic control -- Periodicals; Control theory -- Periodicals; Nonlinear systems -- Periodicals; 629.836
• DOI: 10.1002/rnc.3526
• Languages: English
• ISSNs: 1049-8923

Abstract

Summary: A modification of the particle filter algorithm that allows using it also in cases with incorrect measurements has been presented in the paper. The use of anti‐zero bias does not require a large computational effort (a single additional operation for each measurement value), and simultaneously does not deteriorate results for the case of good measurements (if the bias value is not too large). It has been shown that the bias which provides the best estimation quality depends on the particles number. The obtained results have been compared with unscented Kalman filter method with bad measurement data identification. As an object, power system has been used, with main task set as estimation of the state of this system. Copyright © 2016 John Wiley & Sons, Ltd.