Adaptive Extended Kalman Filter using Exponencial Moving Average⁎. Issue 25 (2018)

• Record Type: Journal Article
• Authors: Silva, Jean Gonzalez; Aquino Limaverde Filho, José Oniram De; Feitosa Fortaleza, Eugênio Liborio
• Is Part Of: IFAC-PapersOnLine. Volume 51:Issue 25(2018)
• Page Start: 208
• Page End: 211
• Publication Date: 2018
• Subjects: Estimators -- Adaptive filters -- Covariance matrices -- Nonlinear systems -- Extended Kalman filters; Automatic control -- Periodicals; 629.805
• DOI: 10.1016/j.ifacol.2018.11.106
• Languages: English
• ISSNs: 2405-8963

Abstract

Abstract: This paper presents a novel adaptive extended Kalman filter for nonlinear discrete-time systems that deals with variations in the process noise covariance. The process noise covariance is estimated at each sample time by calculating the innovation term covariance through exponential moving average. Numerical simulations were carried out to illustrate the proposed algorithm based on a closed-loop nonlinear mass-spring-damper system.