Adaptive Kalman Filter Based on Evolutionary Algorithm and Fuzzy Interference System

Abstract

This is a survey paper.The performance of the Kalman filter (KF), which is Algoritstandard as an outstanding implementation for dynamic system state estimation, greatly depends on its parameter R, called the measurement noise covariance matrix. . However, it’s dif?cult to obtain the accurate value of R before the ?lter starts, and the value of R is possible to change with the measurement environment once the ?lter is working. To solve this difficulty, a new parameter adaptive Kalman ?lter is proposed in this paper. In this new Kalman ?lter, the initial value of R is of?ine determined by Evolutionary hm (EA), and the value of R determined by EA is online updated by Fuzzy Inference System (FIS). The new adaptive Kalman ?lter proposed in this paper (HYdGeFuzKF) has a stronger adaptableness to time-varying measurement noises than regular Kalman ?lter (RegularKF).