ADAPTIVE AND ROBUST SINGULAR VALUE DECOMPOSITION AIDED CUBATURE KALMAN FILTER WITH CHI-SQUARE TEST
Wei Zhao, Huiguang Li, Liying Zou, and Renhui Yuan
Keywords
Robustness, adaptability, Cubature Kalman filter, chi-square test, singular value decomposition
Abstract
The state variation and gross error will cause decline or divergence
of performance of Kalman filter, during the state estimation in a
stochastic dynamic system. To solve this problem, the paper gives
an adaptive and robust Cubature Kalman filter aided by singular
value decomposition. Since when state variation and gross error
occur, the covariance of innovation cannot keep orthogonal, the
adaptive and robust method uses chi-square test as the judgment
of the state variation and gross error. Then the strong-tracking-
filter-based adaptive algorithm is used for adjusting process noise
covariance matrix to eliminate the impact of state variation, and the
measurement-noise-inflating robust algorithm is used for adjusting
measurement noise to eliminate the impact of gross error. Results
of numerical simulation show that, the proposed method is adaptive
to the state variation and robust to the gross error, especially keeps
adaptability and robustness when the two cases occur at the same
time.
Important Links:
Go Back