Information fusion algorithm based on adaptive cubature strong tracking filter

Abstract

The adaptive cubature strong tracking information filter(ACSTIF), which combines the strong tracking filter with the variational Bayesian method, has a good performance when the sudden change of state and the unknown variance of measurement noise appear. However, it also remains two problems. Firstly, the iteration of estimating the unknown variance of measurement noise is not very accurate. Secondly, it is designed for single sensor system so it's measure range and the accuracy is not very good. Aiming at these two problems, this paper proposed a sequential fusion algorithm based on improved adaptive cubature strong tracking information filter. We optimize the algorithm by modifying the process of estimating the unknown variance of measurement noise and combining the improved ACSTIF algorithm with sequential filtering fusion. Our analysis shows that the improved ACSTIF has a better performance than the original ACSTIF. Finally, a four dimensional simulation example is given to verify the validity and accuracy of the analysis results obtained in this paper.