Efficient adaptive tracking algorithm

In this paper, the authors propose a new algorithm - IMM Adaptive Innovation Filter algorithm. The contributions are: 1. Model error of the filter is reduced. This is accomplished through the adoption of 'current' statistic model as the sample of maneuvering target model of IMM. As it is well known, 'current' statistic model and the adaptive algorithm possess rather good performance for tracking maneuvering target. 2. Sensitivity of probability weight to measuring noises is reduced. This is accomplished by employing innovation filter to filter out the residual noises. Theoretical analysis and simulation show that the algorithm improves effectively the performance of Blom's method, reduces the time delay, and improves the performance of tracking.