원문정보
초록
영어
This study is devoted to overcome the underflow problem and poorly cost-effective limitation of model-set adaptive IMM algorithm. Cause of underflow problem in Novel-IMM is addressed firstly, based on which an underflow prevented selection probabilities (UPSP) algorithm is presented to solve this problem. This paper then presents a fast model-set adaptive (FAIMM) IMM algorithm based on steady state Kalman filters that decrease the computational burden greatly while keeping acceptable tracking accuracy. Finally, the threshold choosing strategy of UPSP algorithm is presented, which could make the FAIMM algorithm achieves ideal performance. Simulation results demonstrate that the FAIMM algorithm can be an effective estimator in real-time application.
목차
1. Introduction
2. Novel-IMM and Underflow Problem
2.1. Standard Novel-IMM Algorithm
2.2. Underflow Problem In Novel-IMM
2.3. Solutions For Underflow Problem
3. Cost-Effective Implementation of Model-Set Adaptive IMM
3.1. Steady State Kalman Filters
3.2. The Fast Model-Set Adaptive IMM (FAIMM) Algorithm
4. Results and Analyses
4.1. Simulation Settings
4.2. The Effectiveness of UPSP Algorithm
4.3. Performances of FAIMM Algorithm
4.4. Threshold Choosing Principle of UPSP Algorithm
5. Conclusion
References