Improvement of algorithm on the track recognition

摘要

For more exact navigation and positioning on the navigation systems, this paper discusses the improvement of statistical algorithms based on fuzzy knowledge with a brand-new theory. The improved fusion algorithms based on Bayesian theory and fuzzy methods on the track recognition are also given, respectively. However, this paper mainly makes a comparison of the improved new algorithms and two old algorithms for track recognition. The computer simulation shows the validity and feasibility of two improved algorithms. At the same time, the simulation results show that the overall performance of two improved algorithms for the track recognition all is better than that of the probabilistic algorithms. Based on the experiments in the dense target environment, the correct average recognition rate of statistical algorithms is 66.96% at the highest; that of the improved new algorithm based on fuzzy knowledge is 77.98%, that of the Bayesian fusion algorithm is 86.75%, however, that of the fuzzy fusion algorithm can be 90.14%. These results support the usefulness of the two improved algorithms. © 2002 Elsevier Science. All rights reserved.