Markov Switching System Of A Class Of Delay Estimation Algorithm,

Target maneuvering, camouflage, deceiving and jamming become more and more complicated while the performance of target tracking and recognition is demanded to be real-time, accurate, reliable and integrated. Hence joint target tracking and recognition has been paid much attention. Through considering tracking and recognition integratedly, it may make full use of information available to improve the system performance. So it has great importance and good outlook. Though this kind of research includes many aspects, here we focus on the joint estimation and recognition of Markov switching systems with state measurement and delayed mode measurement. Such delay may derive from complicated recognition process, or even data transmission. We have done some work to improve tracking performance, and the main contributions are as follows:1. We consider a system in which mode measurement is one-step delayed. Through state-augmentation approach, this kind of estimating problem for Markov switching systems with time-delays is changed into a generalized filtering problem for common Markov switching systems. This algorithm (mode-measurement-enhanced interacting multiple model generalized filtering algorithm, namely mme-IMM-gfa) not only is deduced using Bayes theory, but also is illustrated via a maneuvering target tracking simulation example.2. In maneuvering target tracking environment, some typical algorithms based on IMM are introduced and compared. Their applied conditions, keystones, similarities, differences, performances and complexities are analyzed.3. Besides, we do some study of the current state estimating problem where mode measurement is multi-step delayed. We also present an algorithm and give its application condition.