INTERACTING MULTIPLE MODEL METHODS IN TARGET TRACKING - A SURVEY

The Interacting Multiple Model (IMM) estimator is a suboptimal hybrid biter that has been shown to be one of the most cost-effective hybrid state estimation schemes. The main feature of this algorithm is its ab ility to estimate the state of a dynamic system with several behavior modes which can ''switch'' from one to another. In particular, the IMM estimator can be a self-adjusting variable-bandwidth filter, which ma kes it natural for tracking maneuvering targets. The importance of thi s approach is that it is the best compromise available currently betwe en complexity and performance: its computational requirements are near ly linear in the size of the problem (number of models) while its perf ormance is almost the same as that of an algorithm with quadratic comp lexity. The objective of this work is to survey and put in perspective the existing IMM methods for target tracking problems. Special attent ion is given to the assumptions underlying each algorithm and its appl icability to various situations.