AN IMM-JVC ALGORITHM FOR MULTI-TARGET TRACKING WITH ASYNCHRONOUS SENSORS

摘要

The tracking of closely maneuvering targets represents a challenge for both the contact-to-track association and the positional estimation algorithms. This paper demonstrates on a simulated scenario the performance played by each of them in a typical single-scan positional estimation problematic. Two association algorithms were used, one based on Nearest Neighbor (NN) rules the other using the Jonker-Volgenant-Castanon (JVC) optimization. These algorithms were combined to two positional estimators as well, one based on a simple Adaptive Kalman estimator and the other using an Interacting Multiple Model (IMM) estimator. The chosen combination have been modified to correct the state and covariance positional estimation for out-of-sequence reports which are likely when using asynchronous sensors.