An implementation of multiple model Gaussian mixture probability hypothesis density(GM-PHD) filter is
proposed. Based on the GM-PHD, the multiple model GM-PHD filter predicts and updates the state of the Gaussian
component by using the multiple model method, and the updated state is used to describe the PHD distribution of the targets.
It has the characters of both the PHD filter and the multiple model method, and can deal with multi-targets’ maneuvering
with unknown number of the targets. Compared with single model GM-PHD, the algorithm gives more accurate estimation
on the number and state of the targets. Compared with existing multiple model GM-PHD algorithm, the proposed methed
saves computation time more than 30%.