It is difficult to obtain the ML(Maximum Likelihood) parameter estimation of GTD (Geometric Theory of Diffraction) scattering center model because we must solve a high-dimensional

nonlinearity

mixed-parameter optimization problem.A CPSO (Cooperative Particle Swarm Optimization) based algorithm is developed for obtaining the ML estimation of GTD model parameter.This method can obtain the amplitude

type and position parameter simultaneously

and it is insensitive to the initialization value.Comparing with RELAX-based method

the CPSO-algorithm need not iterative estimate repeatedly

so reduces the complexity of computation.The simulation tests show that the CPSO-based method can get quite accurate estimation result of GTD model.