A Novel Iterated DDF Algorithm Based on  Maximum Likelihood Probability

With the development of the location and tracking technology, the nonlinear filtering is becoming the research emphasis, which is the key to realize the nonlinear location and tracking. DDF is a nonlinear interpolation filtering algorithm based on stirling interpolation formula. When the system observability is weak and the observation error is large, the convergence speed, accuracy and stability of DDF are inferior. A novel iterated DDF (IDDF) is proposed by introducing the maximum likelihood probability iterated means into DDF. Since the likelihood probability is always increased in the iterated process of IDDF, its tracking accuracy and convergence speed are improved. Simulations are designed and carried out. The results indicate that, compared with EKF and DDF, higher accuracy of estimation and faster convergence are obtained using IDDF.