The time delay estimation based on cubature particle filter

At the case of non-stationary, non-Gaussian noise and time-varying, the characteristics of the time delay estimation was improved based on the particle filter. In the process of the time delay estimation based on the particle filter, importance density function is the key to the performance of time delay estimation. A new method of the time delay estimation based on cubature particle filter (BCPF-TDE) is presented in this paper. The BCPF-TDE method used the latest measurements to generate the importance density function with Cubature Kalman Filter (CKF). This importance density function approximated to the posterior probability distribution of the time delay parameter by use of the BCPF-TDE. The simulation results show that the estimation error of BCPF-TDE is a lower than those of the time delay estimation based on unscented particle filter (BUPF-TDE) when particle number is the same. Compared with the method BUPF-TDE and BCPF-TED, run time of BCPF-TDE is drastically reduced at the case of the estimation accuracy is similar. These mean that the new method BCPF-TDE is effective and reliability.