## Abstract

A modified truncation Gauss probability density model was set up based on the truncation Gauss probability density model. The algorithm of overall adaptive filtering (OAF) for tracking maneuvering target was proposed by combining the modified truncation Gauss probability density model with adaptive velocity estimation model. In the OAF algorithm, the preestablishing of the maximum acceleration is avoided and the maneuvering frequency is adjusted adaptively. The system noise covariance and the maneuvering frequency were fused by BP neural network. The fused system covariance was gotten by regulating the veight value of neural network online, to reduce precision loss due to regulating parameter improperly. The results of theoretical analyse and simulations show that the OAF algorithm can improve the precision of system by 49.61% and 48.34% respectively in tracking maneuvering and nonmaneuvering targets, in comparison with only adaptive velocity estimation algorithm.

Original language | English |
---|---|

Pages (from-to) | 63-68 |

Number of pages | 6 |

Journal | Binggong Xuebao/Acta Armamentarii |

Volume | 30 |

Issue number | 1 |

Publication status | Published - Jan 2009 |

## Keywords

- Adaptiveness
- Automatic control technology
- Kalman filter
- Target tracking
- Truncation gauss probability model