基于自适应SSUKF的多弹协同导航算法

Multi-missile Cooperative Navigation Algorithm Based on Adaptive SSUKF

在现代战争中,导弹协同作战已经变得越来越普遍。针对多弹协同作战,导弹在复杂、恶劣的战场环境下,卫星导航系统功能严重受限甚至出现不定位状况,造成传统SINS/GNSS组合导航系统精度降低等诸多问题,提出了基于自适应SSUKF的多弹协同导航算法。首先,以弹群中各弹上的惯性导航系统作为基准,利用数据链的量测得到该弹与弹群中其他弹之间的相对位置关系,再结合该弹惯性导航系统的误差方程,建立了基于秩亏约束级联滤波异构弹群的协同导航解算模型。进而针对UKF计算量大的缺点,利用超球体单形采样策略代替传统的比例对称采样策略,并基于所建立模型的线性状态方程和加性噪声的特点,提出了简化的自适应SSUKF算法进行协同导航解算,在保证解算精度的同时大大提高了计算效率。最后,仿真结果表明,所提出的算法能够有效提高复杂战场环境下异构弹群组网作战导弹导航系统的精度与适应能力,其精度提高约为纯惯导的两倍,且解算效率相较于传统算法有了显著的提升。

In modern warfare,cooperative operation among missiles has become increasingly prevalent.To address the challenges faced by missiles in complex and harsh battlefield environments,where satellite navigation system functionality is severely constrained or even not positioning,leading to reduced accuracy in traditional SINS/GNSS integrated navigation system,a multi-missile cooperative navigation algorithm based on adaptive SSUKF is proposed.Firstly,treating the SINS of each missile in the swarm as a reference,the relationships of relative position between the missiles within the swarm are obtained using data link measurements.Combining the information with the error equations of the SINS for each missile,a solution model for heterogeneous swarm cooperative navigation based on rank-deficient constraint serial filtering is established.Then,in response to the disadvantage of large computational complexity in UKF,the spherical simplex is employed instead of the traditional proportional symmetric sampling strategy.Leveraging the linear state equations of the model and the characteristics of additive noise,a simplified adaptive SSUKF algorithm for cooperative navigation is proposed.This algorithm significantly improves computational efficiency while ensuring accuracy.Lastly,simulation results show that the proposed algorithm effectively enhances the accuracy and adaptability of multi-missile navigation systems within heterogeneous swarm networks operating in complex battlefield environments.The accuracy improvement is approximately twice that of pure SINS.Meanwhile,the computational efficiency is significantly enhanced compared to the traditional algorithms.