基于改进多模型的火星大气进入自适应估计方法

针对火星进入过程中大气密度等不确定参数对导航系统状态估计精度的影响,提出了一种基于改进混合专家框架的多模型自适应估计方法。该方法对进入过程中不同的测量信息进行规范化处理,以克服传统多模型自适应估计方法稳定性差、数值下溢等固有缺陷,进一步提高状态估计精度。将其应用于火星不同进入探测方式下的导航场景进行仿真分析。仿真结果表明:该方法在动力学系统模型参数存在不确定扰动时能获得精确的状态估计,可以满足未来定点着陆探测对导航系统的精度需求。

Mars Entry Adaptive Estimation Method Based on Modified Multiple Models

In this paper, aiming at the adverse effects of uncertain parameters, such as Mars atmospheric density uncertainty, on state estimation, a modified multiple model adaptive estimation (MMAE) method using mixture-of-experts framework is proposed. Normalization dimensional processing of different measurement information during entry phase is conducted by the proposed method to overcome the inherent defects such as poor stability and numerical underflow of traditional MMAE method, further improving the state estimation accuracy. The method is also applied to two different navigation scenarios under different Mars entry detection modes. The simulation results show that the proposed method can obtain precise state estimation when there is uncertain disturbance in dynamics system.