Stochastic Multi-Pass Noncoplanar Aeroassisted Orbital Transfer Problem using Log-Domain Sequential Covariance Control

Noncoplanar aeroassisted orbital transfer is a trans-medium, nonlinear, multi-phase optimal control problem. This technology relies on large bank angle maneuvers to generate the requisite orbital normal force; however, this characteristic inevitably compromises the vehicle's control margin available for disturbance suppression. Consequently, efficiently and accurately evaluating critical heat rate uncertainties to guarantee flight safety amid environmental stochasticity remains a formidable challenge. To address this, this paper proposes a log-domain sequential covariance control method, achieving efficient and exact quantification of heat rate uncertainties. Specifically, a log-domain mapping is introduced to losslessly transform the multiplicative atmospheric density noise, which follows a log-normal distribution, into an additive Gaussian structure. This exact transformation enables high-precision analysis utilizing linear covariance propagation theory within a sequential convex programming framework. To further enhance algorithmic efficiency, key techniques including stochastic state space reduction, trans-medium linkage constraint reconstruction, and convex feasible subset approximation are developed. Extensive high-fidelity Monte Carlo simulations demonstrate that the proposed method significantly mitigates the risk of underestimating extreme heat rate boundaries. Furthermore, it exhibits superior fuel economy compared to the theoretically optimal traditional Hohmann transfer. The proposed method provides a robust and high-efficiency trajectory planning solution for onboard autonomous aeroassisted orbital transfer missions.