No abstract

No abstract

In this paper, a passive hazard detection and avoidance (HDA) system is presented, relying only on images as observations. To process these images, convolutional neural networks (CNNs) are used to perform semantic segmentation and identify hazards corresponding to three different layers, namely feature detection, shadow detection, and slope estimation. The absence of active sensors such as light detection and ranging (LiDAR) makes it challenging to assess the surface geometry of a celestial body, and the training of the neural networks in this work is oriented towards coping with that drawback. The image data set for the training is generated using blender, and different body shape models (also referred to as meshes) are included, onto which stochastic feature populations and illumination conditions are imposed to produce a more diverse database. The CNNs are trained following a transfer learning approach to reduce the training effort and take advantage of previously trained networks. The results accurately predict the hazards in the images belonging to the data set, but struggle to yield successful predictions for the slope estimation, when images external to the data set are used, indicating that including the geometry of the target body in the training phase makes an impact on the quality of these predictions. The obtained predictions are composed to create safety maps, which are meant to be given as input to the guidance block of the spacecraft to evaluate the need for a manoeuvre to avoid hazardous areas. Additionally, preliminary hardware-in-the-loop (HIL) test results are included, in which the algorithms developed are confronted against images taken using real hardware.

Missions to asteroids have been the trend in space exploration for the last years. They provide information about the formation and evolution of the Solar System, contribute to direct planetary defense tasks, and could be potentially exploited for resource mining. Be their purpose as it may, the factor that all these mission types have in common is the challenging dynamical environment they have to deal with. The gravitational environment of a certain asteroid is most of the times not accurately known until very late mission phases when the spacecraft has already orbited the body for some time.Shape models help to estimate the gravitational potential with a density distribution assumption (usually constant value) and some optical measurements of the body. These measurements, unlike the ones needed for harmonic coefficient estimation, can be taken from well before arriving at the asteroid’s sphere of influence, which allows to obtain a better approximation of the gravitational dynamics much sooner. The disadvantage they pose is that obtaining acceleration values from these models implies a heavy computational burden on the on-board processing unit, which is very often too time-consuming for the mission profile.In this paper, the technique developed on [1] is used to create a validated Python-based tool that obtains spherical harmonic coefficients from the shape model of an asteroid or comet, given a certain density for the body. This validated software suite, called AstroHarm, is used to analyse the accuracy of the models obtained and the improvements in computational efficiency in a simulated spacecraft orbiting a small body.The results obtained are shown offering a qualitative comparison between different order spherical harmonic models and the original shape model. Finally, the creation of a catalogue for harmonics is proposed together with some thoughts on complex modelling using this tool.