Attitude planning with pointing constraints:

https://pdf.sciencedirectassets.com/271447/1-s2.0-S0094576517X00064/1-s2.0-

tl;dr:

We want the angle between the sun vector and the vector corresponding to our star tracker to be at least theta during pointing maneuvers. This inequality can be expressed in terms of a quadratic form of the quaternion. The dynamical equations, initial&final attitudes can also be expressed in terms of quaternions. This gives us yet another constrained optimization problem (eq 14) to solve (though this one is much simpler, we should be able to implement the pseudospectral ****method in python fairly easily)

Optimal task scheduling and path planning https://www.sciencedirect.com/science/article/pii/S1270963819300537

tl;dr:

Suppose we have n imaging sites, what’s the best order to observe them? For each possible order (execution strategy), what is the minimum transition time between each imagine site? What is the attitude path that’ll cost the least amount of energy?

The answer boils down to solving a constrained optimization problem given by equations 12-18. They found the minimum time and minimum-energy path for each execution strategy first, then find the best execution strategy using a genetic algorithm.

overview paper, very long: https://stacks.stanford.edu/file/druid:fp397ds6833/eddy_thesis_final-augmented.pdf