Chance constraints transcription and failure risk estimation for stochastic trajectory optimization

Abstract

Stochastic trajectory optimization ensures mission success under uncertainty by enforcing chance constraints. Transcription methods convert these into tractable deterministic forms, but existing approaches do not support general multidimensional constraints. We introduce three such methods: spectral radius, first-order, and d-th order with low conservatism. A risk estimation technique and conservatism metric enable systematic evaluation. Applied to aerospace scenarios, our methods outperform existing ones, with the d-th order transcription excelling in high-dimensional settings.