Improving energy aware nanosatellite task scheduling by a branch-cut-and-price algorithm

Abstract

Listen

Translate article

CubeSats are small satellite platforms that have become increasingly popular for space research due to their small size, modular design, and affordability. However, their limited power and communication capabilities present significant challenges for the development and operation of these satellites. One of the main challenges is to effectively manage the limited power resources available on CubeSats for supporting the various tasks and payloads of the satellite. Addressing these challenges requires careful task scheduling that considers the power requirements and importance of different tasks and the trade-offs between power consumption and other factors such as data rate, volume, and accuracy. Decision-making in task scheduling for CubeSats often involves solving large mixed-integer programming (MIP) problems, which can be computationally intensive and time-consuming. To address this challenge, we propose a branch-cut-and-price algorithm for CubeSats that improves upon previous approaches in the literature. We apply dual stabilization to column generation and propose new valid inequalities; resort to an effectively implemented dynamic programming (DP) algorithm to generate multiple columns in parallel; and rely on a new branching strategy based on pseudocosts to reduce the number of nodes and improve the overall performance of the method. Our results demonstrate the effectiveness of these enhancements for solving a complex task scheduling problem in the context of CubeSats compared to the state-of-the-art approach, and further highlight its potential to improve the efficiency and accuracy of mission planning and operations for these small satellite platforms.