Elsevier

Acta Astronautica

Volume 123, June–July 2016, Pages 51-61
Acta Astronautica

Campaign-level dynamic network modelling for spaceflight logistics for the flexible path concept

https://doi.org/10.1016/j.actaastro.2016.03.006Get rights and content

Abstract

This paper develops a network optimization formulation for dynamic campaign-level space mission planning. Although many past space missions have been designed mainly from a mission-level perspective, a campaign-level perspective will be important for future space exploration. In order to find the optimal campaign-level space transportation architecture, a mixed-integer linear programming (MILP) formulation with a generalized multi-commodity flow and a time-expanded network is developed. Particularly, a new heuristics-based method, a partially static time-expanded network, is developed to provide a solution quickly. The developed method is applied to a case study containing human exploration of a near-Earth object (NEO) and Mars, related to the concept of the Flexible Path. The numerical results show that using the specific combinations of propulsion technologies, in-situ resource utilization (ISRU), and other space infrastructure elements can reduce the initial mass in low-Earth orbit (IMLEO) significantly. In addition, the case study results also show that we can achieve large IMLEO reduction by designing NEO and Mars missions together as a campaign compared with designing them separately owing to their common space infrastructure pre-deployment. This research will be an important step toward efficient and flexible campaign-level space mission planning.

Introduction

As space becomes more and more accessible through technology development, space systems design has also become increasingly complex. Thus, we need a campaign-level perspective for space systems design in addition to the conventional mission-level perspective. A campaign contains multiple missions that may or may not be for the same destinations. A design method taking the whole campaign into consideration can enable efficient and flexible space systems.

We can see a transition from mission-level design into campaign-level design in our history of space exploration. In the Apollo project, all missions used a carry-along strategy, where they transported everything they needed by themselves. This was possible due to their short mission durations (e.g., 2 weeks) and small demands in consumables and equipment. However, that type of mission-level strategy is not necessarily desirable for long-term missions. For example, the International Space Station (ISS) programme could not use a mission-level strategy because it aimed to have a human presence in space over the long term. Instead, it used a campaign-level resupply strategy, where the vehicles are launched regularly to resupply the hardware and consumables and even replace the crew.

The campaign-level strategies include uses of technologies and infrastructure considering the whole campaign. Particularly, a combination of pre-deployment, carry-along, and resupply strategies will be very important for interplanetary missions over the long term. In addition, effective uses of different technologies and infrastructures such as advanced propulsion system [1], [2], propellant depot [3], [4], [5], [6], and in-situ resource utilization (ISRU) [7], [8], [9], [10], [11] have also been proposed to be useful.

Given that background, we aim to find an efficient optimization framework to consider a campaign-level mission planning for future human and robotic space exploration. This paper particularly intends to improve and extend the most advanced work we recognize in space logistics optimization literature, which uses time-expanded generalized multi-commodity network flow (GMCNF) approach for space logistics modelling [12], [13], [14], [15]. In the previous work, dynamic linear programming (LP) based network optimization approach has been proposed to be effective [12], [13], [14], [15]. However, using continuous variables for all flows might not provide a realistic solution. For example, a crew member can be split into multiple pieces in the optimal solution. In order to avoid that situation, this paper proposes a mixed-integer linear programming (MILP) formulation. A more important contribution of this paper is about the time-expanded network. The previous work used a bi-scale time-expanded network for dynamic optimization [14], [15]. This method was useful in providing a logistics flow over time and provided a more realistic solution than static optimization, but it required a large computational resources. In this paper, a new efficient approximate method, a partially static time-expanded network, is proposed. This method combines the ideas from both static and bi-scale time-expanded network approaches, which enables significant improvement in computational effort.

The resulting efficient dynamic optimization formulation is proposed and applied to a case study containing human exploration of a near-Earth object (NEO) and Mars, related to the concept of the Flexible Path. This case study shows that a design from a campaign-level perspective would significantly improve the performance. This is the very reason why we need to pursue campaign-level space mission planning, and the proposed method can be a very strong support and evaluation tool for that purpose.

The rest of the paper is organized in the following: Section 2 shows the past research in space logistics modelling, and Section 3 shows the detailed method. Section 4 applies the proposed method to the case study. Section 5 concludes the discussions.

Section snippets

Literature review

Space logistics has been studied recently in the context of the efficient human exploration of Mars. A detailed literature review and motivation behind space logistics research can be found in the past papers [14], [15], but a few important studies are shown here.

Network modelling has been proposed to be an effective technique for efficient optimization of space logistics. It converts the space logistics map into a mathematical graph as shown in Fig. 1. Here, the nodes correspond to physical

Campaign-level dynamic space logistics modelling

This chapter summarizes the mathematics for campaign-level dynamic space logistics modelling. There are two important components in the mathematics of space logistics modelling: MILP formulation and the partially static time-expanded network. The details of these two components are shown in the next sections.

Case study: Flexible Path

This chapter applies the proposed method to a case study. The objective of the case study is (1) to show the usefulness of the newly proposed MILP-based partially static time-expanded network and (2) to evaluate the benefits of campaign-level system design compared with the mission-level system design.

Conclusions

In this paper, a campaign-level dynamic logistics network formulation for lifecycle optimization of human-robotic mission sequences is proposed and applied to the Flexible Path case study. The proposed method using a partially static time-expanded network can find the optimal combination of technologies and operations to be used at each stage of the campaign. The idea is to combine the GMCNF with a time-expanded network to model the dynamic movement of commodities and infrastructure elements

References (33)

  • G. Sanders, W. Larson, R. Sacjsteder, A. Mclemore, NASA in-situ resource utilization (ISRU) project - development and...
  • G. Sanders, K. Romig, W. Larson, R. Johnson, D. Rapp, K. Johnson, K. Sacksteder, D. Linne, P. Curreri, M. Duke, B....
  • S. Schreiner, Molten Regolith Electrolysis Reactor Modeling and Optimization of In-Situ Resource Utilization Systems...
  • T. Ishimatsu

    Generalized Multi-Commodity Network Flows: Case Studies in Space Logistics and Complex Infrastructure Systems (Ph.D. thesis)

    (2013)
  • K. Ho, O. de Weck, J. Hoffman, R. Shishko, Dynamic network modelling for spaceflight logistics, in: Proceedings of...
  • K. Ho

    Dynamic Network Modeling for Spaceflight Logistics with Time-Expanded Networks (Ph.D. thesis)

    (2015)
  • Cited by (47)

    • Integrated in-situ resource utilization system design and logistics for Mars exploration

      2020, Acta Astronautica
      Citation Excerpt :

      They can produce and store space resources in space, especially spacecraft propellant, to support space transportation and reduce mission costs. Studies on campaign-level space mission planning have shown the effectiveness of ISRU and propellant depots for space transportation [3–7]. On the other hand, several testbeds have been built by NASA [11] and Lockheed Martin [12] to evaluate the performance of the hydrogen reduction reaction plant in oxygen production.

    • A location-routing problem for the design of an asteroid mining supply chain network

      2019, Acta Astronautica
      Citation Excerpt :

      A single-facility location problem has also been developed for determining the optimal radius of an orbital factory in heliocentric orbit [25]. The space transportation networks developed in Refs. [13–17] contain only a limited number of orbital nodes, defining generic orbits such as Low Earth Orbit, Geostationary Transfer Orbit, and Geostationary Orbit (as well as comparable orbits around the Moon and Mars), each with a defined orbital radius and shape. While these nodes allow for the determination of optimal routing strategies of spacecraft through the network, there is little ability in the optimization of the shape and orientation of the individual orbits.

    View all citing articles on Scopus
    View full text