CoChat

Optimization of Reconfigurable Satellite Constellations Using Simulated Annealing and Genetic Algorithm

Open access
Paek, Sung Wook, Kim, Sangtae, de Weck, Olivier · 2019 · Sensors

The DOI matches its Crossref record; 47 of the 52 references it cites were found in a registry or an index; 2 of the 20 sources its claims cite were checked for support.

Abstract

Agile Earth observation can be achieved with responsiveness in satellite launches, sensor pointing, or orbit reconfiguration.

Overview

CoChat Summary

TL;DR

This paper presents a reconfigurable satellite constellation (ReCon) framework for agile Earth observation, co-optimizing satellite design and orbits for global and regional observation modes. It benchmarks simulated annealing (SA) and genetic algorithm (GA) against a gradient-based method, finding GA produced the most optimal solution but at higher computation cost.

Key findings

  • GA outperformed SA and gradient-based steepest descent; the GA solution was slightly (4%) more optimal than the SA solution in terms of the J value, but GA required 7 times longer computation time than SA. SA performed similar or slightly better than the gradient-based method.
  • Both SA and GA produced very similar optimal ReCon configurations with 5 orbit planes, 2 satellites per plane, and a field of regard of 47°; only the altitude difference differed in sign. The GA individual satellite weighs 2 ton and the aperture has a diameter of 1.2 m, within the Earth-observation satellite database ranges.
  • Sensitivity analysis at the GA optimum x* = (15/1, −54.7, 5, 2, 47.1) found the RGT ratio had the largest impact, followed by the number of planes; increasing the RGT ratio by 0.5 decreases the RGT altitude by approximately 100 km. Increasing mutation rate from 0.001 to 0.002 and 0.004 led to increased average J scores of 1.73 and 2.30 with a statistically significant p-value at 0.0.
  • The ReCon offers performance-to-cost (mass) superior to static constellations; a static constellation would have a J score of 2 at least. LHS reduced 4096 combinations down to 100 design points.

Methods

  • Systems engineering and multidisciplinary system design optimization formulate ReCon as nonlinear programming with five design variables: altitude (RGT ratio), altitude difference, number of orbit planes, number of satellites per plane, and field of regard. The objective J is a weighted sum of four figures of merit (ROM revisit time, GOM coverage, constellation mass, reconfiguration time) plus constraint penalties.
  • Simulation modules include astrodynamics (STK simulations run for two days, up to J4 effects), optics (GSD/aperture relationships), maneuvers (reconfiguration time), and propulsion/constellation (mass via rocket equation). Preliminary Latin hypercube sampling created 100 design points from 4096 combinations.
  • Optimization methods: SA (single-point, initial x0 = [31/2, −200, 2, 1, 50], penalty gain g = 1000), GA (population 50, crossover 0.95, mutation 0.001, constraint gain 0.1), and gradient-based steepest descent benchmark.

Limitations

  • The paper states detailed cost analysis is not the scope of this study; direct comparison with static satellite constellations is not straightforward because they cannot adopt the “choice and concentration” observation strategy of a ReCon. Future work includes political and policy aspects, low-thrust electric propulsion, radar observation, GPS radio occultation, and planetary observation.
  • The paper states GA required 7 times longer computation time than SA; gradient-based methods were poorly suited to the ReCon problem and rarely converged when LHS results or random initialization were used as starting points. Only ROM revisit time and GOM coverage are considered among the four possible observation objectives.
  • Evident from the text: the single-objective weighted sum uses chosen weights and scaling factors, and several parameters are fixed (e.g., 10 reconfiguration moves over a 5-year lifetime, 0.5 m GSD, 55°N regional latitude, Walker phasing parameter of 1). Simulations run for two days, and observation scheduling and constellation management are left to ongoing or future research.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.3390/s19040765

Cites

47 of 47 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure A1 Sensor payload mass vs. aperture diameter.
Figure A1 Sensor payload mass vs. aperture diameter.CC BY · PubMed Central
Figure A2 Tank mass vs. tank volume.
Figure A2 Tank mass vs. tank volume.CC BY · PubMed Central
Figure 1 Two observation modes of a ReCon: (a) Global observation mode (GOM); (b) Regional observation mode (ROM). Note that the satellite is moving along solid lines from left to right in prograde orbits (this case) and vice versa in retrograde orbits.
Figure 1 Two observation modes of a ReCon: (a) Global observation mode (GOM); (b) Regional observation mode (ROM). Note that the satellite is moving along solid lines from left to right in prograde orbits (this case) and vice versa in retrograde orbits.CC BY · PubMed Central
Figure 2 Data flow in ReCon simulation.
Figure 2 Data flow in ReCon simulation.CC BY · PubMed Central
Figure 3 Walker pattern of Globalstar satellite constellation.
Figure 3 Walker pattern of Globalstar satellite constellation.CC BY · PubMed Central
Figure 4 ReCon simulations: (a) 3D view; (b) 2D view.
Figure 4 ReCon simulations: (a) 3D view; (b) 2D view.CC BY · PubMed Central
Figure 5 Optical relationship of ground sample distance (GSD) with other parameters (not to scale).
Figure 5 Optical relationship of ground sample distance (GSD) with other parameters (not to scale).CC BY · PubMed Central
Figure 6 Reconfiguration time: (a) Target identification at t = T0 (satellite not aligned with repeat groundtracks); (b) Orbit transfer from GOM to ROM at t = T0 + TR (satellite aligned with repeat groundtracks).
Figure 6 Reconfiguration time: (a) Target identification at t = T0 (satellite not aligned with repeat groundtracks); (b) Orbit transfer from GOM to ROM at t = T0 + TR (satellite aligned with repeat groundtracks).CC BY · PubMed Central
Figure 7 Mass breakdown of a ReCon satellite.
Figure 7 Mass breakdown of a ReCon satellite.CC BY · PubMed Central
Figure 8 ReCon optimization framework.
Figure 8 ReCon optimization framework.CC BY · PubMed Central
Figure 9 Latin hypercubes: (a) Two factors (A, B) and four levels; (b) three factors (A, B, C) and four levels.
Figure 9 Latin hypercubes: (a) Two factors (A, B) and four levels; (b) three factors (A, B, C) and four levels.CC BY · PubMed Central
Figure 10 Simulated Annealing (SA) algorithm.
Figure 10 Simulated Annealing (SA) algorithm.CC BY · PubMed Central
Figure 11 SA convergence history.
Figure 11 SA convergence history.CC BY · PubMed Central
Figure 12 Genetic algorithm (GA) for ReCon optimization.
Figure 12 Genetic algorithm (GA) for ReCon optimization.CC BY · PubMed Central
Figure 13 GA convergence history: (a) Population mean; (b) best individual.
Figure 13 GA convergence history: (a) Population mean; (b) best individual.CC BY · PubMed Central
Figure 14 Illustration of modified step size calculation for discrete design variables.
Figure 14 Illustration of modified step size calculation for discrete design variables.CC BY · PubMed Central
Figure 15 Illustration of objective (J) space for discrete samples.
Figure 15 Illustration of objective (J) space for discrete samples.CC BY · PubMed Central
Figure 16 Time taken by each method and solution optimality.
Figure 16 Time taken by each method and solution optimality.CC BY · PubMed Central
Figure 17 Normalized sensitivity of design variables (RGT ratio divided by 10).
Figure 17 Normalized sensitivity of design variables (RGT ratio divided by 10).CC BY · PubMed Central
Figure 18 GA problem size and time taken for optimization.
Figure 18 GA problem size and time taken for optimization.CC BY · PubMed Central

Verification

Claims the paper makes

0 Valid · 4 Not valid · 13 Not checked · 70 Own results and interpretations

29 more claims

Create a free account to see all 87 claims and to verify the 6 references still open.

See all claims — free account
  • Probably not a substantial claim. 20
  • It may state two separate claims as one. 42
  • The quote only partly states this claim. 3

Create a free account to see these claims.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

3 of 3

Citations

Cites

Show all 47 references

Research works printed but not linked: 9

  1. Blackwell, W.J., Allen, G., Galbraith, C., Leslie, R., Osaretin, I., Scariton, M., Shields, M., Thompson, E., Toher, D., and Townzen, D. (2013, January 14). MicroMAS: A First Step Towards a Nanosatellite Constellation for Global Storm. Proceedings of the AIAA/USU Conference on Small Satellites, SSC13-XI-1, Logan, UT, USA.
  2. Paek, S.W., and de Weck, O. (October, January 29). Delta-v map of useful orbits for Earth observation missions. Proceedings of the International Astronautical Congress, IAC-14.B1.P.4.x24733, Toronto, ON, Canada.
  3. Paek, S.W. (2012). Reconfigurable Satellite Constellations (ReCon) for Geo-Spatially Adaptive Earth observation Missions. [Master’s Thesis, Massachusetts Institute of Technology].
  4. Szigethy, 2017, Chapter 9. Concurrent Design Optimization of Satellites and Reconfigurable Constellations, Horizons in Earth Science Research, Volume 15, 151
  5. Camps, A., Milne, T., and Heron, M. (2017, January 17–18). Challenges and activity in the Polar Regions: Technology, measurement and information. Proceedings of the IEEE Technical Activities Board 2017 Ad-hoc Committee, Phoenix, AZ, USA.
  6. Vallado, D.A. (2007). Fundamentals of Astrodynamics and Applications, Springer. [3rd ed.].
  7. Driggers, R.G. (2003). Encyclopedia of Optical Engineering, CRC Press.
  8. Wertz, J.R., Everett, D.F., and Puschell, J.J. (2010). Space Mission Engineering: The New SMAD, Microcosm Press.
  9. Gonzalez, J., and Saccoccia, G. (2011, January 11–15). ESA Electric Propulsion Activities. Proceedings of the 32nd International Electric Propulsion Conference, Wiesbaden, Germany.

Cited by

See all 77

Metadata

Authors
Paek, S. W., Kim, S., & de Weck, O.
Affiliations
  • Samsung (South Korea) (KR)
  • Korea Institute of Science and Technology (KR)
  • Department of Aeronautics and Astronautics, Massachusetts Institute of Technology, Cambridge, MA 02139, USA
Corresponding author
Paek, Sung Wook
Venue
Sensors
Publisher
MDPI AG
Published
2019-02-13
Volume
19
Issue
4
Pages
765
ISSN
1424-8220 (ISSN-L)
OpenAlex
W2914498930
Cited by
75
References
52
Citations by year
2026: 8 · 2025: 10 · 2024: 10 · 2023: 16 · 2022: 11 · 2021: 9 · 2020: 9 · 2019: 2
FWCI
5.70
Type
journal article
Language
en
Open access
Locations
  • Sensors (publishedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, other-oa)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DSpace@MIT (Massachusetts Institute of Technology) (submittedVersion, cc-by)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • Multidisciplinary Digital Publishing Institute
Topics
  • Physical Sciences › Engineering › Aerospace Engineering › Spacecraft Dynamics and Control
  • Physical Sciences › Engineering › Aerospace Engineering › Spacecraft Design and Technology
  • Physical Sciences › Engineering › Aerospace Engineering › Satellite Communication Systems
Keywords
Earth observation satellites, simulated annealing, genetic algorithm, disaster monitoring, observation scheduling

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (qwen3.8-omni-flash).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Optimization of Reconfigurable Satellite Constellations Using Simulated Annealing and Genetic Algorithm. https://app.cochat.ai/papers/10.3390/s19040765
The paper
Paek, S. W., Kim, S., & de Weck, O. (2019). Optimization of Reconfigurable Satellite Constellations Using Simulated Annealing and Genetic Algorithm. Sensors. https://doi.org/10.3390/s19040765

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.