pith. sign in

arxiv: 2512.18985 · v2 · submitted 2025-12-22 · 🧮 math.OC

On-Orbit Servicing-Integrated Maintenance Strategy for Satellite Constellation

Jaewoo Kim , Taehyun Sung , Woonam Hwang , Jaemyung Ahn This is my paper

Pith reviewed 2026-05-16 20:58 UTC · model grok-4.3

classification 🧮 math.OC
keywords on-orbit servicingsatellite constellationmaintenance strategyinventory managementbi-objective optimizationPareto frontierspace logisticscost reduction
0
0 comments X

The pith

Integrating on-orbit servicing into satellite constellation maintenance reduces annual costs by up to 14.5 percent while preserving required service levels.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper develops a maintenance model for satellite constellations that combines launching replacement satellites with recovering failed ones through on-orbit servicing. It uses an inventory management framework built on a parametric replenishment policy to track system performance, then casts the operator's cost-minimization goal and the provider's profit-maximization goal as a bi-objective optimization problem whose Pareto-optimal solutions are identified. A case study on a real-world-scale constellation demonstrates that the OOS-integrated strategies achieve the reported cost reductions in maintenance, launches, and manufacturing without sacrificing availability. The analysis further shows that operator-side replenishment rules stay nearly constant across the efficient frontier while provider pricing decisions drive most of the variation, and that the fraction of failures recoverable by OOS has the largest structural impact.

Core claim

The paper claims that an OOS-integrated maintenance strategy, modeled via an inventory system with parametric replenishment and solved through bi-objective optimization, produces Pareto-optimal outcomes in which annual maintenance costs fall by as much as 14.5 percent and both launch and manufacturing costs fall by approximately 25 percent relative to a pure-replacement benchmark, all while the constellation continues to meet its service-level targets. Within any given scenario the operator's maintenance policy remains almost invariant across solutions, so that movement along the Pareto frontier is driven primarily by the OOS provider's price and performance choices. Among OOS parameters, it

What carries the argument

The bi-objective optimization problem that balances the constellation operator's maintenance-cost minimization under service-level constraints against the OOS provider's profit maximization, constructed on top of an inventory model that uses a parametric replenishment policy.

Load-bearing premise

The parametric replenishment policy and the chosen recovery fraction for OOS failures accurately capture the real operational uncertainties and yield implementable decisions for both operator and provider.

What would settle it

Running the model with historical failure rates, actual OOS mission success statistics, and observed launch delays from an existing constellation and comparing the predicted cost savings against the real recorded expenses would show whether the claimed reductions hold.

read the original abstract

This paper proposes a maintenance strategy for a satellite constellation that utilizes on-orbit servicing (OOS). Under this strategy, the constellation operator addresses satellite failures in two ways: by deploying new satellites and by recovering failed satellites through OOS. We develop an inventory management model with a parametric replenishment policy for the maintenance process, which can evaluate the performance of the satellite constellation system. Based on this model, we formulate the interaction between the constellation operator -- who seeks to maintain the required service level of the constellation while minimizing maintenance cost -- and the OOS provider -- who seeks to maximize profit by selecting service price and performance levels -- as a bi-objective optimization problem and identify the corresponding Pareto-optimal solutions. A case study based on real-world-scale constellation and launch service shows that, relative to the benchmark strategy without OOS, the OOS-integrated solutions can reduce annual maintenance cost by up to 14.5%, while reducing annual launch and manufacturing costs by approximately 25% each and maintaining the required service levels. The results further show that, within a given scenario, the Pareto-optimal set is generally generated by an almost invariant maintenance strategy on the constellation operator side, whereas most variation along the Pareto frontier is driven by pricing decisions of the OOS provider side. Among the OOS-related parameters, the fraction of failures that can be recovered through OOS has the strongest structural effect.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The paper develops an inventory management model incorporating a parametric replenishment policy to evaluate satellite constellation maintenance that combines new satellite launches with on-orbit servicing (OOS) recovery of failed units. It formulates the interaction between the constellation operator (minimize maintenance cost subject to service level) and OOS provider (maximize profit via price and performance choices) as a bi-objective optimization problem, solves for Pareto-optimal solutions, and reports a case study on a real-world-scale constellation showing up to 14.5% reduction in annual maintenance cost and approximately 25% reductions in annual launch and manufacturing costs relative to a no-OOS benchmark, while preserving required service levels. The results indicate that operator-side maintenance strategies remain nearly invariant across the Pareto front, with most variation driven by OOS provider pricing, and that the recoverable failure fraction exerts the strongest structural influence.

Significance. If the inventory model and chosen parameters prove robust, the work supplies a quantitative bi-objective framework for assessing OOS integration into large satellite constellations, with concrete cost-saving estimates and the practical observation that operator decisions can remain stable while provider pricing varies. The emphasis on the recoverable-fraction parameter as the dominant driver offers a clear target for future operational studies.

major comments (3)
  1. [Case study] Case study section: the headline reductions (14.5% maintenance cost, ~25% launch and manufacturing costs) are reported as point estimates for a single chosen recovery fraction, yet the text states that this fraction has the strongest structural effect on the Pareto set. No sensitivity analysis is shown for how the savings and service-level compliance change when the fraction is varied over plausible operational ranges (e.g., 0.4–0.8), which is load-bearing for the reliability of the quantitative claims.
  2. [Inventory management model] Inventory model and parametric replenishment policy: the performance evaluation rests on a parametric replenishment rule whose parameters are not validated against stochastic failure simulations or historical data, nor accompanied by error bounds or convergence checks. Because the claimed service-level maintenance and cost savings are direct outputs of this model, the absence of such validation undermines the central quantitative results.
  3. [Bi-objective optimization] Bi-objective formulation: recovery fraction and service price are treated as decision variables whose values drive the Pareto solutions, but the case study presents only single-point outcomes without exploring the stability of the front or the operator strategy invariance under perturbations of these parameters within realistic uncertainty ranges.
minor comments (2)
  1. [Abstract] The abstract and introduction could more explicitly separate the operator’s cost-minimization objective from the provider’s profit-maximization objective when describing the Pareto set.
  2. [Model formulation] Notation for the replenishment policy parameters and the recovery fraction should be introduced with a single consolidated table to improve readability.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive and detailed comments. We address each major comment below, indicating the revisions we will incorporate to strengthen the robustness and reliability of the quantitative results.

read point-by-point responses

  1. Referee: Case study section: the headline reductions (14.5% maintenance cost, ~25% launch and manufacturing costs) are reported as point estimates for a single chosen recovery fraction, yet the text states that this fraction has the strongest structural effect on the Pareto set. No sensitivity analysis is shown for how the savings and service-level compliance change when the fraction is varied over plausible operational ranges (e.g., 0.4–0.8), which is load-bearing for the reliability of the quantitative claims.

    Authors: We agree that sensitivity analysis on the recoverable failure fraction is necessary to support the reported savings, given its structural role. In the revised manuscript we will add results varying this fraction over 0.4–0.8 and show the resulting changes in annual maintenance cost, launch and manufacturing costs, and service-level compliance. revision: yes

  2. Referee: Inventory model and parametric replenishment policy: the performance evaluation rests on a parametric replenishment rule whose parameters are not validated against stochastic failure simulations or historical data, nor accompanied by error bounds or convergence checks. Because the claimed service-level maintenance and cost savings are direct outputs of this model, the absence of such validation undermines the central quantitative results.

    Authors: The replenishment policy follows standard inventory-theoretic principles with parameters drawn from satellite-reliability literature. We will add stochastic simulation results to the revised manuscript, including convergence diagnostics and error bounds on service levels and costs, to provide explicit validation of the policy outputs. revision: yes

  3. Referee: Bi-objective formulation: recovery fraction and service price are treated as decision variables whose values drive the Pareto solutions, but the case study presents only single-point outcomes without exploring the stability of the front or the operator strategy invariance under perturbations of these parameters within realistic uncertainty ranges.

    Authors: We will extend the case-study section with perturbations of the recovery fraction and service price over realistic ranges. The added results will confirm the stability of the Pareto front and the near-invariance of the operator maintenance strategy, thereby substantiating the reported invariance property. revision: yes

Circularity Check

0 steps flagged

No circularity: model outputs are computed results, not tautological

full rationale

The paper builds an inventory model with a parametric replenishment policy, evaluates system performance, then solves a bi-objective optimization to obtain Pareto solutions for the case study. The headline cost reductions are explicit numerical outputs of that optimization under chosen parameter values (including recovery fraction). No quoted equation or step equates a prediction to its own input by construction, and no self-citation chain is shown to be load-bearing. The derivation therefore remains self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The model rests on standard inventory assumptions plus one key domain parameter (recoverable failure fraction) whose value is not derived from first principles.

free parameters (1)
  • fraction of failures recoverable through OOS
    Central parameter that controls cost savings; its value is selected or estimated to produce the reported Pareto set.
axioms (1)
  • domain assumption Parametric replenishment policy sufficiently represents the maintenance process under uncertainty
    Invoked to evaluate constellation performance without explicit stochastic simulation details.

pith-pipeline@v0.9.0 · 5551 in / 1196 out tokens · 20787 ms · 2026-05-16T20:58:08.857245+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

38 extracted references · 38 canonical work pages

  1. [1]

    An Updated Comparison of Four Low Earth Orbit Satellite Constellation Systems to Provide Global Broadband,

    Pachler, N., del Portillo, I., Crawley, E. F., and Cameron, B. G., “An Updated Comparison of Four Low Earth Orbit Satellite Constellation Systems to Provide Global Broadband,”2021 IEEE International Conference on Communications Workshops (ICC Workshops), IEEE, 2021, pp.1–7. https://doi.org/10.1109/ICCWorkshops50388.2021.9473799

    work page doi:10.1109/iccworkshops50388.2021.9473799 2021

  2. [2]

    Pachler, E

    Pachler, N., Crawley, E. F., and Cameron, B. G., “Flooding the Market: Comparing the Performance of Nine Broadband Megaconstellations,”IEEE Wireless Communications Letters, Vol. 13, No. 9, 2024, pp. 2397–2401. https://doi.org/10.1109/LWC.2024.3416531

    work page doi:10.1109/lwc.2024.3416531 2024

  3. [3]

    Space Logistics Modeling and Optimization: Review of the State of the Art,

    Ho, K., “Space Logistics Modeling and Optimization: Review of the State of the Art,”Journal of Spacecraft and Rockets, Published Online. https://doi.org/10.2514/1.A35982

    work page doi:10.2514/1.a35982

  4. [4]

    Optimal Satellite Constellation Spare Strategy Using Multi-Echelon Inventory Control,

    Jakob. P., Shimizu, S., Yoshikawa, S., and Ho, K., “Optimal Satellite Constellation Spare Strategy Using Multi-Echelon Inventory Control,”Journal of Spacecraft and Rockets, Vol. 56, No. 5, 2019, pp. 1449–1461. https://doi.org/10.2514/1.A34387

    work page doi:10.2514/1.a34387 2019

  5. [5]

    Replenishment Strategy for Satellite Constellation with Dual Supply Modes,

    Kim, J., Ahn, J., and Sung, T., “Replenishment Strategy for Satellite Constellation with Dual Supply Modes,”Journal of Spacecraft and Rockets, Vol. 62, No. 5, 2025, pp. 1567–1583. https://doi.org/10.2514/1.A36281

    work page doi:10.2514/1.a36281 2025

  6. [6]

    A., Pyke, D

    Silver, E. A., Pyke, D. F., and Thomas, D. J.,Inventory and Production Management in Supply Chains,4th ed., CRC Press, Boca Raton, FL, 2016, Chap. 6

    work page 2016

  7. [7]

    Optimal Deployment of Satellite Mega-Constellation,

    Sung, T., and Ahn, J., “Optimal Deployment of Satellite Mega-Constellation,”Acta Astonautica, Vol. 202, Jan. 2023, pp. 653–669. https://doi.org/10.1016/j.actaastro.2022.10.027

    work page doi:10.1016/j.actaastro.2022.10.027 2023

  8. [8]

    Li, D.-Y

    Li, W.-J., Cheng, D.-Y., Liu, X.-G., Wang, Y.-B., Shi, W.-H., Tang, Z.-X., Gao, F., Zeng, F.-M., Chai, H.-Y, Luo, W.-B., Cong, Q., and Gao, Z.-L., “On-Orbit Service (OOS) of Spacecraft: A Review of Engineering Developments,”Progress in Aerospace Sciences, Vol. 108, Jul. 2019, pp. 32-120. https://doi.org/10.1016/j.paerosci.2019.01.004

    work page doi:10.1016/j.paerosci.2019.01.004 2019

  9. [9]

    On-Orbit Servicing, Assembly, and Manufacturing (OSAM) State of Play,

    Arney, D., Sutherland, R., Mulvaney, J., Steinkoenig, D., Stockdale, C., and Farley, M., “On-Orbit Servicing, Assembly, and Manufacturing (OSAM) State of Play,” NASA TR 20210022660, 2021

    work page 2021

  10. [10]

    Game Changer: On-Orbit Servicing,

    Davis, J. P., Mayberry, J. P., and Penn, J. P., “Game Changer: On-Orbit Servicing,” The Aerospace Corporation Center for Space Policy and Strategy, May 2019. https://csps.aerospace.org/papers/game-changer-orbit-servicing

    work page 2019

  11. [11]

    Game Changer: In-Space Servicing, Assembly, and Manufacturing for the New Space Economy,

    Cavaciuti, A. J., Heying, J. H., and Davis, J., “Game Changer: In-Space Servicing, Assembly, and Manufacturing for the New Space Economy,” The Aerospace Corporation Center for Space Policy and Strategy, Jul. 2022. https://csps.aerospace.org/papers/game-changer-space-servicing-assembly-and-manufacturing-new-space-economy

    work page 2022

  12. [12]

    Space Systems Flexibility Provided by On-Orbit Servicing: Part 1,

    Saleh, J. H., Lamassoure, E., and Hastings, D. E., “Space Systems Flexibility Provided by On-Orbit Servicing: Part 1,”Journal of Spacecraft and Rockets, Vol. 39, No. 4, 2002, pp. 551-560. https://doi.org/10.2514/2.3844 30

    work page doi:10.2514/2.3844 2002

  13. [13]

    Space Systems Flexibility Provided by On-Orbit Servicing: Part 2,

    Lamassoure, E., Saleh, J. H., and Hastings, D. E., “Space Systems Flexibility Provided by On-Orbit Servicing: Part 2,”Journal of Spacecraft and Rockets, Vol. 39, No. 4, 2002, pp. 561-570. https://doi.org/10.2514/2.3845

    work page doi:10.2514/2.3845 2002

  14. [14]

    On-Orbit Upgrade and Repair: The Hubble Space Telescope Example,

    Joppin, C., and Hastings, D. E., “On-Orbit Upgrade and Repair: The Hubble Space Telescope Example,”Journal of Spacecraft and Rockets, Vol. 43, No. 3, 2006, pp. 614-625. https://doi.org/10.2514/1.15496

    work page doi:10.2514/1.15496 2006

  15. [15]

    On-Orbit Servicing: A New Value Proposition for Satellite Design and Operation,

    Long, A. M., Richards, M. G., and Hastings, D. E., “On-Orbit Servicing: A New Value Proposition for Satellite Design and Operation,”Journal of Spacecraft and Rockets, Vol. 44, No. 4, 2007, pp. 964-976. https://doi.org/10.2514/1.27117

    work page doi:10.2514/1.27117 2007

  16. [16]

    On-Orbit Servicing System Assessment and Optimization Methods Based on Lifecycle Simulation Under Mixed Aleatory and Epistemic Uncertainties,

    Yao, W., Chen, X., Huang, Y., and van Tooren, M., “On-Orbit Servicing System Assessment and Optimization Methods Based on Lifecycle Simulation Under Mixed Aleatory and Epistemic Uncertainties,”Acta Astronautica, Vol. 87, Jun.-July 2013, pp. 107-126. https://doi.org/10.1016/j.actaastro.2013.02.005

    work page doi:10.1016/j.actaastro.2013.02.005 2013

  17. [17]

    Economic Value Analysis of On-Orbit Servicing for Geosynchronous Communication Satellites,

    Liu, Y., Zhao, Y., Tan, C., Liu, H., and Liu, Y., “Economic Value Analysis of On-Orbit Servicing for Geosynchronous Communication Satellites,”Acta Astronautica, Vol. 180, Mar. 2021, pp. 176–188. https://doi.org/10.1016/j.actaastro.2020.11.040

    work page doi:10.1016/j.actaastro.2020.11.040 2021

  18. [18]

    System Architecting of GEO Communication Satellite Considering On-Orbit Refueling,

    Kim, J., and Ahn, J., “System Architecting of GEO Communication Satellite Considering On-Orbit Refueling,”Journal of Spacecraft and Rockets, 2025, published online. https://doi.org/10.2514/1.A36403

    work page doi:10.2514/1.a36403 2025

  19. [19]

    Intelsat-901 Satellite, With MEV-1 Servicer Attached, Resumes Service,

    Henry, C., “Intelsat-901 Satellite, With MEV-1 Servicer Attached, Resumes Service,” Space News, 2020, https://spacenews.com/intelsat-901-satellite-with-mev-1-servicer-attached-resumes-service/[accessed 14 Oct. 2025]

    work page 2020

  20. [20]

    On-Orbit Servicing System Architectures for Proliferated Low-Earth-Orbit Constellations,

    Luu, M. A., and Hastings, D. E., “On-Orbit Servicing System Architectures for Proliferated Low-Earth-Orbit Constellations,” Journal of Spacecraft and Rockets, Vol. 59, No. 6, pp. 1946–1965. https://doi.org/10.2514/1.A35393

    work page doi:10.2514/1.a35393 1946

  21. [21]

    R., Meissinger, H

    Wertz, J. R., Meissinger, H. F., Newman, L. K., and Smit, G. N.,Mission Geometry; Orbit and Constellation Design and Management, Microcosm Press, El Segundo, CA, 2001, Chap. 12, 13

    work page 2001

  22. [22]

    A.,Fundamentals of Astrodynamics and Applications, 5th ed., Microcosm Press, Torrance, CA, 2022, Chap

    Vallado, D. A.,Fundamentals of Astrodynamics and Applications, 5th ed., Microcosm Press, Torrance, CA, 2022, Chap. 9

    work page 2022

  23. [23]

    Circle-to-Circle Constant-Thrust Orbit Raising,

    Alfano, S., and Thorne, J. D., “Circle-to-Circle Constant-Thrust Orbit Raising,”The Journal of the Astronautical Sciences, Vol. 42, No. 1, 1994, pp. 35–45

    work page 1994

  24. [24]

    Mathematical Relation Between the Hohmann Transfer and Continuous-Low Thrust Maneuvers,

    Bettinger, R. A., and Black, J. T., “Mathematical Relation Between the Hohmann Transfer and Continuous-Low Thrust Maneuvers,”Acta Astronautica, Vol. 96, Mar.-Apr. 2014, pp. 42–44. https://doi.org/10.1016/j.actaastro.2013.11.020

    work page doi:10.1016/j.actaastro.2013.11.020 2014

  25. [25]

    E.,An Introduction to Reliability and Maintainability Engineering, 3rd ed., Waveland Press, Long Grove, IL, 2019, Chap

    Ebling, C. E.,An Introduction to Reliability and Maintainability Engineering, 3rd ed., Waveland Press, Long Grove, IL, 2019, Chap. 9, 10

    work page 2019

  26. [26]

    TheFrequencyDistributionoftheDifferenceBetweenTwoPoissonVariatesBelongingtoDifferentPopulations,

    Skellam,J.G.,“TheFrequencyDistributionoftheDifferenceBetweenTwoPoissonVariatesBelongingtoDifferentPopulations,” Journal of the Royal Statistical Society, Vol. 109, No. 3, 1946, p. 296. https://www.jstor.org/stable/2981372

    work page arXiv 1946

  27. [27]

    A.,Handbook of Mathematical Functions: with Formulas, Graphs, and Mathematical Tables, Dover Publications, New York, NY, 1965, Chap

    Abramowitz, M., and Stegun, I. A.,Handbook of Mathematical Functions: with Formulas, Graphs, and Mathematical Tables, Dover Publications, New York, NY, 1965, Chap. 9

    work page 1965

  28. [28]

    Managing Supply Chain Inventories: A Multiple Retailer, One Warehouse, Multiple Supplier Model,

    Ganeshan, R., “Managing Supply Chain Inventories: A Multiple Retailer, One Warehouse, Multiple Supplier Model,” International Journal of Production Economics, Vol. 59, No. 1, 1999, pp. 341-354. https://doi.org/10.1016/S0925-5273(98)00115-7

    work page doi:10.1016/s0925-5273(98)00115-7 1999

  29. [29]

    Modelling a Two-Echelon (s, Q) Distribution System,

    Kim, J. D., “Modelling a Two-Echelon (s, Q) Distribution System,” Ph.D. Dissertation, The Pennsylvania State University, University Park, PA, 1991

    work page 1991

  30. [30]

    Chopra, S.,Supply Chain Management: Strategy, Planning, and Operation, 7th ed., Pearson Education, London, England, 2019. 31

    work page 2019

  31. [31]

    Valuation of On-Orbit Servicing in Proliferated Low-Earth Orbit Constellations,

    Luu, M., and Hastings, D. E., “Valuation of On-Orbit Servicing in Proliferated Low-Earth Orbit Constellations,”Accelerating Space Commerce, Exploration, and New Discovery Conference, ASCEND 2020, AIAA Paper 2020-4127, 2020. https://doi.org/10.2514/6.2021-4207

    work page doi:10.2514/6.2021-4207 2020

  32. [32]

    Quantification of the Responsiveness of On-Orbit Servicing Infrastructure for Modularized Earth-Orbiting Platforms,

    du Jonchay, T. S., and Ho, K., “Quantification of the Responsiveness of On-Orbit Servicing Infrastructure for Modularized Earth-Orbiting Platforms,”Acta Astronautica, Vol. 132, 2017, pp. 192–203. https://doi.org/10.1016/j.actaastro.2016.12.021

    work page doi:10.1016/j.actaastro.2016.12.021 2017

  33. [33]

    Semi-Analytical Model for Design and Analysis of On-Orbit Servicing Architecture,

    Ho, K., Wang, H., DeTrempe, P. A., du Jonchay, T. S., and Tomita, K., “Semi-Analytical Model for Design and Analysis of On-Orbit Servicing Architecture,”Journal of Spacecraft and Rockets, Vol. 57, No. 6, 2020, pp. 1129–1138. https://doi.org/10.2514/1.A34663

    work page doi:10.2514/1.a34663 2020

  34. [34]

    In-SpaceServicing,Assembly,andManufacturing: Benefits,Challenges,andPolicyOptions,

    Howard, K. L., Pekar-Carpenter, K., Mai, C. L., Burgott, K., Reid, J., Shouse, B., Rando, J., Han, R., Harner, P., and Powell, T., “In-SpaceServicing,Assembly,andManufacturing: Benefits,Challenges,andPolicyOptions,”U.S.GovernmentAccountability Office, GAO-25-107555, Jul. 2025. https://www.gao.gov/products/gao-25-107555

    work page 2025

  35. [35]

    Survey of Multi-Objective Optimization Methods for Engineering,

    Marler, R. T., and Arora, J. S., “Survey of Multi-Objective Optimization Methods for Engineering,”Structural and Multidisci- plinary Optimization, Vol. 26, Mar. 2004, pp. 396-395. https://doi.org/10.1007/s00158-003-0368-6

    work page doi:10.1007/s00158-003-0368-6 2004

  36. [36]

    Y.,Mathematical Psychics, P

    Edgeworth, F. Y.,Mathematical Psychics, P. Keagan, London, England, 1881

    work page

  37. [37]

    S., and McLure, M., Oxford University Press, Oxford, England, 2014

    Pareto, V.,Manual of Political Economy: A Critical and Variorum Edition, edited by Montesano, A., Zanni, A., Bruni, L., Chipman, J. S., and McLure, M., Oxford University Press, Oxford, England, 2014

    work page 2014

  38. [38]

    and Pratap, A

    Deb, K., Pratap, A., Agarwal, S., and Meyarivan, T., “A Fast and Elitist Multiobjective Genetic Algorithm: NSGA-II,”IEEE Transactions on Evolutionary Computation, Vol. 6, No. 2, 2009, pp. 182-197. https://doi.org/10.1109/4235.996017 32

    work page doi:10.1109/4235.996017 2009