Predicting the performance of different electric vertical take-off and landing (eVTOL) vehicle designs is paramount to vehicle manufacturers and hobbyists. These vehicles' maximum flight time (endurance) and maximum flight distance (range) depend on design and operational parameters relating to their structure, propulsion system, payload, and mission profile. In recent years, sophisticated physics-based models have been developed to estimate and optimize their aerodynamic, propulsion, and electrical performance. Integrating and simulating those models can closely estimate a vehicle's endurance and range. However, this demands advanced knowledge of different subsystems utilized and extensive computational resources limiting the wide-scale utilization of such models. This paper showcases the development and implementation of a framework to train simpler machine learning-based surrogates. The surrogate models are trained on a limited number of eVTOL performance estimates generated by physics-based models and can mimic them accurately. Forty-seven thousand eVTOL vehicle designs were simulated to generate the training data for various machine-learning models. These include several decision tree models, K-nearest neighbor models, linear regression models, and a multi-perceptron neural network model. Vehicle design and operational parameters such as propeller size, payload mass, drag coefficient, velocity, and motor and battery parameters are used as features, and vehicle endurance and range estimates are used as targets. Compared to the alternative approaches, these surrogate models are computationally very efficient and easy to understand and use. Testing on hold-out datasets shows excellent performance, with multiple models having a mean average percentage error of less than 2% in estimating vehicle endurance and range.
Citation: Jubilee Prasad Rao, Sai Naveen Chimata. Machine learning-based surrogates for eVTOL performance prediction and design optimization[J]. Metascience in Aerospace, 2024, 1(3): 246-267. doi: 10.3934/mina.2024011
Predicting the performance of different electric vertical take-off and landing (eVTOL) vehicle designs is paramount to vehicle manufacturers and hobbyists. These vehicles' maximum flight time (endurance) and maximum flight distance (range) depend on design and operational parameters relating to their structure, propulsion system, payload, and mission profile. In recent years, sophisticated physics-based models have been developed to estimate and optimize their aerodynamic, propulsion, and electrical performance. Integrating and simulating those models can closely estimate a vehicle's endurance and range. However, this demands advanced knowledge of different subsystems utilized and extensive computational resources limiting the wide-scale utilization of such models. This paper showcases the development and implementation of a framework to train simpler machine learning-based surrogates. The surrogate models are trained on a limited number of eVTOL performance estimates generated by physics-based models and can mimic them accurately. Forty-seven thousand eVTOL vehicle designs were simulated to generate the training data for various machine-learning models. These include several decision tree models, K-nearest neighbor models, linear regression models, and a multi-perceptron neural network model. Vehicle design and operational parameters such as propeller size, payload mass, drag coefficient, velocity, and motor and battery parameters are used as features, and vehicle endurance and range estimates are used as targets. Compared to the alternative approaches, these surrogate models are computationally very efficient and easy to understand and use. Testing on hold-out datasets shows excellent performance, with multiple models having a mean average percentage error of less than 2% in estimating vehicle endurance and range.
| [1] | Gupta SG, Ghonge DM, Jawandhiya PM (2013) Review of unmanned aircraft system (uas). Int J Adv Res Comput Eng Technol 2. https://ssrn.com/abstract = 3451039 |
| [2] | Singhal G, Bansod B, Mathew L (2018) Unmanned aerial vehicle classification, applications and challenges: A review. |
| [3] |
Zhang B, Song ZX, Zhao F, et al. (2022) Overview of propulsion systems for unmanned aerial vehicles. Energies 15: 455. https://doi.org/10.3390/en15020455 doi: 10.3390/en15020455
|
| [4] | Bershadsky D, Haviland S, Johnson EN (2016) Electric multirotor uav propulsion system sizing for performance prediction and design optimization. In 57th AIAA/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference, 0581. https://doi.org/10.2514/6.2016-0581 |
| [5] | Al Hammadi M, AlMesafri N, Zafar S, et al. (2023) Design and analytical analyses of evtol uav performance calculator for power and energy. In 2023 10th International Conference on Recent Advances in Air and Space Technologies (RAST), 1–6. IEEE. |
| [6] |
Villegas A, Mishkevich V, Gulak Y, et al. (2017) Analysis of key elements to evaluate the performance of a multirotor unmanned aerial–aquatic vehicle. Aerosp Sci Technol 70: 412–418. https://doi.org/10.1016/j.ast.2017.07.046 doi: 10.1016/j.ast.2017.07.046
|
| [7] |
Song YH, Kim DK (2018) Development of coaxial propeller test facility and experimental study on hover performance characteristics for drone. J Korean Soc Aeronaut Space Sci 46: 59–67. https://doi.org/10.5139/JKSAS.2018.46.1.59 doi: 10.5139/JKSAS.2018.46.1.59
|
| [8] |
Prasad Rao J, Holzsager JE, Maia MM, et al. (2022) Experimental study into optimal configuration and operation of two-four rotor coaxial systems for evtol vehicles. Aerospace 9: 452. https://doi.org/10.3390/aerospace9080452 doi: 10.3390/aerospace9080452
|
| [9] |
Prasad RJ, Diez FJ (2018) Novel cyclic blade pitching mechanism for wind and tidal energy turbine applications. Energies 11: 3328. https://doi.org/10.3390/en11123328 doi: 10.3390/en11123328
|
| [10] |
Traub LW (2011) Range and endurance estimates for battery-powered aircraft. J Aircraft 48: 703–707. https://doi.org/10.2514/1.C031027 doi: 10.2514/1.C031027
|
| [11] | Lawrence D, Mohseni K (2005) Efficiency analysis for long duration electric mavs. In Infotech@ Aerospace, 7090. https://doi.org/10.2514/6.2005-7090 |
| [12] |
Chang T, Yu H (2015) Improving electric powered uavs' endurance by incorporating battery dumping concept. Procedia Eng 99: 168–179. https://doi.org/10.1016/j.proeng.2014.12.522 doi: 10.1016/j.proeng.2014.12.522
|
| [13] | Abdilla A, Richards A, Burrow S (2015) Power and endurance modelling of battery-powered rotorcraft. In 2015 IEEE/RSJ international conference on intelligent robots and systems (IROS), 675–680. IEEE. https://doi.org/10.1109/IROS.2015.7353445 |
| [14] | Drela M (1989) Xfoil: An analysis and design system for low reynolds number airfoils. In Low Reynolds Number Aerodynamics: Proceedings of the Conference Notre Dame, Indiana, USA, 5–7, 1–12. Springer. |
| [15] | Drela M (2006) Qprop theory document. web. mit. edu/drela/Public/Qprop, 2006. |
| [16] | Bajwa OI, Baluch HA, Saeed HA (2022) Machine learning techniques for predicting uav design parameters: Overview, methodology and potential benefits. In 2022 19th International Bhurban Conference on Applied Sciences and Technology (IBCAST), 249–255. IEEE. |
| [17] | Smart JT, Alonso JJ (2019) Primary weight estimation for evtols via explicit analysis and surrogate regression. In AIAA Aviation 2019 Forum, 3679. https://doi.org/10.2514/6.2019-3679 |
| [18] | Karali H, Demirezen MU, Yukselen MA, et al. (2020) Design of a deep learning based nonlinear aerodynamic surrogate model for uavs. In AIAA Scitech 2020 forum, 1288. https://doi.org/10.2514/6.2020-1288 |
| [19] |
Dupuis R, Jouhaud JC, Sagaut P (2018) Surrogate modeling of aerodynamic simulations for multiple operating conditions using machine learning. Aiaa J 56: 3622–3635. https://doi.org/10.2514/1.J056405 doi: 10.2514/1.J056405
|
| [20] | Poggi C, Rossetti M, Bernardini G, et al. (2022) Surrogate models for predicting noise emission and aerodynamic performance of propellers. Aerosp Sci Technol 125: 107016. |
| [21] | Elrefaie M, Ayman T, Elrefaie M, et al. (2024) Surrogate modeling of the aerodynamic performance for airfoils in transonic regime. In AIAA SCITECH 2024 Forum, 2220. https://doi.org/10.2514/6.2024-2220 |
| [22] |
Sun G, Wang S (2019) A review of the artificial neural network surrogate modeling in aerodynamic design. P I Mech Eng G-J Aer Eng 233: 5863–5872. https://doi.org/10.1177/0954410019864485 doi: 10.1177/0954410019864485
|
| [23] | Giorgiani do Nascimento R, Fricke K, Viana F (2020) Quadcopter control optimization through machine learning. In AIAA Scitech 2020 Forum, 1148. https://doi.org/10.2514/6.2020-1148 |
| [24] |
Koch W, Mancuso R, West R, et al. (2019) Reinforcement learning for uav attitude control. ACM Trans Cyber-Phys Syst 3: 1–21. https://doi.org/10.1145/3301273 doi: 10.1145/3301273
|
| [25] |
Li P, Khan J (2023) Feature extraction and analysis of landscape imaging using drones and machine vision. Soft Comput 1–19. https://doi.org/10.1007/s00500-023-09352-w doi: 10.1007/s00500-023-09352-w
|
| [26] |
Al-Sharman MK, Zweiri Y, Jaradat MAK, et al. (2019) Deep-learning-based neural network training for state estimation enhancement: Application to attitude estimation. IEEE Transactions on Instrumentation and Measurement, 69: 24–34. https://doi.org/10.1109/TIM.2019.2895495 doi: 10.1109/TIM.2019.2895495
|
| [27] | Pedrioli A, Capone P, Righi M, et al. (2024) Model-si: Modeling and simulation-multi-fidelity surrogate model of an evtol for certification. In AIAA SCITECH 2024 Forum, 1624. https://doi.org/10.2514/6.2024-1624 |
| [28] |
Yin J, Hu Z, Mourelatos ZP, et al. (2023) Efficient reliability-based path planning of off-road autonomous ground vehicles through the coupling of surrogate modeling and rrt. IEEE T Intell Transp Syst. https://doi.org/10.1109/TITS.2023.3296651 doi: 10.1109/TITS.2023.3296651
|
| [29] |
Stolfi DH, Danoy G (2023) Optimising robot swarm formations by using surrogate models and simulations. Appl Sci 13: 5989. https://doi.org/10.3390/app13105989 doi: 10.3390/app13105989
|
| [30] | Ozisikyilmaz B, Memik G, Choudhary A (2008) Machine learning models to predict performance of computer system design alternatives. In 2008 37th International Conference on Parallel Processing, 495–502. IEEE. https://doi.org/10.1109/ICPP.2008.36 |
| [31] | Pudaruth S (2014) Predicting the price of used cars using machine learning techniques. Int J Inf Comput Technol 4: 753–764. |
| [32] |
Lu H, Cantero-Chinchilla S, Yang X, et al. (2024) Deep learning uncertainty quantification for ultrasonic damage identification in composite structures. Compos Struct 338: 118087. https://doi.org/10.1016/j.compstruct.2024.118087 doi: 10.1016/j.compstruct.2024.118087
|
| [33] | Sihombing DJC, Dexius JU, Manurung J, et al. (2022) Design and analysis of automated machine learning (automl) in powerbi application using pycaret. In 2022 International Conference of Science and Information Technology in Smart Administration (ICSINTESA), 89–94. IEEE. https://doi.org/10.1109/ICSINTESA56431.2022.10041543 |
| [34] | Sundararajan M, Najmi A (2020) The many Shapley values for model explanation. In International conference on machine learning, 9269–9278. |
| [35] |
Visani G, Bagli E, Chesani F, et al. (2022) Statistical stability indices for lime: Obtaining reliable explanations for machine learning models. J Oper Res Soc 73: 91–101. https://doi.org/10.1080/01605682.2020.1865846 doi: 10.1080/01605682.2020.1865846
|
Figures(13) / Tables(2)
Jubilee Prasad Rao, Sai Naveen Chimata. Machine learning-based surrogates for eVTOL performance prediction and design optimization[J]. Metascience in Aerospace, 2024, 1(3): 246-267. doi: 10.3934/mina.2024011
DownLoad: