Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
4.5
2.7
Journals
Applied Sciences
Special Issues
Applications of Machine Learning and Optimal Control to Aerospace Systems
share announcement

Applications of Machine Learning and Optimal Control to Aerospace Systems

A special issue of Applied Sciences (ISSN 2076-3417). This special issue belongs to the section "Aerospace Science and Engineering".

Deadline for manuscript submissions: closed (30 October 2023) | Viewed by 10110

Special Issue Editor

Prof. Dr. Sungsu Park

E-Mail Website
Guest Editor
Department of Aerospace Engineering, Sejong University, Seoul 05006, Korea
Interests: optimal control and reinforcement learning with applications to aerospace systems
Special Issues, Collections and Topics in MDPI journals

Special Issue Information

Dear Colleagues,

Although machine learning and optimal control have been successfully applied in various fields, their application within aerospace systems is still in its infancy. Due to the high standards of safety—a critical aspect of aerospace engineering—the data-driven machine learning approach must be both verifiable and interpretable, and the optimal control technology must be computationally tractable for onboard autonomy, even in complex and dynamic environments.

In response to these technological challenges, a variety of novel approaches and algorithms have arisen, offering many ways that aerospace systems can reap the benefits of machine learning and optimal control, especially in the areas of guidance, navigation, and control systems.

In this Special Issue, we would like to explore novel research and recent advances in machine learning and optimal control in aerospace applications. For this purpose, authors are invited to submit full research articles, as well as comprehensive review and survey papers, including, but not limited to, the following topics:

  • Emerging technology in machine learning and optimal control;
  • Machine learning techniques for safety-critical applications;
  • Onboard optimal guidance and control for reusable rockets;
  • Onboard autonomy to operate aerospace systems safely within urban environments;
  • Collision avoidance maneuvers using machine learning techniques;
  • Reinforcement learning-based flight control system design;
  • Multi-objective optimization-based flight control system design;
  • Trajectory optimization using machine learning;
  • Machine learning-based digital twin of aerospace system;
  • Intelligent navigation systems;
  • Autonomous air traffic control;
  • Explainable deep learning.

Prof. Dr. Sungsu Park
Guest Editor

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Applied Sciences is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2400 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

Keywords

  • machine learning
  • deep learning
  • reinforcement learning
  • optimal control
  • embedded optimization
  • aerospace system
  • trajectory optimization
  • guidance
  • navigation and control
  • urban air mobility

Published Papers (5 papers)

Download All Papers
Order results
Result details
Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:

Research

17 pages, 3127 KiB  
Article
Three-Dimensional Dubins-Path-Guided Continuous Curvature Path Smoothing
by Sungsu Park
Appl. Sci. 2022, 12(22), 11336; https://doi.org/10.3390/app122211336 - 8 Nov 2022
Cited by 1 | Viewed by 2032
Abstract
This paper presents an efficient three-dimensional (3D) Dubins path design and a new continuous curvature path-smoothing algorithm. The proposed 3D Dubins path is a simple extension of the 2D path and serves as a reference path generator to guide the proposed smoothing algorithm. [...] Read more.
This paper presents an efficient three-dimensional (3D) Dubins path design and a new continuous curvature path-smoothing algorithm. The proposed 3D Dubins path is a simple extension of the 2D path and serves as a reference path generator to guide the proposed smoothing algorithm. In the smoothing algorithm, the reference path is approximated by several cubic Bezier curves to generate a parametric path that simultaneously satisfies curvature continuity and maximum curvature requirements. The algorithm also provides a criterion to select a required number of Bezier curves to strictly meet the maximum curvature constraint. Therefore, our algorithm can be used for 3D path-following applications in many areas such as aerospace and robotics. The numerical simulation results show that the proposed algorithm produces a continuous curvature path that passes through all waypoints with a slight increase in path length compared with the original 3D Dubins reference path. Full article
(This article belongs to the Special Issue Applications of Machine Learning and Optimal Control to Aerospace Systems)
►▼ Show Figures

Figure 1

19 pages, 6073 KiB  
Article
A Technical Device for Determining the Predispositions of Students—Air Traffic Controllers and Pilots during Multitasking Training
by Matej Antosko and Pavol Lipovsky
Appl. Sci. 2022, 12(21), 11171; https://doi.org/10.3390/app122111171 - 4 Nov 2022
Viewed by 1382
Abstract
The specific professions of aviation personnel include the professions of the pilot and air traffic controller. These occupations are specific in that while performing their work, they must be able to simultaneously operate the devices in the handling area and in the pedipulation [...] Read more.
The specific professions of aviation personnel include the professions of the pilot and air traffic controller. These occupations are specific in that while performing their work, they must be able to simultaneously operate the devices in the handling area and in the pedipulation area, supplemented by acoustic sensations in the form of correspondence between flying and ground stations. The performance requirements of pilots and air traffic controllers place high demands on their health, psychological condition, attention, and concentration, due to being in constant pursuit of minimization of erroneous decisions, otherwise defined as the human factor in aviation. This article is focused on the development and testing of a technical device for measuring the relative error rate of students in multitasking tasks in preparation for employment. The main result is a designed technical device with hardware (HW) and software (SW) parts. An experimental method was used to measure the qualitative and quantitative performance indicators of the test subjects. The results of the experiment were observed and evaluated based on the analytical-synthetic method based on critical thinking. By comparing and abstracting the measured data, the reference values of the performance indicators of the tested subjects were determined. The selection of the final sample of subjects consisted of two phases. In the first phase, questionnaires were evaluated, and in the second phase, reaction time measurements during multitasking tasks using technical equipment were evaluated. Based on the measurements, an error ratio was defined, which could be graphically represented. The testing proved the full functionality of the designed technical equipment for these purposes in aviation education. Full article
(This article belongs to the Special Issue Applications of Machine Learning and Optimal Control to Aerospace Systems)
►▼ Show Figures

Figure 1

16 pages, 5124 KiB  
Article
Heat Conduction Plate Layout Optimization Using Physics-Driven Convolutional Neural Networks
by Yang Sun, Abdussalam Elhanashi, Hao Ma and Mario Rosario Chiarelli
Appl. Sci. 2022, 12(21), 10986; https://doi.org/10.3390/app122110986 - 30 Oct 2022
Cited by 2 | Viewed by 1532
Abstract
Optimizing heat conduction layout is essential during engineering design, especially for sensible thermal products. However, when the optimization algorithm iteratively evaluates different loading cases, the traditional numerical simulation methods usually lead to a substantial computational cost. To effectively reduce the computational effort, data-driven [...] Read more.
Optimizing heat conduction layout is essential during engineering design, especially for sensible thermal products. However, when the optimization algorithm iteratively evaluates different loading cases, the traditional numerical simulation methods usually lead to a substantial computational cost. To effectively reduce the computational effort, data-driven approaches are used to train a surrogate model as a mapping between the prescribed external loads and various geometry. However, the existing model is trained by data-driven methods, which require intensive training samples from numerical simulations and do not effectively solve the problem. Choosing the steady heat conduction problems as examples, this paper proposes a physics-driven convolutional neural networks (PD-CNNs) method to infer the physical field solutions for randomly varied loading cases. After that, the particle swarm optimization (PSO) algorithm is used to optimize the sizes, and the positions of the hole masks in the prescribed design domain and the average temperature value of the entire heat conduction field is minimized. The goal of reducing heat transfer is achieved. Compared with the existing data-driven approaches, the proposed PD-CNN optimization framework predicts field solutions that are highly consistent with conventional simulation results. However, the proposed method generates the solution space without pre-obtained training data. We obtained thermal intensity results for holes 1, hole 2, hole 3, and hole 4 with 0.3948, 0.007, 0.0044, and 0.3939, respectively, by optimization PD-CNN model. Full article
(This article belongs to the Special Issue Applications of Machine Learning and Optimal Control to Aerospace Systems)
►▼ Show Figures

Figure 1

27 pages, 665 KiB  
Article
Comparison of Optimization Techniques and Objective Functions Using Gas Generator and Staged Combustion LPRE Cycles
by Suniya Sadullah Khan, Ihtzaz Qamar, Muhammad Umer Sohail, Raees Fida Swati, Muhammad Azeem Ahmad and Saad Riffat Qureshi
Appl. Sci. 2022, 12(20), 10462; https://doi.org/10.3390/app122010462 - 17 Oct 2022
Cited by 1 | Viewed by 1646
Abstract
This paper compares various optimization techniques and objective functions to obtain optimum rocket engine performances. This research proposes a modular optimization framework that provides an optimum design for Gas Generator (GG) and Staged Combustion (SC) Liquid Propellant Rocket Engines. This process calculates the [...] Read more.
This paper compares various optimization techniques and objective functions to obtain optimum rocket engine performances. This research proposes a modular optimization framework that provides an optimum design for Gas Generator (GG) and Staged Combustion (SC) Liquid Propellant Rocket Engines. This process calculates the ideal rocket engine performance by applying seven different optimization techniques: Simulated Annealing (SA), Nelder Mead (NM), Cuckoo Search Algorithm (CSA), Particle Swarm Optimization (PSO), Pigeon-Inspired Optimization (PIO), Genetic Algorithm (GA) and a novel hybrid GA-PSO technique named GA-Swarm. This new technique combines the superior search capability of GA with the efficient constraint matching capability of PSO. This research also compares objective functions to determine the most suitable function for GG and SC cycle rocket engines. Three single objective functions are used to minimize the Gross Lift-Off Weight and to maximize Specific Impulse and the Thrust-to-Weight ratio. A fourth multiobjective function is used to simultaneously maximize both Specific Impulse and Thrust-to-Weight ratio. This framework is validated against a pump-fed rocket, and results are within 1% of the actual rocket engine mass. The results of this research indicate that PSO and GA-Swarm produce optimum results for all objective functions. Finally, the most suitable objective function to use while comparing these two cycles is the Gross Lift-Off Weight. Full article
(This article belongs to the Special Issue Applications of Machine Learning and Optimal Control to Aerospace Systems)
►▼ Show Figures

Figure 1

18 pages, 4150 KiB  
Article
Avoiding Obstacles via Missile Real-Time Inference by Reinforcement Learning
by Daseon Hong and Sungsu Park
Appl. Sci. 2022, 12(9), 4142; https://doi.org/10.3390/app12094142 - 20 Apr 2022
Cited by 4 | Viewed by 2560
Abstract
In the contemporary battlefield where complexity has increased, the enhancement of the role and ability of missiles has become crucial. Thus, missile guidance systems are required to be further developed in a more intelligent and autonomous way to deal with complicated environments. In [...] Read more.
In the contemporary battlefield where complexity has increased, the enhancement of the role and ability of missiles has become crucial. Thus, missile guidance systems are required to be further developed in a more intelligent and autonomous way to deal with complicated environments. In this paper, we propose novel missile guidance laws using reinforcement learning, which can autonomously avoid obstacles and terrains in complicated environments with limited prior information and even without the need of off-line trajectory or waypoint generation. The proposed guidance laws are focused on two mission scenarios: the first is with planar obstacles, which is used to cope with maritime operations, and the second is with complex terrain, which is used to cope with land operations. We present the detailed design processes for both scenarios, including a neural network architecture, reward function selection, and training method. Simulation results are provided to show the feasibility and effectiveness of the proposed guidance laws and some important aspects are discussed in terms of their advantages and limitations. Full article
(This article belongs to the Special Issue Applications of Machine Learning and Optimal Control to Aerospace Systems)
►▼ Show Figures

Figure 1

Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:
 
Displaying articles 1-5
Back to TopTop