Search NASA⌕ Search

SEARCH · Search NASA

Results for “multi objective”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 19 records
All sourcesNASA NTRSDOE OSTI
NASA NTRS2010

Multi-Objective Multi-User Scheduling for Space Science Missions

We have developed an architecture called MUSE (Multi-User Scheduling Environment) to enable the integration of multi-objective evolutionary algorithms with existing domain planning and scheduling tools. Our approach is intended to make it possible to re-use existing software, while obtaining the advantages of multi-objective optimization algorithms. This approach enables multiple participants to actively engage in the optimization process, each representing one or more objectives in the optimization problem. As initial applications, we apply our approach to scheduling the James Webb Space Telescope, where three objectives are modeled: minimizing wasted time, minimizing the number of observations that miss their last planning opportunity in a year, and minimizing the (vector) build up of angular momentum that would necessitate the use of mission critical propellant to dump the momentum. As a second application area, we model aspects of the Cassini science planning process, including the trade-off between collecting data (subject to onboard recorder capacity) and transmitting saved data to Earth. A third mission application is that of scheduling the Cluster 4-spacecraft constellation plasma experiment. In this paper we describe our overall architecture and our adaptations for these different application domains. We also describe our plans for applying this approach to other science mission planning and scheduling problems in the future.

science planning↗
NASA NTRS2011

Developer Tools for Evaluating Multi-Objective Algorithms

Multi-objective algorithms for scheduling offer many advantages over the more conventional single objective approach. By keeping user objectives separate instead of combined, more information is available to the end user to make trade-offs between competing objectives. Unlike single objective algorithms, which produce a single solution, multi-objective algorithms produce a set of solutions, called a Pareto surface, where no solution is strictly dominated by another solution for all objectives. From the end-user perspective a Pareto-surface provides a tool for reasoning about trade-offs between competing objectives. From the perspective of a software developer multi-objective algorithms provide an additional challenge. How can you tell if one multi-objective algorithm is better than another? This paper presents formal and visual tools for evaluating multi-objective algorithms and shows how the developer process of selecting an algorithm parallels the end-user process of selecting a solution for execution out of the Pareto-Surface.

multi-objective optimization↗
NASA NTRSDate not supplied

Exploring the Low-Thrust Transfer Design Space in an Ephemeris Model via Multi-Objective Reinforcement Learning

Multi-Reward Proximal Policy Optimization (MRPPO) is a multi-objective reinforcement learning algorithm used to train multiple policies to uncover solutions within a multi-objective solution space. MRPPO is used in this paper to train policies to construct low-thrust transfers for a SmallSat from the vicinity of !2 to an !5 short period orbit in the Sun-Earth-Moon system. First, the policies are trained in this scenario in the circular restricted three-body problem. This information is used to initialize the policies before training in a higher-fidelity ephemeris model; a process known as transfer learning. The recovered segments of the solution space will be compared to fundamental dynamical structures to both examine the results of MRPPO in this complex design scenario and explore the effectiveness of transfer learning.

Christopher J Sullivan↗
NASA NTRSDate not supplied

Exploring the Low-Thrust Transfer Design Space in an Ephemeris Model via Multi-Objective Reinforcement Learning

Multi-Reward Proximal Policy Optimization (MRPPO) is a multi-objective reinforcement learning algorithm used to train multiple policies to uncover solutions within a multi-objective solution space. MRPPO is used in this paper to train policies to construct low-thrust transfers for a SmallSat from the vicinity of L2 to an L5 short period orbit in the Sun-Earth-Moon system. First, the policies are trained in this scenario in the circular restricted three-body problem. This information is used to initialize the policies before training in a higher-fidelity ephemeris model; a process known as transfer learning. The recovered segments of the solution space will be compared to fundamental dynamical structures to both examine the results of MRPPO in this complex design scenario and explore the effectiveness of transfer learning.

Christopher J. Sullivan↗
NASA NTRS2022

Exploring the Low-Thrust Transfer Design Space in an Ephemeris Model via Multi-Objective Reinforcement Learning

Multi-Reward Proximal Policy Optimization (MRPPO) is a multi-objective reinforcement learning algorithm used to train multiple policies to uncover solutions within a multi-objective solution space. MRPPO is used in this paper to train policies to construct low-thrust transfers for a SmallSat from the vicinity of 𝐿2 to an 𝐿5 short period orbit in the Sun-Earth-Moon system. First, the policies are trained in this scenario in the circular restricted three-body problem. This information is used to initialize the policies before training in a higher-fidelity ephemeris model; a process known as transfer learning. The recovered segments of the solution space will be compared to fundamental dynamical structures to both examine the results of MRPPO in this complex design scenario and explore the effectiveness of transfer learning.

Mashiku, Alinda K.↗
NASA NTRS2017

Global, Multi-Objective Trajectory Optimization With Parametric Spreading

Mission design problems are often characterized by multiple, competing trajectory optimization objectives. Recent multi-objective trajectory optimization formulations enable generation of globally-optimal, Pareto solutions via a multi-objective genetic algorithm. A byproduct of these formulations is that clustering in design space can occur in evolving the population towards the Pareto front. This clustering can be a drawback, however, if parametric evaluations of design variables are desired. This effort addresses clustering by incorporating operators that encourage a uniform spread over specified design variables while maintaining Pareto front representation. The algorithm is demonstrated on a Neptune orbiter mission, and enhanced multidimensional visualization strategies are presented.

trajectory design↗
NASA NTRS2021

Exploring Transfers Between Earth-Moon Halo Orbits via Multi-Objective Reinforcement Learning

Multi-Reward Proximal Policy Optimization, a multi-objective deep reinforcement learning algorithm, is used to examine the design space of low-thrust trajectories for a SmallSat transferring between two libration point orbits in the Earth- Moon system. Using Multi-Reward Proximal Policy Optimiza- tion, multiple policies are simultaneously and efficiently trained on three distinct trajectory design scenarios. Each policy is trained to create a unique control scheme based on the trajectory design scenario and assigned reward function: a unique combination of weights scaling competing objectives that guide the spacecraft to the target mission orbit, incentivize faster flight times, and penalize propellant mass usage. Then, the policies are evaluated on the same set of perturbed initial conditions in each scenario to generate the propellant mass usages, flight times, and state discontinuities from a reference trajectory for each control scheme. This solution space of low-thrust trajectories for a SmallSat is used to examine the multi-objective trade space for the trajectory design scenario. By autonomously constructing the solution space, insights into the required propellant mass, flight time, and transfer geometry are rapidly achieved.

Mashiku, Alinda K.↗
NASA NTRSDate not supplied

Exploring Transfers Between Earth-Moon Halo Orbits via Multi-Objective Reinforcement Learning

Multi-Reward Proximal Policy Optimization, a multi-objective deep reinforcement learning algorithm, is used to examine the design space of low-thrust trajectories for a SmallSat transferring between two libration point orbits in the Earth-Moon system. Using Multi-Reward Proximal Policy Optimization, multiple policies are simultaneously and efficiently trained on three distinct trajectory design scenarios. Each policy is trained to create a unique control scheme based on the trajectory design scenario and assigned reward function: a unique combination of weights scaling competing objectives that guide the spacecraft to the target mission orbit, incentivize faster flight times, and penalize propellant mass usage. Then, the policies are evaluated on the same set of perturbed initial conditions in each scenario to generate the propellant mass usages, flight times, and state discontinuities from a reference trajectory for each control scheme. This solution space of low-thrust trajectories for a SmallSat is used to examine the multi-objective trade space for the trajectory design scenario. By autonomously constructing the solution space, insights into the required propellant mass, flight time, and transfer geometry are rapidly achieved.

Christopher J Sullivan↗
NASA NTRS2002

Subsystem Imaging Performance and Modeling of the Infrared Multi-Object Spectrograph

The Infrared Multi-Object Spectrograph (IRMOS) is a facility instrument for the Kitt Peak National Observatory Mayall Telescope (3.8 meter). IRMOS is a near-IR (0.8 - 2.5 micron) spectrograph with low to mid resolution (R=lambda/delta, lambda = 300 - 3800). The IRMOS spectrograph produces simultaneous spectra of - 100 objects in its 2.8 x 2.0 arc-min field of view using a commercial MEMS multi-mirror array device (MMA). The IRMOS optical design consists of two imaging systems, or "stages." The focal reducer, stage one, images the focal plane of the telescope onto the MMA. The spectrograph, stage two, images the MMA onto the detector. We describe the breadboard alignment method and imaging and scattered light performance for both the focal reducer and spectrograph. This testing provides verification of the optomechanical alignment method, and a measurement of the contribution of scattered light in the system due to mirror small scale surface error. After the stage I and 2 optics are integrated with the instrument, our test results will make it possible to distinguish between scattered light from the mirrors and the MMA. Image testing will be done at four wavelengths in the visible and near-IR. A mercury-argon pencil lamp will provide spectral lines at 546.1 and 1012 nm, and a blackbody radiation source lines at 1600 and 2200 nm. A CCD camera will be used as a detector for the visible wavelengths, and an IR photodiode will be used for the IR wavelengths. We compare our data with a theoretical analysis using a commercial software package. Mirror surface error is modeled by treating each surface as a superposition of various gratings (e.g., diamond turning tool marks, features due to the impurities of Al 6061, and periodic mid-frequency errors due to drift during machining).

Connelly, Joseph A.↗
NASA NTRS2003

Integration, Testing and Performance of the Infrared Multi-Object Spectrometer

The Infrared Multi-Object Spectrometer (IRMOS) is a principle investigator-class instrument for the Kitt Peak National Observatory 2.1 m and Mayall 3.8 m telescopes. IRMOS is a near-IR (0.8 - 2.5 micron) spectrometer with low-to mid-resolving power (R = lambda/delta lambda = 300 - 3000). On the 3.8 m telescope, IRMOS produces simultaneous spectra of approximately 100 objects in its approximately 3 x 2 arcmin field of view using a commercial micro electro-mechanical systems (MEMS) digital micro-mirror device (DMD) from Texas Instruments. The multi-mirror array DMD operates as a real-time programmable slit mask. The all-reflective optical design consists of two imaging subsystems. The focal reducer images the focal plane of the telescope onto the DMD field stop, and the spectrograph images the DMD onto a large-format detector. The instrument operates at approximately 80 K, cooled by a single electro-mechanical cryocooler. The bench and all components are made from aluminum 6061-T651. There are three cryogenic mechanisms. We describe laboratory integration and test of IRMOS before shipment to Kitt Peak. We give an overview of the optical alignment technique and integration of optical, mechanical, electrical and cryogenic subsystems. We compare optical test results to model predictions of point spread function size and morphology, contrast, and stray light. We discuss some lessons learned and conclude with a prediction for performance on the telescope.

Ohl, Raymond G.↗
NASA NTRS2003

Ambient and Cryogenic Alignment Verification and Performance of the Infrared Multi-Object Spectrometer

The Infrared Multi-Object Spectrometer (IRMOS) is a facility instrument for the Kitt Peak National Observatory 4 and 2.1 meter telescopes. IRMOS is a near-IR (0.8 - 2.5 micron) spectrometer with low- to mid-resolving power (R = 300 - 3000). IRMOS produces simultaneous spectra of approximately 100 objects in its 2.8 x 2.0 arc-min field of view using a commercial Micro Electro-Mechanical Systems (MEMS) Digital Micro-mirror Device (DMD) from Texas Instruments. The IRMOS optical design consists of two imaging subsystems. The focal reducer images the focal plane of the telescope onto the DMD field stop, and the spectrograph images the DMD onto the detector. We describe ambient breadboard subsystem alignment and imaging performance of each stage independently, and the ambient and cryogenic imaging performance of the fully assembled instrument. Interferometric measurements of subsystem wavefront error serve to venfy alignment, and are accomplished using a commercial, modified Twyman-Green laser unequal path interferometer. Image testing provides further verification of the optomechanical alignment method and a measurement of near-angle scattered light due to mirror small-scale surface error. Image testing is performed at multiple field points. A mercury-argon pencil lamp provides spectral lines at 546.1 nm and 1550 nm, and a CCD camera and IR camera are used as detectors. We use commercial optical modeling software to predict the point-spread function and its effect on instrument slit transmission and resolution. Our breadboard test results validate this prediction. We conclude with an instrument performance prediction for first light.

Connelly, Joseph A.↗
NASA NTRS2004

Alignment and Performance of the Infrared Multi-Object Spectrometer

The Infrared Multi-Object Spectrometer (IRMOS) is a principle investigator class instrument for the Kitt Peak National Observatory 4 and 2.1 meter telescopes. IRMOS is a near-IR (0.8 - 2.5 micron) spectrometer with low-to mid-resolving power (R = 300 - 3000). IRMOS produces simultaneous spectra of approximately 100 objects in its 2.8 x 2.0 arc-min field of view (4 m telescope) using a commercial Micro Electro-Mechanical Systems (MEMS) micro-mirror array (MMA) from Texas Instruments. The IRMOS optical design consists of two imaging subsystems. The focal reducer images the focal plane of the telescope onto the MMA field stop, and the spectrograph images the MMA onto the detector. We describe ambient breadboard subsystem alignment and imaging performance of each stage independently, and ambient imaging performance of the fully assembled instrument. Interferometric measurements of subsystem wavefront error serve as a qualitative alignment guide, and are accomplished using a commercial, modified Twyman-Green laser unequal path interferometer. Image testing provides verification of the optomechanical alignment method and a measurement of near-angle scattered light due to mirror small-scale surface error. Image testing is performed at multiple field points. A mercury-argon pencil lamp provides a spectral line at 546.1 nanometers, a blackbody source provides a line at 1550 nanometers, and a CCD camera and IR camera are used as detectors. We use commercial optical modeling software to predict the point-spread function and its effect on instrument slit transmission and resolution. Our breadboard and instrument level test results validate this prediction. We conclude with an instrument performance prediction for cryogenic operation and first light in late 2003.

Connelly, Joseph A.↗
NASA NTRS2019

Time-varying Weights in Multi-Objective Optimal Control for Flexible Wing Aircraft

A multi-objective optimal control technique is modified to accommodate changing cost function weights and is used to control a flexible wing aircraft model. Variation of the weights is used to adjust the relative importance of each objective according to either a prescribed function of time or of the state. Several techniques for obtaining a practical approximation to the optimal control solution are presented, and stability of a specific weight structure with the optimal controller is demonstrated. Functionality of the multi-objective control design with weight variation is demonstrated in simulation of a flexible wing transport aircraft and is shown to improve performance over the fixed weight version both at a constant flight condition and across changing flight conditions.

Hashemi, Kelley E.↗
NASA NTRS2022

A Reinforcement Learning Hyper-Heuristic in Multi-Objective Optimization with Application to Structural Damage Identification

Multi-objective optimization allows satisfying multiple decision criteria concurrently, and generally yields multiple solutions. It has the potential to be applied to structural damage identification applications which are oftentimes under-determined. How to achieve high-quality solutions in terms of accuracy, diversity, and completeness is a challenging research subject. The solution techniques and parametric selections are believed to be problem specific. In this research, we formulate a reinforcement learning hyper-heuristic scheme to work coherently with the single-point search algorithm MOSA/R (Multi-Objective Simulated Annealing Algorithm based on Re-seed). The four low-level heuristics proposed can meet various optimization requirements adaptively and autonomously using the domination amount, crowding distance, and hypervolume calculations. The new approach exhibits improved and more robust performance than AMOSA, NSGA-II, and MOEA/D when applied to benchmark test cases. It is then applied to an active damage interrogation scheme for structural damage identification where solution diversity/completeness and accuracy are critically important. Results show that this approach can successfully include the true damage scenario in the solution set identified. The outcome of this research can potentially be extended to a variety of applications.

Pei Cao↗
NASA NTRS2019

Multi-Objective Optimal Control of the 6-DoF Aeroservoelastic Common Research Model with Aspect Ratio 13.5 Wing

A new 6-DoF aeroservoelastic (ASE) Common Research Model (CRM) provided by The Boeing Company with aspect ratio 13.5 and 17 control surfaces per wing is utilized to demonstrate combined tracking and optimal multi-objective control. The multi-objective controller is derived on the closed loop tracking controller, and utilizes state and gust estimates provided by an extended state observer. Various methods of model reduction useful for control and estimation are presented. A computationally efficient MATLAB/Simulink simulation is presented which includes actuator dynamics, rate and deflection saturation limits, and gust disturbance inputs. The platform is used to demonstrate excellent 6-DoF tracking control performance coupled with the multi-objective controller, which is shown to effectively reduce structural mode movement, wing root bending moment, and drag. State and gust estimation is also shown to perform well, even when derived and/or implemented with significantly fewer states than the original full-sized model.

Drew, Michael C.↗
NASA NTRS2017

Multi-Objective Flight Control for Drag Minimization and Load Alleviation of High-Aspect Ratio Flexible Wing Aircraft

As aircraft wings become much more flexible due to the use of light-weight composites material, adverse aerodynamics at off-design performance can result from changes in wing shapes due to aeroelastic deflections. Increased drag, hence increased fuel burn, is a potential consequence. Without means for aeroelastic compensation, the benefit of weight reduction from the use of light-weight material could be offset by less optimal aerodynamic performance at off-design flight conditions. Performance Adaptive Aeroelastic Wing (PAAW) technology can potentially address these technical challenges for future flexible wing transports. PAAW technology leverages multi-disciplinary solutions to maximize the aerodynamic performance payoff of future adaptive wing design, while addressing simultaneously operational constraints that can prevent the optimal aerodynamic performance from being realized. These operational constraints include reduced flutter margins, increased airframe responses to gust and maneuver loads, pilot handling qualities, and ride qualities. All of these constraints while seeking the optimal aerodynamic performance present themselves as a multi-objective flight control problem. The paper presents a multi-objective flight control approach based on a drag-cognizant optimal control method. A concept of virtual control, which was previously introduced, is implemented to address the pair-wise flap motion constraints imposed by the elastomer material. This method is shown to be able to satisfy the constraints. Real-time drag minimization control is considered to be an important consideration for PAAW technology. Drag minimization control has many technical challenges such as sensing and control. An initial outline of a real-time drag minimization control has already been developed and will be further investigated in the future. A simulation study of a multi-objective flight control for a flight path angle command with aeroelastic mode suppression and drag minimization demonstrates the effectiveness of the proposed solution. In-flight structural loads are also an important consideration. As wing flexibility increases, maneuver load and gust load responses can be significant and therefore can pose safety and flight control concerns. In this paper, we will extend the multi-objective flight control framework to include load alleviation control. The study will focus initially on maneuver load minimization control, and then subsequently will address gust load alleviation control in future work.

Aerolasticity↗
NASA NTRS2020

Multi-Objective Gust Load Alleviation Control Designs for an Aeroelastic Wind Tunnel Demonstration Wing

This paper presents several control and gust disturbance estimation techniques applied to a mathematical model of a physical flexible wing wind tunnel model used in ongoing tests at the University of Washington Aeronautical Laboratory's Kirsten Wind Tunnel. Three methods of gust disturbance estimation are presented, followed by three control methods: LQG, Basic Multi-Objective (BMO), and a novel Multi-Objective Prediction Correction (MOPC) controller. The latter of which augments a multi-objective controller, and attempts to correct for errors in the disturbance estimate. A simplified linear simulation of the three controllers is performed and a simple MIMO stability and robustness assessment is performed. Then, the same controllers are simulated in a higher fidelity Simulink environment that captures sampling, saturation and noise effects. This preliminary analysis indicates that the BMO controller provides the best performance and largest stability margins.

Drew, Michael C.↗
NASA NTRS1990

Multi-objective shape and material optimization of composite structures including damping

A multi-objective optimal design methodology is developed for light-weight, low cost composite structures of improved dynamic performance. The design objectives include minimization of resonance amplitudes (or maximization of modal damping), weight, and material cost. The design vector includes micromechanics, laminate, and structural shape parameters. Performance constraints are imposed on static displacements, dynamic amplitudes, and natural frequencies. The effects of damping on the dynamics of composite structures are incorporated. Preliminary applications on a cantilever composite beam illustrated that only the proposed multi-objective optimization, as opposed to single objective functions, simultaneously improved all objectives. The significance of composite damping in the design of advanced composite structures was also demonstrated, indicating the design methods based on undamped dynamics may fail to improve the dynamic performance near resonances.

Saravanos, D. A.↗