Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Global Optimal Control”

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 127 records · Page 7

Adaptive analog video signal conditioner

This paper presents an adaptive analog video signal conditioner. This system supplements conventional irises and automatic gain controls to condition video signals for digitizing with a uniform analog-to-digital converter. Employing two degree of freedom, this circuit adjusts an analog gain and analog offset to increase the average information content of digitized images as defined by an entropy measure. The globally near-optimized implementation tracks smoothly to adapt to changing scene conditions.

Jeong, Cheol-Ho↗

Using SpF to Achieve Petascale for Legacy Pseudospectral Applications

Pseudospectral (PS) methods possess a number of characteristics (e.g., efficiency, accuracy, natural boundary conditions) that are extremely desirable for dynamo models. Unfortunately, dynamo models based upon PS methods face a number of daunting challenges, which include exposing additional parallelism, leveraging hardware accelerators, exploiting hybrid parallelism, and improving the scalability of global memory transposes. Although these issues are a concern for most models, solutions for PS methods tend to require far more pervasive changes to underlying data and control structures. Further, improvements in performance in one model are difficult to transfer to other models, resulting in significant duplication of effort across the research community. We have developed an extensible software framework for pseudospectral methods called SpF that is intended to enable extreme scalability and optimal performance. Highlevel abstractions provided by SpF unburden applications of the responsibility of managing domain decomposition and load balance while reducing the changes in code required to adapt to new computing architectures. The key design concept in SpF is that each phase of the numerical calculation is partitioned into disjoint numerical kernels that can be performed entirely inprocessor. The granularity of domain decomposition provided by SpF is only constrained by the datalocality requirements of these kernels. SpF builds on top of optimized vendor libraries for common numerical operations such as transforms, matrix solvers, etc., but can also be configured to use open source alternatives for portability. SpF includes several alternative schemes for global data redistribution and is expected to serve as an ideal testbed for further research into optimal approaches for different network architectures. In this presentation, we will describe our experience in porting legacy pseudospectral models, MoSST and DYNAMO, to use SpF as well as present preliminary performance results provided by the improved scalability.

DYNAMO↗

SpF: Enabling Petascale Performance for Pseudospectral Dynamo Models

Pseudospectral (PS) methods possess a number of characteristics (e.g., efficiency, accuracy, natural boundary conditions) that are extremely desirable for dynamo models. Unfortunately, dynamo models based upon PS methods face a number of daunting challenges, which include exposing additional parallelism, leveraging hardware accelerators, exploiting hybrid parallelism, and improving the scalability of global memory transposes. Although these issues are a concern for most models, solutions for PS methods tend to require far more pervasive changes to underlying data and control structures. Further, improvements in performance in one model are difficult to transfer to other models, resulting in significant duplication of effort across the research community.We have developed an extensible software framework for pseudospectral methods called SpF that is intended to enable extreme scalability and optimal performance. High-level abstractions provided by SpF unburden applications of the responsibility of managing domain decomposition and load balance while reducing the changes in code required to adapt to new computing architectures. The key design concept in SpF is that each phase of the numerical calculation is partitioned into disjoint numerical kernels that can be performed entirely in-processor. The granularity of domain-decomposition provided by SpF is only constrained by the data-locality requirements of these kernels. SpF builds on top of optimized vendor libraries for common numerical operations such as transforms, matrix solvers, etc., but can also be configured to use open source alternatives for portability. SpF includes several alternative schemes for global data redistribution and is expected to serve as an ideal testbed for further research into optimal approaches for different network architectures.In this presentation, we will describe the basic architecture of SpF as well as preliminary performance data and experience with adapting legacy dynamo codes. We will conclude with a discussion of planned extensions to SpF that will provide pseudospectral applications with additional flexibility with regard to time integration, linear solvers, and discretization in the radial direction.

Pseudospectral (PS)↗

Application of singular perturbation techniques /SPT/ and continuation methods for on-line aircraft trajectory optimization

The singular perturbation theory (SPT) approximation of optimal feedback control laws is presented and methods for on-line application of these approximations are discussed. It is demonstrated that SPT control laws break down when the current state is near the terminal target state. The use of continuation methods to improve the accuracy of the SPT approximation and to obtain global solutions of two-point boundary value problems is also discussed. As an illustration, consideration is given to the minimum-time control of a supersonic aircraft for a three-dimensional intercept problem.

Washburn, R. B.↗

Optimizing Aerobot Exploration of Venus

Venus Flyer Robot (VFR) is an aerobot; an autonomous balloon probe designed for remote exploration of Earth's sister planet in 2003. VFR's simple navigation and control system permits travel to virtually any location on Venus, but it can survive for only a limited duration in the harsh Venusian environment. To help address this limitation, we develop: (1) a global circulation model that captures the most important characteristics of the Venusian atmosphere; (2) a simple aerobot model that captures thermal restrictions faced by VFR at Venus; and (3) one exact and two heuristic algorithms that, using abstractions (1) and (2), construct routes making the best use of VFR's limited lifetime. We demonstrate this modeling by planning several small example missions and a prototypical mission that explores numerous interesting sites recently documented in the plane tary geology literature.

SPACE EXPLORATION↗

Hybrid Power Management Program Continued

Hybrid Power Management (HPM) is the innovative integration of diverse, state-of-the-art power devices in an optimal configuration for space and terrestrial applications. The appropriate application and control of the various power devices significantly improves overall system performance and efficiency. The advanced power devices include ultracapacitors and photovoltaics. HPM has extremely wide potential with applications including power-generation, transportation, biotechnology, and space power systems. It may significantly alleviate global energy concerns, improve the environment, and stimulate the economy.

Eichenberg, Dennis J.↗

Autonomous Spacecraft Communication Interface for Load Planning

Ground-based controllers can remain in continuous communication with spacecraft in low Earth orbit (LEO) with near-instantaneous communication speeds. This permits near real-time control of all of the core spacecraft systems by ground personnel. However, as NASA missions move beyond LEO, light-time communication delay issues, such as time lag and low bandwidth, will prohibit this type of operation. As missions become more distant, autonomous control of manned spacecraft will be required. The focus of this paper is the power subsystem. For present missions, controllers on the ground develop a complete schedule of power usage for all spacecraft components. This paper presents work currently underway at NASA to develop an architecture for an autonomous spacecraft, and focuses on the development of communication between the Mission Manager and the Autonomous Power Controller. These two systems must work together in order to plan future load use and respond to unanticipated plan deviations. Using a nominal spacecraft architecture and prototype versions of these two key components, a number of simulations are run under a variety of operational conditions, enabling development of content and format of the messages necessary to achieve the desired goals. The goals include negotiation of a load schedule that meets the global requirements (contained in the Mission Manager) and local power system requirements (contained in the Autonomous Power Controller), and communication of off-plan disturbances that arise while executing a negotiated plan. The message content is developed in two steps: first, a set of rapid-prototyping "paper" simulations are preformed; then the resultant optimized messages are codified for computer communication for use in automated testing.

Controls↗

On l(1): Optimal decentralized performance

In this paper, the Manousiouthakis parametrization of all decentralized stabilizing controllers is employed in mathematically formulating the l(sup 1) optimal decentralized controller synthesis problem. The resulting optimization problem is infinite dimensional and therefore not directly amenable to computations. It is shown that finite dimensional optimization problems that have value arbitrarily close to the infinite dimensional one can be constructed. Based on this result, an algorithm that solves the l(sup 1) decentralized performance problems is presented. A global optimization approach to the solution of the infinite dimensional approximating problems is also discussed.

Sourlas, Dennis↗

Hybrid Power Management Program Evaluated Ultracapacitors for the Next Generation Launch Transportation Project

The NASA Glenn Research Center initiated baseline testing of ultracapacitors to obtain empirical data in determining the feasibility of using ultracapacitors for the Next Generation Launch Transportation (NGLT) Project. There are large transient loads associated with NGLT that require a very large primary energy source or an energy storage system. The primary power source used for this test was a proton-exchange-membrane (PEM) fuel cell. The energy storage system can consist of batteries, flywheels, or ultracapacitors. Ultracapacitors were used for these tests. NASA Glenn has a wealth of experience in ultracapacitor technology through the Hybrid Power Management (HPM) Program, which the Avionics, Power and Communications Branch of Glenn s Engineering Development Division initiated for the Technology Transfer and Partnership Office. HPM is the innovative integration of diverse, state-ofthe- art power devices in optimal configurations for space and terrestrial applications. The appropriate application and control of the various advanced power devices (such as ultracapacitors and fuel cells) significantly improves overall system performance and efficiency. HPM has extremely wide potential. Applications include power generation, transportation systems, biotechnology systems, and space power systems. HPM has the potential to significantly alleviate global energy concerns, improve the environment, and stimulate the economy.

Eichenberg, Dennis J.↗

System Identification for eVTOL Aircraft Using Simulated Flight Data

This paper describes a system identification method for electric vertical takeoff and landing (eVTOL) aircraft. The approach merges fixed-wing and rotary-wing modeling techniques with new strategies to develop a modeling method for eVTOL vehicles using flight test data. The eVTOL aircraft system identification approach is demonstrated through application to the NASA LA-8 tandem tilt-wing, distributed electric propulsion aircraft using a high-fidelity flight dynamics simulation. Orthogonal phase-optimized multisine inputs are applied to each control surface and propulsor at numerous flight conditions throughout the flight envelope to collect informative flight data. An aero-propulsive model is identified at each flight condition using the equation-error method in the frequency domain. The local model parameters are then blended to create a global model across the nominal flight envelope. Parameter estimation results are shown to provide a good fit to modeling data and have good prediction capability. The methodology is developed with a discussion of unique eVTOL vehicle aerodynamic characteristics and practical strategies intended to inform future flight-based system identification efforts for eVTOL aircraft.

system identification↗

Partially Decentralized Control Architectures for Satellite Formations

In a partially decentralized control architecture, more than one but less than all nodes have supervisory capability. This paper describes an approach to choosing the number of supervisors in such au architecture, based on a reliability vs. cost trade. It also considers the implications of these results for the design of navigation systems for satellite formations that could be controlled with a partially decentralized architecture. Using an assumed cost model, analytic and simulation-based results indicate that it may be cheaper to achieve a given overall system reliability with a partially decentralized architecture containing only a few supervisors, than with either fully decentralized or purely centralized architectures. Nominally, the subset of supervisors may act as centralized estimation and control nodes for corresponding subsets of the remaining subordinate nodes, and act as decentralized estimation and control peers with respect to each other. However, in the context of partially decentralized satellite formation control, the absolute positions and velocities of each spacecraft are unique, so that correlations which make estimates using only local information suboptimal only occur through common biases and process noise. Covariance and monte-carlo analysis of a simplified system show that this lack of correlation may allow simplification of the local estimators while preserving the global optimality of the maneuvers commanded by the supervisors.

Carpenter, J. Russell↗

Hybrid Power Management (HPM)

The NASA Glenn Research Center s Avionics, Power and Communications Branch of the Engineering and Systems Division initiated the Hybrid Power Management (HPM) Program for the GRC Technology Transfer and Partnership Office. HPM is the innovative integration of diverse, state-of-the-art power devices in an optimal configuration for space and terrestrial applications. The appropriate application and control of the various power devices significantly improves overall system performance and efficiency. The advanced power devices include ultracapacitors and fuel cells. HPM has extremely wide potential. Applications include power generation, transportation systems, biotechnology systems, and space power systems. HPM has the potential to significantly alleviate global energy concerns, improve the environment, and stimulate the economy. One of the unique power devices being utilized by HPM for energy storage is the ultracapacitor. An ultracapacitor is an electrochemical energy storage device, which has extremely high volumetric capacitance energy due to high surface area electrodes, and very small electrode separation. Ultracapacitors are a reliable, long life, maintenance free, energy storage system. This flexible operating system can be applied to all power systems to significantly improve system efficiency, reliability, and performance. There are many existing and conceptual applications of HPM.

Eichenberg, Dennis J.↗

SPOT Program

A Spacecraft Position Optimal Tracking (SPOT) program was developed to process Global Positioning System (GPS) data, sent via telemetry from a spacecraft, to generate accurate navigation estimates of the vehicle position and velocity (state vector) using a Kalman filter. This program uses the GPS onboard receiver measurements to sequentially calculate the vehicle state vectors and provide this information to ground flight controllers. It is the first real-time ground-based shuttle navigation application using onboard sensors. The program is compact, portable, self-contained, and can run on a variety of UNIX or Linux computers. The program has a modular objec-toriented design that supports application-specific plugins such as data corruption remediation pre-processing and remote graphics display. The Kalman filter is extensible to additional sensor types or force models. The Kalman filter design is also strong against data dropouts because it uses physical models from state and covariance propagation in the absence of data. The design of this program separates the functionalities of SPOT into six different executable processes. This allows for the individual processes to be connected in an a la carte manner, making the feature set and executable complexity of SPOT adaptable to the needs of the user. Also, these processes need not be executed on the same workstation. This allows for communications between SPOT processes executing on the same Local Area Network (LAN). Thus, SPOT can be executed in a distributed sense with the capability for a team of flight controllers to efficiently share the same trajectory information currently being computed by the program. SPOT is used in the Mission Control Center (MCC) for Space Shuttle Program (SSP) and International Space Station Program (ISSP) operations, and can also be used as a post -flight analysis tool. It is primarily used for situational awareness, and for contingency situations.

Smith, Jason T.↗

NASA's Rodent Research Project on ISS: Validation of a New Platform for Conducting Biomedical and Basic Research into the Consequences of Long Duration Habitation in Space

Rodent research has played a key role in advancing biomedical discoveries both on Earth and in space. The National Research Counsel’s Decadal survey(1) emphasized the importance of expanding NASAs life sciences research to perform long duration, rodent experiments on the International Space Station (ISS). To accomplish this objective, flight hardware, operations, and science capabilities were developed at NASA ARC to support both commercial and government-sponsored rodent research.Rodent Research-1 (RR-1) was the first mission in which animals were delivered and maintained in the ISS for a long duration mission in modified Animal Enclosure Module hardware. Both RR validation and commercial science objectives were pursued on the RR-1 mission. Adult female mice (20 total Flight, FLT) were launched Sept 21, 2014 in RR hardware within a Dragon Capsule (SpaceX4), then after 4 days in transit, were transferred for habitation on the ISS for 17 days (commercial) or 33 days (validation), when animals were euthanized and select tissues recovered on orbit. Various controls groups consisted of: 1) Basal mice from the same cohorts as FLT mice, but tissues were recovered at time of launch, 2) Vivarium (VIV) were housed in standard cages 3) Ground Controls (GC) were housed in flight hardware within an environmental chamber at Kennedy Space Center. The health and behavior of all mice on the ISS were monitored by video feed on a daily basis. Mice were euthanized by injection of Euthasol, then either fast frozen intact or dissected to preserve livers (fast frozen) and spleens (RNAlater). Samples were stored at ≤ -80˚C until their return to Earth for later analyses.Hardware performed nominally throughout the mission and the planned in-flight science operations were completed successfully. FLT mice appeared generally more physically active on orbit than respective GC groups. After 33 days on the ISS, mean body weights of FLT mice did not differ from GC, with both groups showing a 6% rise compared to time of launch, while VIV mice showed an 8% rise over the same period. Importantly, there were no significant differences in body weights between groups at the end of 33 days on the ISS, providing an indication that the RR hardware supported the health of the mice both on Earth and in space. Based on the preliminary data obtained from the livers and spleens of mice after 17 days on the ISS, purified RNA was of high quality (RIN values of spleen: FLT=9.48 +0.40, GC=9.28 +0.44, n=5/group); therefore, RNA quality from samples retrieved on orbit was acceptable for even the most demanding transcriptomic analyses. In addition, liver enzyme activity levels (units/mg protein) of FLT mice (after 17d on ISS) and all control mice were similar in magnitude to samples that were optimally prepared by freezing in liquid nitrogen in the laboratory (enzymes analyzed included catalase, glutathione reductase and glyceraldehyde-3-phosphate dehydrogenase). Validation analyses still in progress include behavior and tissue biochemistries, as well as optimization of science return by post-flight recovery of tissues for biospecimen sharing and global expression analyses.Together, these preliminary findings demonstrate new capability for supporting long duration rodent research on the ISS to achieve both basic science and biomedical objectives.

Globus, R. K.↗

Minimum fuel horizontal flightpaths in the terminal area

The problem of minimum fuel airplane trajectories from arbitrary initial states to be fixed final state is considered. There are four state variables (two position coordinates, heading, and constrained velocity) and two constrained controls (thrust and bank angle). The fuel optimality of circular and straight line flightpaths is examined. Representative extremals (trajectories satisfying the necessary conditions of the minimum principle) of various types are computed and used to evaluate trajectories generated by an on line algorithm. Attention is paid to the existence of Darboux points (beyond which an extremal ceases to be globally optimal). One fuel flow rate model includes a term quadratic in thrust; hence, the optimal thrust is continuous and nonsingular. The other fuel flow rate model is linear in thrust, and consequently the optimal thrust is discontinuous and singular.

Kreindler, E.↗

Combined control-structural optimization

An approach to combined control-structural optimization aimed at enhancing early design trade-offs is outlined and illustrated by numerical examples. The approach employs a homotopic strategy and is capable of generating families of designs that can be used in early trade studies. Analytical results are obtained for classes of structure/control objectives with LQG and LQR costs. For these, it is demonstrated that global optima can be computed for small values of the homotopy parameter.

Milman, M.↗

Characterization and control of self-motions in redundant manipulators

The presence of redundant degrees of freedom in a manipulator structure leads to a physical phenomenon known as a self-motion, which is a continuous motion of the manipulator joints that leaves the end-effector motionless. In the first part of the paper, a global manifold mapping reformulation of manipulator kinematics is reviewed, and the inverse kinematic solution for redundant manipulators is developed in terms of self-motion manifolds. Global characterizations of the self-motion manifolds in terms of their number, geometry, homotopy class, and null space are reviewed using examples. Much previous work in redundant manipulator control has been concerned with the redundancy resolution problem, in which methods are developed to determine, or resolve, the motion of the joints in order to achieve end-effector trajectory control while optimizing additional objective functions. Redundancy resolution problems can be equivalently posed as the control of self-motions. Alternatives for redundancy resolution are briefly discussed.

Burdick, J.↗

Neutron Characterization for Additive Manufacturing

Oak Ridge National Laboratory (ORNL) is leveraging decades of experience in neutron characterization of advanced materials together with resources such as the Spallation Neutron Source (SNS) and the High Flux Isotope Reactor (HFIR) shown in Fig. 1 to solve challenging problems in additive manufacturing (AM). Additive manufacturing, or three-dimensional (3-D) printing, is a rapidly maturing technology wherein components are built by selectively adding feedstock material at locations specified by a computer model. The majority of these technologies use thermally driven phase change mechanisms to convert the feedstock into functioning material. As the molten material cools and solidifies, the component is subjected to significant thermal gradients, generating significant internal stresses throughout the part (Fig. 2). As layers are added, inherent residual stresses cause warping and distortions that lead to geometrical differences between the final part and the original computer generated design. This effect also limits geometries that can be fabricated using AM, such as thin-walled, high-aspect- ratio, and overhanging structures. Distortion may be minimized by intelligent toolpath planning or strategic placement of support structures, but these approaches are not well understood and often "Edisonian" in nature. Residual stresses can also impact component performance during operation. For example, in a thermally cycled environment such as a high-pressure turbine engine, residual stresses can cause components to distort unpredictably. Different thermal treatments on as-fabricated AM components have been used to minimize residual stress, but components still retain a nonhomogeneous stress state and/or demonstrate a relaxation-derived geometric distortion. Industry, federal laboratory, and university collaboration is needed to address these challenges and enable the U.S. to compete in the global market. Work is currently being conducted on AM technologies at the ORNL Manufacturing Demonstration Facility (MDF) sponsored by the DOE's Advanced Manufacturing Office. The MDF is focusing on R&D of both metal and polymer AM pertaining to in-situ process monitoring and closed-loop controls; implementation of advanced materials in AM technologies; and demonstration, characterization, and optimization of next-generation technologies. ORNL is working directly with industry partners to leverage world-leading facilities in fields such as high performance computing, advanced materials characterization, and neutron sciences to solve fundamental challenges in advanced manufacturing. Specifically, MDF is leveraging two of the world's most advanced neutron facilities, the HFIR and SNS, to characterize additive manufactured components.

Watkins, Thomas↗