Critical degenerateness in large linear systems
Critical degenerateness in large linear systems represented by coefficient matrices of differential equations
SEARCH · Engineering Papers
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.
Critical degenerateness in large linear systems represented by coefficient matrices of differential equations
Uncertainties of a large space system (LSS) can be deterministic or stochastic in nature. The former may result in, for example, an energy spillover problem by which the interaction between unmodeled modes and controls may cause system instability. The stochastic uncertainties are responsible for mode localization and estimation errors, etc. We will address the effects of uncertainties on structural model formulation, use of available test data to verify and modify analytical models before orbiting, and how the system model can be further improved in the on-orbit environment.
When planning large scale operations in space, environmental impact must be considered in addition to radiation, spacecraft charging, contamination, high power and size. Pollution of the atmosphere and space is caused by rocket effluents and by photoelectrons generated by sunlight falling on satellite surfaces even light pollution may result (the SPS may reflect so much light as to be a nuisance to astronomers). Large (100 Km 2) structures also will absorb the high energy particles that impinge on them. Altogether, these effects may drastically alter the Earth's magnetosphere. It is not clear if these alterations will in any way affect the Earth's surface climate. Large structures will also generate large plasma wakes and waves which may cause interference with communications to the vehicle. A high energy, microwave beam from the SPS will cause ionospheric turbulence, affecting UHF and VHF communications. Although none of these effects may ultimately prove critical, they must be considered in the design of large structures.
Two classes of large-scale systems are identified, which can always be stabilized by decentralized feedback control. For the class of systems composed of interconnected linear subsystems, we can choose local controllers for the subsystems to achieve stability of the overall system. The same linear feedback scheme can be used to stabilize a class of linear systems with bilinear interconnections. In this case, however, the scheme is used to establish a finite region of stability for the overall system. The stabilization algorithm is applied to the design of a control system for the Large-Space Telescope.
Proposed large space structures have many characteristics that make them difficult to analyze and control. They are highly flexible, with components mathematically modeled by partial differential equations or very large systems of ordinary differential equations. They have many resonant frequencies, typically low and closely spaced. Natural damping may be low and/or improperly modeled. Coupled with stringent operational requirements of orientation, shape control, and vibration suppression, and the inability to perform adequate ground testing, these characteristics present an unconventional identification and control design problem to the systems theorist. Some of the research underway within Langley's Spacecraft Control Branch, Guidance and Control Division aimed at developing theory and algorithms to treat large space structures systems identification and control problems is described. The research areas to be considered are computer-aided design algorithms, and systems identification and control of distributed systems.
A large-scale system is considered as a system constituted of subsystems which may be connected or disconnected from each other during operation. A new concept of connective stability is introduced by which a large-scale system is regarded as stable if it remains stable (in the sense of Lyapunov) under structural perturbations produced by the on-off participation of the subsystems. Algebraic conditions are developed that guarantee exponential connective stability of large-scale systems which may be composed of linear and nonlinear time-varying subsystems coupled by linear or nonlinear connections.
Large-scale complex systems are characterized by multiple interacting subsystems and the analysis of multiple disciplines. The design and development of such systems inevitably requires the resolution of multiple conflicting objectives. The size of complex systems, however, prohibits the development of comprehensive system models, and thus these systems must be partitioned into their constituent parts. Because simultaneous solution of individual subsystem models is often not manageable iteration is inevitable and often excessive. In this dissertation these issues are addressed through the development of a method for hierarchical robust preliminary design exploration to facilitate concurrent system and subsystem design exploration, for the concurrent generation of robust system and subsystem specifications for the preliminary design of multi-level, multi-objective, large-scale complex systems. This method is developed through the integration and expansion of current design techniques: Hierarchical partitioning and modeling techniques for partitioning large-scale complex systems into more tractable parts, and allowing integration of subproblems for system synthesis; Statistical experimentation and approximation techniques for increasing both the efficiency and the comprehensiveness of preliminary design exploration; and Noise modeling techniques for implementing robust preliminary design when approximate models are employed. Hierarchical partitioning and modeling techniques including intermediate responses, linking variables, and compatibility constraints are incorporated within a hierarchical compromise decision support problem formulation for synthesizing subproblem solutions for a partitioned system. Experimentation and approximation techniques are employed for concurrent investigations and modeling of partitioned subproblems. A modified composite experiment is introduced for fitting better predictive models across the ranges of the factors, and an approach for constructing partitioned response surfaces is developed to reduce the computational expense of experimentation for fitting models in a large number of factors. Noise modeling techniques are compared and recommendations are offered for the implementation of robust design when approximate models are sought. These techniques, approaches, and recommendations are incorporated within the method developed for hierarchical robust preliminary design exploration. This method as well as the associated approaches are illustrated through their application to the preliminary design of a commercial turbofan turbine propulsion system. The case study is developed in collaboration with Allison Engine Company, Rolls Royce Aerospace, and is based on the Allison AE3007 existing engine designed for midsize commercial, regional business jets. For this case study, the turbofan system-level problem is partitioned into engine cycle design and configuration design and a compressor modules integrated for more detailed subsystem-level design exploration, improving system evaluation. The fan and low pressure turbine subsystems are also modeled, but in less detail. Given the defined partitioning, these subproblems are investigated independently and concurrently, and response surface models are constructed to approximate the responses of each. These response models are then incorporated within a commercial turbofan hierarchical compromise decision support problem formulation. Five design scenarios are investigated, and robust solutions are identified. The method and solutions identified are verified by comparison with the AE3007 engine. The solutions obtained are similar to the AE3007 cycle and configuration, but are better with respect to many of the requirements.
Items of technology and developmental efforts in support of the large space systems technology programs are described. The major areas of interest are large antennas systems, large space platform systems, and activities that support both antennas and platform systems.
A class of large scale linear systems, stabilizable by local feedback control applied around each decoupled subsystem, is established. The control scheme, outlined in the paper, is based upon the decomposition-aggregation method, and involves only operations on the subsystem level. The systems, stabilized by the proposed method, have a tolerance to a wide class of non-linearities, and are reliable with respect to the structural perturbations.
Large scale axisymmetric inlet systems performance capabilities, reviewing theoretical and experimental programs
Large scale axisymmetric inlet systems performance capabilities, reviewing theoretical and experimental programs
Large nuclear space power systems capable of continuously producing over one megawatt of electrical power for a several year period will be needed in the future. This paper presents the results of a study to compare applicable conversion technologies which were deemed to be ready for a time period of 1995 and beyond. A total of six different conversion technologies were studied in detail and compared on the basis of conversion efficiency, radiator area, overall system mass, and feasibility. Three static, modular conversion technologies were considered; these include: AMTEC, thermionic, and thermoelectric conversion. The other three conversion technologies are heat engines which involve dynamic components. The dynamic systems analyzed were Brayton, Rankine, and the free piston Stirling engine. Each of the conversion techniques was also examined for limiting characteristics and an attempt was made to identify common research needs and enabling technologies.
Technology pertinent to large antenna systems, technology related to large space platform systems, and base technology applicable to both antenna and platform systems are discussed. Design studies, structural testing results, and theoretical applications are presented with accompanying validation data. A total systems approach including controls, platforms, and antennas is presented as a cohesive, programmatic plan for large space systems.
The ability of large information systems to support the changing data requirements of active science projects is being tested in a NASA collaborative study. This paper briefly profiles both the active science project and the large information system involved in this effort and offers some observations about the effectiveness of the project support. This is followed by lessons that are important for those participating in large information systems that need to support active science projects or that make available the valuable data produced by these projects. We learned in this work that it is difficult for a large information system focused on long term data management to satisfy the requirements of an on-going science project. For example, in order to provide the best service, it is important for all information system staff to keep focused on the needs and constraints of the scientists in the development of appropriate services. If the lessons learned in this and other science support experiences are not applied by those involved with large information systems of the EOS (Earth Observing System) era, then the final data products produced by future science projects may not be robust or of high quality, thereby making the conduct of the project science less efficacious and reducing the value of these unique suites of data for future research.
DYANA /dynamic analyzer/ is an advanced programming system which performs automatically the computing of a problem, as well as a major portion of the programming and analysis. The system is divided into time response and frequency response of dynamic and equivalent systems.
Solving large, sparse, linear systems of equations is one of the fundamental problems in large scale scientific and engineering computation. A model of a general class of asynchronous, iterative solution methods for linear systems is developed. In the model, the system is solved by creating several cooperating tasks that each compute a portion of the solution vector. This model is then analyzed to determine the expected intertask data transfer and task computational complexity as functions of the number of tasks. Based on the analysis, recommendations for task partitioning are made. These recommendations are a function of the sparseness of the linear system, its structure (i.e., randomly sparse or banded), and dimension.
Potential NASA missions utilizing large space systems are outlined. With reference to the Mobile Communications Satellite, four conceptual antenna configurations are described and compared. The Science and Applications Space Platform is described and the development of a power utilization package is examined. The package would be left on orbit and would provide a cost-effective means for collecting sunlight energy, storing it and providing power to Shuttle and payloads.
Reduction of a large system of equations to a lower-dimensional system of similar dynamics is investigated. For dynamical systems with disparate timescales, a criterion for determining redundant dimensions and a general reduction method based on the minimization of evolution rate are proposed.