SEARCH · Search NASA

Results for “Decentralized optimization”

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.

SourceResearch connectionFrom yearThrough year

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

At least 55 records · Page 3

All sources NASA NTRS DOE OSTI

NASA NTRS1999

Feasibility of Decentralized Linear-Quadratic-Gaussian Control of Autonomous Distributed Spacecraft

A distributed satellite formation, modeled as an arbitrary number of fully connected nodes in a network, could be controlled using a decentralized controller framework that distributes operations in parallel over the network. For such problems, a solution that minimizes data transmission requirements, in the context of linear-quadratic-Gaussian (LQG) control theory, was given by Speyer. This approach is advantageous because it is non-hierarchical, detected failures gracefully degrade system performance, fewer local computations are required than for a centralized controller, and it is optimal with respect to the standard LQG cost function. Disadvantages of the approach are the need for a fully connected communications network, the total operations performed over all the nodes are greater than for a centralized controller, and the approach is formulated for linear time-invariant systems. To investigate the feasibility of the decentralized approach to satellite formation flying, a simple centralized LQG design for a spacecraft orbit control problem is adapted to the decentralized framework. The simple design uses a fixed reference trajectory (an equatorial, Keplerian, circular orbit), and by appropriate choice of coordinates and measurements is formulated as a linear time-invariant system.

Carpenter, J. Russell↗

NASA NTRS1999

Feasibility of Decentralized Linear-Quadratic-Gaussian Control of Autonomous Distributed Spacecraft

Carpenter, J. Russell↗

NASA NTRS1985

Thermal storage analysis for large manned space platforms

High electrical power and waste heat rejection is projected for future manned low earth orbit space platforms, such as Space Station. The high heat rejection, optical coating degradation, long operating life with minimum maintenance requirements pose a challenging thermal management design problem. System optimization, with respect to radiator area and weight, indicate the requirement for thermal storage. This paper examines the thermal storage benefits, determines the characteristics as applied to different TMS concepts (e.g., centralized, decentralized), and examines the similarities and differences of thermal storage integration with single-phase and two-phase systems for a study baseline 75 kWe low earth orbit platform.

Lehtinen, A. M.↗

NASA NTRS1987

Task allocation among multiple intelligent robots

Researchers describe the design of a decentralized mechanism for allocating assembly tasks in a multiple robot assembly workstation. Currently, the approach focuses on distributed allocation to explore its feasibility and its potential for adaptability to changing circumstances, rather than for optimizing throughput. Individual greedy robots make their own local allocation decisions using both dynamic allocation policies which propagate through a network of allocation goals, and local static and dynamic constraints describing which robots are elibible for which assembly tasks. Global coherence is achieved by proper weighting of allocation pressures propagating through the assembly plan. Deadlock avoidance and synchronization is achieved using periodic reassessments of local allocation decisions, ageing of allocation goals, and short-term allocation locks on goals.

Gasser, L.↗

NASA NTRS1994

Computer code for controller partitioning with IFPC application: A user's manual

A user's manual for the computer code for partitioning a centralized controller into decentralized subcontrollers with applicability to Integrated Flight/Propulsion Control (IFPC) is presented. Partitioning of a centralized controller into two subcontrollers is described and the algorithm on which the code is based is discussed. The algorithm uses parameter optimization of a cost function which is described. The major data structures and functions are described. Specific instructions are given. The user is led through an example of an IFCP application.

Schmidt, Phillip H.↗

NASA NTRS1991

Decentralized hierarchical partitioning of centralized integrated controllers

A framework for a decentralized hierarchical controller partitioning structure is developed. This structure allows for the design of separate airframe and propulsion controllers which, when assembled, will meet the overall design criterion for the integrated airframe/propulsion system. An algorithm based on parameter optimization of the state-space representation for the subsystem controllers is described. The algorithm is currently being applied to an integrated flight propulsion control design example.

Schmidt, Phillip↗

NASA NTRS1990

Substructuring decomposition and controller synthesis

A decentralized design procedure is proposed for the control design for flexible structures. The structure to be controlled is decomposed into several substructures by using a natural decomposition called substructuring decomposition. For each substructure, a subcontroller is designed by using the linear quadratic optimal control theory. Then, a controller synthesis scheme called Substructural Controller Synthesis (SCS) is used to assemble the subcontrollers into a system controller, which is to be used to control the whole structure. A plane truss example is used to illustrate the efficacy of the proposed method.

Su, Tzu-Jeng↗

NASA NTRS1989

Substructural controller synthesis

A decentralized design procedure which combines substructural synthesis, model reduction, decentralized controller design, subcontroller synthesis, and controller reduction is proposed for the control design of flexible structures. The structure to be controlled is decomposed into several substructures, which are modeled by component mode synthesis methods. For each substructure, a subcontroller is designed by using the linear quadratic optimal control theory. Then, a controller synthesis scheme called Substructural Controller Synthesis (SCS) is used to assemble the subcontrollers into a system controller, which is to be used to control the whole structure.

Su, Tzu-Jeng↗

NASA NTRS1999

Integration of a Decentralized Linear-Quadratic-Gaussian Control into GSFC's Universal 3-D Autonomous Formation Flying Algorithm

A decentralized control is investigated for applicability to the autonomous formation flying control algorithm developed by GSFC for the New Millenium Program Earth Observer-1 (EO-1) mission. This decentralized framework has the following characteristics: The approach is non-hierarchical, and coordination by a central supervisor is not required; Detected failures degrade the system performance gracefully; Each node in the decentralized network processes only its own measurement data, in parallel with the other nodes; Although the total computational burden over the entire network is greater than it would be for a single, centralized controller, fewer computations are required locally at each node; Requirements for data transmission between nodes are limited to only the dimension of the control vector, at the cost of maintaining a local additional data vector. The data vector compresses all past measurement history from all the nodes into a single vector of the dimension of the state; and The approach is optimal with respect to standard cost functions. The current approach is valid for linear time-invariant systems only. Similar to the GSFC formation flying algorithm, the extension to linear LQG time-varying systems requires that each node propagate its filter covariance forward (navigation) and controller Riccati matrix backward (guidance) at each time step. Extension of the GSFC algorithm to non-linear systems can also be accomplished via linearization about a reference trajectory in the standard fashion, or linearization about the current state estimate as with the extended Kalman filter. To investigate the feasibility of the decentralized integration with the GSFC algorithm, an existing centralized LQG design for a single spacecraft orbit control problem is adapted to the decentralized framework while using the GSFC algorithm's state transition matrices and framework. The existing GSFC design uses both reference trajectories of each spacecraft in formation and by appropriate choice of coordinates and simplified measurement modeling is formulated as a linear time-invariant system. Results for improvements to the GSFC algorithm and a multiple satellite formation will be addressed. The goal of this investigation is to progressively relax the assumptions that result in linear time-invariance, ultimately to the point of linearization of the non-linear dynamics about the current state estimate as in the extended Kalman filter. An assessment will then be made about the feasibility of the decentralized approach to the realistic formation flying application of the EO-1/Landsat 7 formation flying experiment.

Folta, David C.↗

NASA NTRS1989

Figure control concepts for segmented reflector telescopes

Control analysis activities related to the development of figure control technologies for large space telescopes with precision segmented actively controlled primary reflectors are reviewed. The Precision Segmented Reflector (PSR) configuration is described along with the development and use of the PSR models. Geometric and dynamic models, characterization of figure estimation errors and optimal sensor placement, and the development of quasi-static and dynamic control concepts are outlined. The structure of a quasi-static controller is presented, that utilizes edge sensor measurements to estimate displacement errors at the actuator level and generates figure correction commands. The second approach considers decentralization of the reflector system at the panel level, while the third considers decentralization at the actuator level.

Boussalis, D.↗

NASA NTRS1994

Communication network for decentralized remote tele-science during the Spacelab mission IML-2

The ESA communication network for decentralized remote telescience during the Spacelab mission IML-2, called Interconnection Ground Subnetwork (IGS), provided data, voice conferencing, video distribution/conferencing and high rate data services to 5 remote user centers in Europe. The combination of services allowed the experimenters to interact with their experiments as they would normally do from the Payload Operations Control Center (POCC) at MSFC. In addition, to enhance their science results, they were able to make use of reference facilities and computing resources in their home laboratory, which typically are not available in the POCC. Characteristics of the IML-2 communications implementation were the adaptation to the different user needs based on modular service capabilities of IGS and the cost optimization for the connectivity. This was achieved by using a combination of traditional leased lines, satellite based VSAT connectivity and N-ISDN according to the simulation and mission schedule for each remote site. The central management system of IGS allows minimization of staffing and the involvement of communications personnel at the remote sites. The successful operation of IGS for IML-2 as a precursor network for the Columbus Orbital Facility (COF) has proven the concept for communications to support the operation of the COF decentralized scenario.

Christ, Uwe↗

NASA NTRS1991

Robust decentralized control laws for the ACES structure

Control system design for the Active Control Technique Evaluation for Spacecraft (ACES) structure at NASA Marshall Space Flight Center is discussed. The primary objective of this experiment is to design controllers that provide substantial reduction of the line-of-sight pointing errors. Satisfaction of this objective requires the controllers to attenuate beam vibration significantly. The primary method chosen for control design is the optimal projection approach for uncertain systems (OPUS). The OPUS design process allows the simultaneous tradeoff of five fundamental issues in control design: actuator sizing, sensor accuracy, controller order, robustness, and system performance. A brief description of the basic ACES configuration is given. The development of the models used for control design and control design for eight system loops that were selected by analysis of test data collected from the structure are discussed. Experimental results showing that very significant performance improvement is achieved when all eight feedback loops are closed are presented.

Collins, Emmanuel G., Jr.↗

NASA NTRS1990

Randomly Sampled-Data Control Systems

The purpose is to solve the Linear Quadratic Regulator (LQR) problem with random time sampling. Such a sampling scheme may arise from imperfect instrumentation as in the case of sampling jitter. It can also model the stochastic information exchange among decentralized controllers to name just a few. A practical suboptimal controller is proposed with the nice property of mean square stability. The proposed controller is suboptimal in the sense that the control structure is limited to be linear. Because of i. i. d. assumption, this does not seem unreasonable. Once the control structure is fixed, the stochastic discrete optimal control problem is transformed into an equivalent deterministic optimal control problem with dynamics described by the matrix difference equation. The N-horizon control problem is solved using the Lagrange's multiplier method. The infinite horizon control problem is formulated as a classical minimization problem. Assuming existence of solution to the minimization problem, the total system is shown to be mean square stable under certain observability conditions. Computer simulations are performed to illustrate these conditions.

Han, Kuoruey↗

NASA NTRS2002

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↗

NASA NTRS1976

Trends in modern system theory

The topics considered are related to linear control system design, adaptive control, failure detection, control under failure, system reliability, and large-scale systems and decentralized control. It is pointed out that the design of a linear feedback control system which regulates a process about a desirable set point or steady-state condition in the presence of disturbances is a very important problem. The linearized dynamics of the process are used for design purposes. The typical linear-quadratic design involving the solution of the optimal control problem of a linear time-invariant system with respect to a quadratic performance criterion is considered along with gain reduction theorems and the multivariable phase margin theorem. The stumbling block in many adaptive design methodologies is associated with the amount of real time computation which is necessary. Attention is also given to the desperate need to develop good theories for large-scale systems, the beginning of a microprocessor revolution, the translation of the Wiener-Hopf theory into the time domain, and advances made in dynamic team theory, dynamic stochastic games, and finite memory stochastic control.

Athans, M.↗

NASA NTRS1990

H2, fixed architecture, control design for large scale systems

The H2, fixed architecture, control problem is a classic linear quadratic Gaussian (LQG) problem whose solution is constrained to be a linear time invariant compensator with a decentralized processing structure. The compensator can be made of p independent subcontrollers, each of which has a fixed order and connects selected sensors to selected actuators. The H2, fixed architecture, control problem allows the design of simplified feedback systems needed to control large scale systems. Its solution becomes more complicated, however, as more constraints are introduced. This work derives the necessary conditions for optimality for the problem and studies their properties. It is found that the filter and control problems couple when the architecture constraints are introduced, and that the different subcontrollers must be coordinated in order to achieve global system performance. The problem requires the simultaneous solution of highly coupled matrix equations. The use of homotopy is investigated as a numerical tool, and its convergence properties studied. It is found that the general constrained problem may have multiple stabilizing solutions, and that these solutions may be local minima or saddle points for the quadratic cost. The nature of the solution is not invariant when the parameters of the system are changed. Bifurcations occur, and a solution may continuously transform into a nonstabilizing compensator. Using a modified homotopy procedure, fixed architecture compensators are derived for models of large flexible structures to help understand the properties of the constrained solutions and compare them to the corresponding unconstrained ones.

Mercadal, Mathieu↗

NASA NTRS1986

Design and evaluation of a failure detection and isolation algorithm for restructurable control systems

The use of a decentralized approach to failure detection and isolation for use in restructurable control systems is examined. This work has produced: (1) A method for evaluating fundamental limits to FDI performance; (2) Application using flight recorded data; (3) A working control element FDI system with maximal sensitivity to critical control element failures; (4) Extensive testing on realistic simulations; and (5) A detailed design methodology involving parameter optimization (with respect to model uncertainties) and sensitivity analyses. This project has concentrated on detection and isolation of generic control element failures since these failures frequently lead to emergency conditions and since knowledge of remaining control authority is essential for control system redesign. The failures are generic in the sense that no temporal failure signature information was assumed. Thus, various forms of functional failures are treated in a unified fashion. Such a treatment results in a robust FDI system (i.e., one that covers all failure modes) but sacrifices some performance when detailed failure signature information is known, useful, and employed properly. It was assumed throughout that all sensors are validated (i.e., contain only in-spec errors) and that only the first failure of a single control element needs to be detected and isolated. The FDI system which has been developed will handle a class of multiple failures.

Weiss, Jerold L.↗

NASA NTRS1988

Alpha: A real-time decentralized operating system for mission-oriented system integration and operation

Alpha is a new kind of operating system that is unique in two highly significant ways. First, it is decentralized transparently providing reliable resource management across physically dispersed nodes, so that distributed applications programming can be done largely as though it were centralized. And second, it provides comprehensive, high technology support for real-time system integration and operation, an application area which consists predominately of aperiodic activities having critical time constraints such as deadlines. Alpha is extremely adaptable so that it can be easily optimized for a wide range of problem-specific functionality, performance, and cost. Alpha is the first systems effort of the Archons Project, and the prototype was created at Carnegie-Mellon University directly on modified Sun multiprocessor workstation hardware. It has been demonstrated with a real-time C(sup 2) application. Continuing research is leading to a series of enhanced follow-ons to Alpha; these are portable but initially hosted on Concurrent's MASSCOMP line of multiprocessor products.

Jensen, E. Douglas↗