SEARCH · Search NASA
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.
This paper presents a method for utilizing artificial neural networks for direct adaptive control of dynamic systems with poorly known dynamics. The neural network weights (controller gains) are adapted in real time using state measurements and a random search optimization algorithm. The results are demonstrated via simulation using two highly nonlinear systems.
Presented here is an identification technique that uses the sensor information to choose a new model out of a finite set of discrete model space, in order to follow the observed changes to the given time varying flexible structure. Boundary condition sets or other information on model variations are used to organize the set of possible models laterally into a search tree with levels of abstraction used to order the models vertically within branches. An object-oriented programming approach is used to represent the model set in the search tree. A modified A (asterisk) best first search algorithm finds the model where the model response best matches the current observations. Several extensions to this methodology are discussed. Methods of possible integration of rules with the current search algorithm are considered to give weight to interpreted trends that may be found in a series of observations. This capability might lead, for instance, to identifying a model that incorporates a progressive damage rather than with incorrect paramenters such as added mass. Another new direction is to consider the use of noisy time domain sensor feedback rather than frequency domain information in the search algorithm to improve the real-time capability of the developed procedure.
KBSDE is a knowledge compiler that uses a classification-based approach to map solution constraints in a task specification onto particular search algorithm components that will be responsible for satisfying those constraints (e.g., local constraints are incorporated in generators; global constraints are incorporated in either testers or hillclimbing patchers). Associated with each type of search algorithm component is a subcompiler that specializes in mapping constraints into components of that type. Each of these subcompilers in turn uses a classification-based approach, matching a constraint passed to it against one of several schemas, and applying a compilation technique associated with that schema. While much progress has occurred in our research since we first laid out our classification-based approach [Ton91], we focus in this paper on our reformulation research. Two important reformulation issues that arise out of the choice of a schema-based approach are: (1) compilability-- Can a constraint that does not directly match any of a particular subcompiler's schemas be reformulated into one that does? and (2) Efficiency-- If the efficiency of the compiled search algorithm depends on the compiler's performance, and the compiler's performance depends on the form in which the constraint was expressed, can we find forms for constraints which compile better, or reformulate constraints whose forms can be recognized as ones that compile poorly? In this paper, we describe a set of techniques we are developing for partially addressing these issues.
Integrated Search Environment (ISE), a software package that implements hierarchical searches with meta-control, is described in this manual. ISE is a collection of problem-independent routines to support solving searches. Mainly, these routines are core routines for solving a search problem and they handle the control of searches and maintain the statistics related to searches. By separating the problem-dependent and problem-independent components in ISE, new search methods based on a combination of existing methods can be developed by coding a single master control program. Further, new applications solved by searches can be developed by coding the problem-dependent parts and reusing the problem-independent parts already developed. Potential users of ISE are designers of new application solvers and new search algorithms, and users of experimental application solvers and search algorithms. The ISE is designed to be user-friendly and information rich. In this manual, the organization of ISE is described and several experiments carried out on ISE are also described.
MT's algorithm was developed as an aid in the design of space telecommunications systems when utilized with simultaneous range/command/telemetry operations. This algorithm provides selection of modulation indices for: (1) suppression of undesired signals to achieve desired link performance margins and/or to allow for a specified performance degradation in the data channel (command/telemetry) due to the presence of undesired signals (interferers); and (2) optimum power division between the carrier, the range, and the data channel. A software program using this algorithm was developed for use with MathCAD software. This software program, called the MT program, provides the computation of optimum modulation indices for all possible cases that are recommended by the Consultative Committee on Space Data System (CCSDS) (with emphasis on the squarewave, NASA/JPL ranging system).
NASA and the international community are investing in the development of a commercial transportation infrastructure that includes the increased use of rotorcraft, specifically helicopters and civil tilt rotors. However, there is significant concern over the impact of noise on the communities surrounding the transportation facilities. One way to address the rotorcraft noise problem is by exploiting powerful search techniques coming from artificial intelligence coupled with simulation and field tests to design low-noise flight profiles which can be tested in simulation or through field tests. This paper investigates the use of simulation based on predictive physical models to facilitate the search for low-noise trajectories using a class of automated search algorithms called local search. A novel feature of this approach is the ability to incorporate constraints directly into the problem formulation that addresses passenger safety and comfort.
NASA and the international community are investing in the development of a commercial transportation infrastructure that includes the increased use of rotorcraft, specifically helicopters and aircraft such as a 40-passenger civil tilt rotors. Rotorcraft have a number of advantages over fixed wing aircraft, primarily in not requiring direct access to the primary fixed wing runways. As such they can operate at an airport without directly interfering with major air carrier and commuter aircraft operations. However, there is significant concern over the impact of noise on the communities surrounding the transportation facilities. In this paper we propose to address the rotorcraft noise problem by exploiting powerful search techniques coming from artificial intelligence, coupled with simulation and field tests, to design trajectories that are expected to improve on the amount of ground noise generated. This paper investigates the use of simulation based on predictive physical models to facilitate the search for low-noise trajectories using a class of automated search algorithms called local search. A novel feature of this approach is the ability to incorporate constraints into the problem formulation that addresses passenger safety and comfort.
Laser Interferometer Space Antenna (LISA) is a planned NASA-ESA mission to be launched around 2012. The Gravitational Wave detection is fundamentally the determination of frequency, source parameters, and waveform amplitude derived in a specific order from the interferometric time-series of the rotating LISA spacecrafts. The LISA Science Team has developed a Mock LISA Data Challenge intended to promote the testing of complicated nested search algorithms to detect the 100-1 millihertz frequency signals at amplitudes of 10E-21. However, it has become clear that, sequential search of the parameters is very time consuming and ultra-sensitive; hence, a new strategy has been developed. Parallelization of existing sequential search algorithms of Gravitational Wave signal identification consists of decomposing sequential search loops, beginning with outermost loops and working inward. In this process, the main challenge is to detect interdependencies among loops and partitioning the loops so as to preserve concurrency. Existing parallel programs are based upon either shared memory or distributed memory paradigms. In PVM, master and node programs are used to execute parallelization and process spawning. The PVM can handle process management and process addressing schemes using a virtual machine configuration. The task scheduling and the messaging and signaling can be implemented efficiently for the LISA Gravitational Wave search process using a master and 6 nodes. This approach is accomplished using a server that is available at NASA Ames Research Center, and has been dedicated to the LISA Data Challenge Competition. Historically, gravitational wave and source identification parameters have taken around 7 days in this dedicated single thread Linux based server. Using PVM approach, the parameter extraction problem can be reduced to within a day. The low frequency computation and a proxy signal-to-noise ratio are calculated in separate nodes that are controlled by the master using message and vector of data passing. The message passing among nodes follows a pattern of synchronous and asynchronous send-and-receive protocols. The communication model and the message buffers are allocated dynamically to address rapid search of gravitational wave source information in the Mock LISA data sets.
This paper investigates the automated detection of jet contrails using data from the Advanced Very High Resolution Radiometer. A preliminary algorithm subtracts the 11.8-micron image from the 10.8-micron image, creating a difference image on which contrails are enhanced. Then a three-stage algorithm searches the difference image for the nearly-straight line segments which characterize contrails. First, the algorithm searches for elevated, linear patterns called 'ridges'. Second, it applies a Hough transform to the detected ridges to locate nearly-straight lines. Third, the algorithm determines which of the nearly-straight lines are likely to be contrails. The paper applies this technique to several test scenes.
An algorithm searches a star catalog to identify guide stars within the field of view of a telescope or camera. The algorithm is fast: the number of computations needed to perform the search is approximately proportional to the logarithm of the number of stars in the catalog. The algorithm requires the prior organization of the star catalog into a hierarchy utilizing independent spherical coverings (see figure), such that each successively higher level contains fewer elements. In the lowest and most numerous level of the hierarchy, the elements are individual stars in the star catalog. The next higher level contains a spherical covering (a constellation of n points on a sphere that minimizes the maximum distance of any point on the sphere from the closest one of the n points), the next higher level contains a smaller spherical covering, and so forth, ending at the highest level, which contains one element representing the point of entry into the search structure. With necessary exceptions at the lowest and highest levels, each element at each level is labeled in terms of the element to which it is linked in the next higher level and the first element to which it is linked in the next lower level. Each element is also labeled in terms of (1) its coordinates on the celestial sphere and (2) the largest angular distance to any element in any lower level in the hierarchy. The elements at all levels of the hierarchy are numbered on a single list, such that the elements of each constellation at each level are numbered consecutively. The algorithm is recursive. The input required to start the algorithm comprises the coordinates of a point on the celestial sphere. Attention is then focused on individual elements of the hierarchy, starting from the topmost one, as follows: The angle between the input point and the element under consideration is calculated. If the calculated angle is larger than the sum of (1) the predetermined angle to the most distant element plus (2) the half field of view of the telescope, then no stars will be within the field of view and this recursive part of the algorithm is terminated.
Three documents discuss the Sustainable Objective Valuation and Attainability (SOVA) algorithm and software as used to plan tasks (principally, scientific observations and associated maneuvers) for the Far Ultraviolet Spectroscopic Explorer (FUSE) satellite. SOVA is a means of managing risk in a complex system, based on a concept of computing the expected return value of a candidate ordered set of tasks as a product of pre-assigned task values and assessments of attainability made against qualitatively defined strategic objectives. For the FUSE mission, SOVA autonomously assembles a week-long schedule of target observations and associated maneuvers so as to maximize the expected scientific return value while keeping the satellite stable, managing the angular momentum of spacecraft attitude- control reaction wheels, and striving for other strategic objectives. A six-degree-of-freedom model of the spacecraft is used in simulating the tasks, and the attainability of a task is calculated at each step by use of strategic objectives as defined by use of fuzzy inference systems. SOVA utilizes a variant of a graph-search algorithm known as the A* search algorithm to assemble the tasks into a week-long target schedule, using the expected scientific return value to guide the search.
Distributed search algorithms are crucial in dealing with large optimization problems, particularly when a centralized approach is not only impractical but infeasible. Many machine learning concepts have been applied to search algorithms in order to improve their effectiveness. In this article we present an algorithm that blends Reinforcement Learning (RL) and hill climbing directly, by using the RL signal to guide the exploration step of a hill climbing algorithm. We apply this algorithm to the domain of a constellations of communication satellites where the goal is to minimize the loss of importance weighted data. We introduce the concept of 'ghost' traffic, where correctly setting this traffic induces the satellites to act to optimize the world utility. Our results indicated that the bi-utility search introduced in this paper outperforms both traditional hill climbing algorithms and distributed RL approaches such as team games.
A method of automated design of complex, modular robots involves an evolutionary process in which generative representations of designs are used. The term generative representations as used here signifies, loosely, representations that consist of or include algorithms, computer programs, and the like, wherein encoded designs can reuse elements of their encoding and thereby evolve toward greater complexity. Automated design of robots through synthetic evolutionary processes has already been demonstrated, but it is not clear whether genetically inspired search algorithms can yield designs that are sufficiently complex for practical engineering. The ultimate success of such algorithms as tools for automation of design depends on the scaling properties of representations of designs. A nongenerative representation (one in which each element of the encoded design is used at most once in translating to the design) scales linearly with the number of elements. Search algorithms that use nongenerative representations quickly become intractable (search times vary approximately exponentially with numbers of design elements), and thus are not amenable to scaling to complex designs. Generative representations are compact representations and were devised as means to circumvent the above-mentioned fundamental restriction on scalability. In the present method, a robot is defined by a compact programmatic form (its generative representation) and the evolutionary variation takes place on this form. The evolutionary process is an iterative one, wherein each cycle consists of the following steps: 1. Generative representations are generated in an evolutionary subprocess. 2. Each generative representation is a program that, when compiled, produces an assembly procedure. 3. In a computational simulation, a constructor executes an assembly procedure to generate a robot. 4. A physical-simulation program tests the performance of a simulated constructed robot, evaluating the performance according to a fitness criterion to yield a figure of merit that is fed back into the evolutionary subprocess of the next iteration. In comparison with prior approaches to automated evolutionary design of robots, the use of generative representations offers two advantages: First, a generative representation enables the reuse of components in regular and hierarchical ways and thereby serves a systematic means of creating more complex modules out of simpler ones. Second, the evolved generative representation may capture intrinsic properties of the design problem, so that variations in the representations move through the design space more effectively than do equivalent variations in a nongenerative representation. This method has been demonstrated by using it to design some robots that move, variously, by walking, rolling, or sliding. Some of the robots were built (see figure). Although these robots are very simple, in comparison with robots designed by humans, their structures are more regular, modular, hierarchical, and complex than are those of evolved designs of comparable functionality synthesized by use of nongenerative representations.
For hard computational problems, stochastic local search has proven to be a competitive approach to finding optimal or approximately optimal problem solutions. Two key research questions for stochastic local search algorithms are: Which algorithms are effective for initialization? When should the search process be restarted? In the present work we investigate these research questions in the context of approximate computation of most probable explanations (MPEs) in Bayesian networks (BNs). We introduce a novel approach, based on the Viterbi algorithm, to explanation initialization in BNs. While the Viterbi algorithm works on sequences and trees, our approach works on BNs with arbitrary topologies. We also give a novel formalization of stochastic local search, with focus on initialization and restart, using probability theory and mixture models. Experimentally, we apply our methods to the problem of MPE computation, using a stochastic local search algorithm known as Stochastic Greedy Search. By carefully optimizing both initialization and restart, we reduce the MPE search time for application BNs by several orders of magnitude compared to using uniform at random initialization without restart. On several BNs from applications, the performance of Stochastic Greedy Search is competitive with clique tree clustering, a state-of-the-art exact algorithm used for MPE computation in BNs.
Two algorithms for the search of signals in noise are constructed. Two independent measurable properties of a discreet time-dependent stochastic process F are described. The functions A sub K and B sub K fully satisfy conditions for the application of a test based on Spearman's rank correlation coefficient. A statistic to which the one-sided signal test can be applied is constructed under sufficiently natural assumptions about the noise process. The constructed statistics are simplified. Processing results from calculations of statistics constructed for concrete processes are presented.
A rapid, grid-based, target-search algorithm is presented to find candidate se-quences of small-body encounters for mission design. The algorithm is especially relevant for cases with large combinatorial spaces. In this paper, the al-gorithm is used to identify candidate flyby sequences of multiple Kuiper-Belt Ob-jects (KBOs). Before reaching the first KBO in the sequence, the trajectories in this paper first use gravity assists at one or more of the giant planets to pump-uptheir orbital energy—reducing launch C3. The target-search algorithm consists offour sequential steps: (1) parameter definition, (2) fine-tuned Lambert-based gridsearch of ballistic trajectories visiting one KBO, (3) rapid, ∆V-based proximitysearch for additional KBOs using the state transition matrices (STMs), and (4) tra-jectory optimization of the most promising KBO sequences using the EvolutionaryMission Trajectory Generator (EMTG). The paper also defines an empirical-basedprocess to characterize the maximum step size for the target arrival dates in theLambert grid search. Lastly, a candidate mission to two KBOs is presented. Theresults indicate that the ∆V computed from the STM propagations is not repre-sentative of the final ∆V computed in EMTG; however, it does serve as a useful‘reachability’ metric to identify nearby KBOs.
Multivariable optimization techniques are applied to a particular class of minimum weight structural design problems: the design of an axially loaded, pressurized, stiffened cylinder. Minimum weight designs are obtained by a variety of search algorithms: first- and second-order, elemental perturbation, and randomized techniques. An exterior penalty function approach to constrained minimization is employed. Some comparisons are made with solutions obtained by an interior penalty function procedure. In general, it would appear that an interior penalty function approach may not be as well suited to the class of design problems considered as the exterior penalty function approach. It is also shown that a combination of search algorithms will tend to arrive at an extremal design in a more reliable manner than a single algorithm. The effect of incorporating realistic geometrical constraints on stiffener cross-sections is investigated. A limited comparison is made between minimum weight cylinders designed on the basis of a linear stability analysis and cylinders designed on the basis of empirical buckling data. Finally, a technique for locating more than one extremal is demonstrated.
Here, we present two deterministic compilation algorithms for single-qutrit unitaries with O ( log 1 / ɛ ) gate depth. Each algorithm selects a nearby approximation to the target unitary and then exactly synthesizes the approximation over the Clifford + R basis. The first algorithm exhaustively searches over the group; while the second algorithm searches only for Householder reflections. The exhaustive search algorithm yields an average R count of 2.193 ( 11 ) + 8.621 ( 7 ) log 10 ( 1 / ɛ ) , albeit with a time complexity of O ( ɛ − 4.4 ) . The Householder search algorithm results in a larger average R count of 3.20 ( 13 ) + 10.77 ( 3 ) log 10 ( 1 / ɛ ) at a reduced time complexity of O ( ɛ − 0.42 ) , greatly extending the reach in ɛ . These costs correspond asymptotically to 35% and 69% more non-Clifford gates compared with synthesizing the same unitary with two qubits. Such initial results are encouraging for using the R gate as the nontransversal gate for qutrit-based computation.