Showing papers by "Worcester Polytechnic Institute published in 1986"

Knowledge and Control for a Mechanical Design Expert System

Abstract: M ost first-generation expert systems have been rule-based with a separate inference engine. However, a large unstructured collection of rules clearly lacks validity as a realistic model of design because reducing all knowledge to a single form does not recognize that there are many different types of knowledge used in any design problem-solving activity. Using such a collection of rules also does not recognize that design knowledge forms into clusters. Nor does it specify where or when this knowledge is to be applied, since different clusters ofknowledge may be applicable at different times. Similarly, using a single, central allpurpose inference engine ignores the richness of design problem-solving. Yet another problem is the potential for unfocused system behavior because all rules have equal status in the system and have equal potential for use. Many systems structure rules into sets. However, these sets are based on subtasks rather than on types of knowledge, 1-3 and the problem solving is still uniform, as the same inference engine acts on each rule set. Advocates of such structuring claim the subtasks can be solved linearly with no backtracking between tasks and with only minimal backtracking within tasks. Such subtask structure tells us more about the nature ofthe domain than about design, since it is clear that design decisions of any kind can often be wrong and, if so, will lead to attempts to recover from failure. The uniform rule representation and the lack of knowledge-dependent structure does not provide clear predictions about an expert's failure-recovery behavior. These problems stem mainly from a basic mismatch between the level of the tools available to build systems and the level of abstraction of the design task.

Modeling General Hysteresis Behavior and Random Vibration Application

Abstract: A simple constructive technique for the development of rate-type hysteresis models for general nonlinear system is presented. The technique is used to develop hysteresis models to incorporate time history-dependent postyield restorting forces, and general pinching behavior in smoothly varying deteriorating models. Applications of these models to random vibration analysis modeling via simulation and equivalent linearization techniques under Gaussian noise excitation is presented.

A portmanteau test for self-exciting threshold autoregressive-type nonlinearity in time series

Abstract: SUMMARY A portmanteau test to detect self-exciting threshold autoregressive-type nonlinearity in time series data is proposed. The test is based on cumulative sums of standardized residuals from autoregressive fits to the data. Significance levels for the test under the hypothesis of linearity are obtain from the asymptotic distribution of the cumulative sums as Brownian motion. The performance of the test is evaluated for simulated data from linear, bilinear and self-exciting threshold autoregressive models. It is also compared with another test which has been suggested for detecting general nonlinearity. Features of the proposed test, which make it useful in identifying the autoregressive order and the lag in threshold models, are discussed.

Multiple Bad Data Detectability and Identifiability: A Geometric Approach

Abstract: A method for detecting and identifying multiple bad data in electric power networks is developed by providing a geometric interpretation of the familiar normalized residuals test for single bad data. This generalized multiple bad data test amounts to determining whether the residual vector lies in a subspace determined by the suspect measurements and whether any portions of that subspace are orthogonal to the residual vector. These tests can be performed efficiently using appropriate projection matrices. Thne notion of critical measurement (removal renders the network unobservable) is extended to critical k-tuples of measurements to determine which bad data hypotheses are actually testable. For example, gross errors in critical measurements are not detectable, and gross errors in a critical pair of measurements are detectable but not identifiable. More generally, k-2 gross errors in a critical k-tuple of measurements are identifiable while k or k-l gross errors are detectable but not identifiable. In essence, the set of testable hypotheses is determined by the geometry of the space spanned by all possible residual vectors. A procedure for selecting and pruning a suspect set of measurements is described. Examples for the IEEE 14 bus network are provided.

Solution of elasticity problems by a self‐adaptive mesh refinement technique for boundary element computation

Abstract: A self-adaptive mesh refinement technique is developed for boundary element solutions of the two-dimensional Navier's equation. The method is based on error reduction and applied on the element and global level to estimate the residual error associated with each mesh. Results for several problems employing constant two-dimensional isotropic elastostatic boundary elements are presented.

Effective heat transfer parameters for transient packed-bed models

Abstract: A general theory is presented for the prediction of effective axial conductivity and overall radial heat transfer coefficient for fluid flow through packed beds, under either transient or steady state conditions. The parameters to be used in each case may be quite different, depending on the dominant phase for heat transfer. The theory unifies all previous results and explains conflicting conclusions arrived at by different workers. The ability of a pseudohomogeneous model to reproduce packedbed transients is shown to depend upon the choice of effective parameters.

A Microprocessor-Based Technique for Detection of High Impedance Faults

Abstract: Detection of high impedance faults on distribution systems is difficult due to the low current levels which flow in the fault. The objective is to detect these faults with some device or mechanism situated at the substation without modification to the distribution lines themselves. The technique presented in this paper monitors the unbalance in the fundamental, third and fifth harmonic feeder currents at the substation and performs a statistical evaluation of the present unbalance relative to past levels of unbalance. Using hypothesis testing, if the level of unbalance exceeds a threshold, a fault is indicated. The technique enploys a real-time algorithm suitable for implementation on a microprocessor-based digital relay located in the substation.

Cadmium and zinc reversibly arrest development of Artemia larvae.

Abstract: Despite the widespread distribution of heavy metals such as cadmium and zinc in the environment and their well-known cytotoxicity and embryotoxicity in mammals, comparatively little is known about their effect on aquatic organisms, particularly invertebrates. Post-gastrula and early larval development of the brine shrimp, Artemia, present some useful advantages for studies of developmental aspects of environmental toxicology. Dormant encysted gastrulae, erroneously called brine shrimp eggs, can be obtained commercially and raised in the laboratory under completely defined conditions. Following a period of post-gastrula development within the cyst, pre-nauplius larvae emerge through a crack in the cyst shell. A few hours later, free-swimming nauplius larvae hatch. Cadmium is acutely toxic to both adults and nauplius larvae of Artemia, but the reported LC50s are as high as 10 mM, depending on larval age. In this paper the authors show that pre-nauplius larvae prior to hatching are much more sensitive to cadmium than are hatched nauplius larvae. At 0.1 ..mu..m, cadmium retards development and hatching of larvae; higher concentrations block hatching almost completely and thus are lethal. However, the larvae arrested at the emergence stage survive for 24 hours or more before succumbing to the effects of cadmium, and during this period themore » potentially lethal effect is reversible if the larvae are placed in cadmium-free medium. The effects of zinc parallel those of cadmium, although zinc is somewhat less toxic than cadmium at equal concentrations.« less

Gradients of behaviour constraints and reanalysis via the integrated force method

Abstract: A matrix formulation of structural analysis, known as the integrated force method (IFM) is an attempt to establish an approach in which internal forces are the unknowns. Gradients of such behaviour constraints as stress, displacement and frequency can be obtained in closed form via the IFM. Closed form reanalysis formulae can also be generated. Computational experiments reported here show that reanalysis and gradient generation can be performed efficiently by this approach.

Contribution to the Theory of Stochastically Periodic Harmonics in Power Systems

Abstract: A basic approach for quantitative evaluation of "Stochastically Periodic" harmonics effects on electrical equipment is presented. The main contribution of this paper is in defining and explaining a mathematical method which enables the estimation of the additional losses and the loss of life caused by stochastically periodic current/voltage harmonics. A practical case of N identical converters, randomly operated, is analyzed. The suggested method is applied to determine admissible levels of randomly periodic voltage harmonics for power capacitors.

Comparison of elderly and younger users on keyboard and voice input computer-based composition tasks

Abstract: An experiment was run in which elderly and younger people used a keyboard editor and a simulated listening typewriter to compose letters. Performance was measured and participants rated the systems used.Our general conclusions were as follows: There are no major differences in performance between elderly computer users and their younger counterparts in carrying out a computer-based composition task.Elders appear to be more enthusiastic users of computer systems than are younger people. This is shown by preference ratings, behavioral observations, and post-experimental debriefings.Voice input does not improve performance on composition tasks, but it is greatly preferred over the traditional keyboard input method.

Stability of continuous culture with recombinant organisms.

Abstract: It was found that both poor selection pressure and a variable rate of plasmid loss were present in the system studied and that both have significant effects on continuous reactor operation. At least some of these effects were analyzed by a simple model. At this point, experimental analysis for extracellular levels of tryptophan sufficient to support X- growth (1-4 mg/l) has given contradictory results. This has at least partially indicated the effect may be an intracellular one, and thus the culture history would be critical in such experiments. Since the system studied is not atypical of recombinant cultures, it leads one to speculate on the generality of the phenomena and its extent in other cultures. If important, the use of double auxotrophs or auxotrophs that are mutant in a metabolite for which the cell has a greater growth requirement should be used. Additionally, the presence of higher copy numbers in yeast at lower growth rates also leads one to speculate on how these apparently contradictory phenomena are related.

A self-adaptive mesh refinement technique for boundary element solution of the Navier's equation

Abstract: A self-adaptive mesh refinement technique is developed for boundary element solutions of the two-dimensional Laplace equation. The method is based on error reduction and applied on the element and global level to estimate the error associated with each mesh. This adaptive technique is then utilized to analyze problems with and without singularities. Results employing constant two-dimensional boundary elements are presented.

Nitrous oxide production by cyanobacteria

Abstract: Evidence is presented here that axenic cultures of Nostoc spp., Aphanocapsa (PCC 6308), and Aphanocapsa (PCC 6714) but not Anacystis nidulans R-2 (PCC 7942) produce N2O and ammonia when grown on nitrite. The data suggest that the cyanobacteria produce N2O by nitrite reduction to ammonia.

A triangular thin-shell finite element based on discrete Kirchhoff theory

Abstract: A three-node, curved thin-shell triangular element with simple nodal connections is developed. The displacement and rotation components are independently interpolated by complete cubic and quadratic polynomials respectively. The Kirchhoff hypothesis is enforced at a discrete number of points in the element. The rigid-body displacement condition is satisfied by isoparametric interpolation of the shell geometry within the element. A detailed numerical evaluation through a number of standard problems is performed.

Detecting non-linearity in time series

Abstract: We report our experiences with using tests of linearity in time series proposed by McLeod & Li (1983) and Keenan (1985). Empirical significance levels are checked and power studies reported when the tests are applied to bilinear and SETAR time serious models. We also report their application to detect nonlinearity on a large collection of time series. In general the performance of the former statistic is poor except for large sample sizes whilst the latter is superior both for bilinear and SETAR time series. In an appendix we give the MINITAB macros so that anyone with this package could easily implement the tests used.

On the consistency of least squares estimators for a threshold ar(1) model

Abstract: . For the SETAR (2; 1,1) model where {at(i)} are i.i.d. random variables with mean 0 and variance σ2(i), i = 1,2, and {at(l)} is independent of {at(2)}, we consider estimators of φ1, φ2 and r which minimize weighted sums of the sum of squares functions for σ2(1) and σ2(2). These include as a special case the usual least squares estimators. It is shown that the usual least squares estimators of φ1, φ2 and r are consistent. If σ2(1) ≠σ2(2) conditions on the weights are found under which the estimators of r and φ1 or φ2 are not consistent.

A nonlinear integral operator arising from a model in population genetics IV. clines

Abstract: We study the existence, uniqueness and stability properties of solutions to the integral equation $\phi = Q[\phi ]$ with $\phi ( - \infty ) = 1$, $\phi ( \infty ) = 0$. Here $Q[u](x) = \smallint K(x - y)g(y,u(y))dy$ is defined on functions bounded between 0 and 1, K is a probability density function and ${{g(x,u) = [s(x)u^2 + u]} / {[1 + s(x)u^2 \sigma (x)(1 - u)^2 ]}}$ according to a population genetics model. The hypotheses on g are based on the biological assumption that the homozygotes, that is individuals with genotypes $AA$ or $aa$, are best fit to survive near opposite ends of the one-dimensional habitat.

Factorization of systems of differential equations

Abstract: It is shown that the classical Infeld–Hull factorization method can be extended to coupled systems of second‐order equations. A complete solution of the factorization equations in two dimensions is given and a partial enumeration of factorizable systems is made.

The role and position of graphics in computer science education

Abstract: The SIGGRAPH Education Committee has been considering recommendations for inclusion of graphics in various curricula for higher education. Several issues of computer graphics in computer science or computer science/engineering curricula are identified here. In particular the course content and support facilities necessary and the status of graphics courses within programs are discussed. A basic premise is stated that Curriculum '78[1] and other guidelines[2, 3, 4, 5, 6] for educational programs are flawed in their lack of computer graphics content. The reasons for this are not immediately apparent, but might in part be attributed to the lower profile of graphics during the time of their initial development. Since that time the significant technical advancements and standardization of terms and concepts have not been incorporated as changes in recommended curricula designs.

Mixed-state SU(2) squeezing and its occurrence in two quantum optical models

Abstract: The atomic pseudospin operators describing a two-level atom, and obeying the SU(2) Lie algebra, can exhibit reduced quantum fluctuations (“squeezing”) in certain states of the atom. These states include both pure and mixed states of the atom. Parameterizing the atom by the spherical polar components S, θ, and ϕ of its pseudospin vector S, we determine the parameter ranges in which squeezing can occur and the magnitude of this squeezing. We illustrate our considerations by examining two models, namely, the coherent-state Jaynes–Cummings model and the optical Bloch equations, in which mixed-state SU(2) squeezing arises dynamically.

Learning computer programming through dynamic representation of computer functioning: evaluation of a new learning package for Pascal

Abstract: This paper describes and evaluates a new approach to teaching the beginning Pascal programming course at Worcester Polytechnic Institute. At the heart of this approach is a new computer learning package, called PASLAB, which allows students to understand what is happening inside the computer relative to statements in Pascal programs constructed by an expert. There were 322 students participating in the evaluation under traditional conditions in 1984, and 296 students participating under the new conditions in 1985. Comparison between the two groups was made by examining the respective regression equations predicting final grade for each set of students. While the characteristics of both sets of students were very similar, the regression equations were markedly different. Under traditional conditions, background characteristics of students, including programming experience, had a major impact on the grades those students received. With Paslab in place, the impact of those background characteristics decreased while the importance of psychological orientations (motivation) increased. There were no differences in performance with respect to gender. Further research is needed to test the impact of Paslab in non-technical institutions. The positive results of this study suggest the fruitfulness of developing additional computer learning packages on the same principles as Paslab to be used in other computer-science courses as well as in other fileds.