Advisor Feed
https://feeds.library.caltech.edu/people/McCann-G-D/advisor.rss
A Caltech Library Repository Feedhttp://www.rssboard.org/rss-specificationpython-feedgenenTue, 16 Apr 2024 15:38:22 +0000An Arbitrary-Function Generator and its Application to the Study of Some Non-Linear Systems on the Electric Analog Computer
https://resolver.caltech.edu/CaltechTHESIS:02202024-220557703
Authors: {'items': [{'id': 'Buchholz-Werner', 'name': {'family': 'Buchholz', 'given': 'Werner'}, 'show_email': 'NO'}]}
Year: 1950
DOI: 10.7907/qafc-cv17
<p>An arbitrary-function generator, capable of
furnishing an output voltage which may be any given single-valued
function of the input voltage, is described. It was
constructed for use with the Electric Analog Computer at the
California Institute of Technology. In principle it is an
optical follower system containing a cathode-ray tube and a
phototube with an opaque mask representing the desired
arbitrary function.</p>
<p>The working model constructed is accurate to about
one or two percent, has a phase shift of 3½ degrees at 2 kc
and a limiting time delay of about 50 microseconds with the
beam traversing a step equal to the maximum height of the
function pattern employed.</p>
<p>Sources of error are analyzed and suggestions are
made for possible improvements. A major problem which has
not been solved yet satisfactorily is the local deterioration
of the cathode-ray tube screen due to fatigue.</p>
<p>A number of non-linear problems are solved with the
aid of this device, all of them dealing with second-order
systems having a single degree of freedom and only one
non-linear term in the equation. Among the systems discussed
are series circuits with a non-linear capacitance, or
mechanical systems with a non-linear spring; non-linear
damping; oscillating systems, including systems behaving
according to Van der Pol's equation; and mechanical systems
made unstable by the presence of static (dry) friction
greater than the dynamic (Coulomb) friction.</p>
<p>The results obtained appear to support the
conclusion that the generator introduces only one or two
percent of distortion, plus a small phase shift at the
higher frequencies. The device makes possible the solution
of many non-linear problems, hitherto inaccessible, at
relatively high natural frequencies and repetition rates.</p>https://thesis.library.caltech.edu/id/eprint/16298Analog Methods of Nonlinear Vibration Analysis
https://resolver.caltech.edu/CaltechTHESIS:03242016-111828484
Authors: {'items': [{'id': 'Greenwood-Donald-Theodore', 'name': {'family': 'Greenwood', 'given': 'Donald Theodore'}, 'show_email': 'NO'}]}
Year: 1951
DOI: 10.7907/0DNR-VB14
<p>This thesis presents methods by which electrical analogies can be obtained for nonlinear systems. The accuracy of these methods is investigated and several specific types of nonlinear equations are studied in detail.</p>
<p>In Part I a general method is given for obtaining electrical analogs of nonlinear systems with one degree of freedom. Loop and node methods are compared and the stability of the loop analogy is briefly considered.</p>
<p>Parts II and III give a description of the equipment and a discussion of its accuracy. Comparisons are made between experimental and analytic solutions of linear systems.</p>
<p>Part IV is concerned with systems having a nonlinear restoring force. In particular, solutions of Duffing's equation are obtained, both by using the electrical analogy and also by approximate analytical methods.</p>
<p>Systems with nonlinear damping are considered in Part V. Two specific examples are chosen: (1) forced oscillations and (2) self-excited oscillations (van der Pol’s equation). Comparisons are made with approximate analytic solutions.</p>
<p>Part VI gives experimental data for a system obeying
Mathieu's equation. Regions of stability are obtained. Examples of subharmonic, ultraharmonic, and ultrasubharmonic oscillat1ons are shown.</p>
https://thesis.library.caltech.edu/id/eprint/9633Factors Affecting the Rate of Dielectric Recovery of Power Arcs in Long Gaps
https://resolver.caltech.edu/CaltechTHESIS:10112017-102446992
Authors: {'items': [{'id': 'Ellis-Harry-McPhee', 'name': {'family': 'Ellis', 'given': 'Harry McPhee'}}]}
Year: 1951
DOI: 10.7907/KAYK-Z254
<p>A satisfactory testing technique is developed for investigating
the recovery characteristics of various power system insulations.</p>
<p>Duration of the fault current, for a vertical test gap, has little
effect on the dielectric recovery. Five cycle fault current duration
decreases the initial rate of recovery, for a horizontal test gap,
below that for a half cycle, but for longer time delays thermal
convection increases the rate of recovery above that for a half cycle
duration.</p>
<p>The 11 inch test gap has a lower percentage recovery than the 6
inch test gap. The lower the boiling point and the lower the ionization
potential of the electrode material, the lower the rate of recovery.</p>
<p>The increase in the rate of recovery produced by winds up to 1000
feet per minute is mainly due to displacement of the ionized gases from
the test electrodes. For recovery based on this displacement of the
ionized gases from the electrodes, the axial wind recovery should be
half of the perpendicular. However, the axial wind decreases the cross
section of the arc, heat transfer away from the ionized gases is
increased, and the axial recovery is greater than half of the perpendicular
recovery. Wind velocities less than 100 feet per minute have a
negligible effect on the dielectric recovery characteristics.</p>
<p>A 3000 frame per second camera is used to make a high speed photographic
study of the arc. By using a system of mirrors, two mutually
perpendicular views of the arc are obtained simultaneously, on each
frame of the film. Breakdown occurs in the weakest path through the
ionized gases, even though this path is 2 to 3 times as long as the
electrode separation.</p>
<p>The extent of ionization of the luminous gases is obtained with
the aid of a microphotometer by taking density readings directly from
the image of the arc on the 16 mm. negative film. The average density
variation with time follows the general shape of the curve based on
diffusion as the only important deionizing agent, for a short time after
current zero.</p>
<p>The local coefficient of heat transfer obtained from the analogy
of the arc to a hot cylindrical solid body increases with the wind
velocity and the density of ionization of the luminous gases decreases
with increasing wind velocity.</p>
<p>The diameter and density of the arc are obtained from the photographic
study and the recovery voltage calculated, at various time delays,
from Slepian's theory for the critical breakdown gradient of the
arc column. The results are in general qualitative agreement with the
experimental results.</p>
<p>The minimum reclosing time for high-speed automatic-reclosing circuit
breakers on a transmission system can be reduced by a factor of
five for a wind of 1000 feet per minute (15 miles per hour) blowing
across the ionized fault path.</p>
https://thesis.library.caltech.edu/id/eprint/10503Temperature Measurements of Large Power Arcs and the Relation of Temperature to Dielectric Recovery
https://resolver.caltech.edu/CaltechETD:etd-05122003-095829
Authors: {'items': [{'id': 'Duesterhoeft-William-Charles', 'name': {'family': 'Duesterhoeft', 'given': 'William Charles'}, 'show_email': 'NO'}]}
Year: 1953
DOI: 10.7907/XRPC-8159
A study of the basic mechanisms of dielectric recovery of power frequency arcs in air is reported. The particular arc studied is that of a standard 6 inch rod gap conducting 300 or 800 amperes crest for 1/2 cycle of a 60 cycle per second current.
The temperature of the arc space and the variation of the arc temperature with time are measured. The temperature is measured by spectrographic and velocity of sound techniques. The temperature at current zero is 5000 degrees Kelvin. The temperature is 720 degrees K. 77 milliseconds after current zero and 415 degrees K. 196 milliseconds after current zero.
The temperature is correlated to the dielectric strength during the recovery period. During the recovery period the arc space is at atmospheric pressure. The gas density is reduced due to the high temperatures existing in the arc space. Evidence is given that the low gas density is reduced due to the high temperatures existing in the arc space. Evidence is given that the low gas density is the major factor causing the reduced dielectric strength.
Residual ions in the arc space also contribute to reducing the dielectric strength. A criterion for recovery breakdown is given including the effect of the residual ions. The ion density after current zero decreases due to electron recombination with a recombination coefficient [alpha] = 2.0 x 10[superscript -9] cubic centimeter per ion-second. The loss of ions by diffusion appears negligible.https://thesis.library.caltech.edu/id/eprint/1730A Special Purpose Electric Analog Computer and its Application to the Solution of Certain Nonlinear Differential Equations
https://resolver.caltech.edu/CaltechETD:etd-05062003-110154
Authors: {'items': [{'id': 'Howard-Robert-Charles', 'name': {'family': 'Howard', 'given': 'Robert Charles'}, 'show_email': 'NO'}]}
Year: 1953
DOI: 10.7907/Y018-YQ64
This thesis consists of three parts. (1) An analog computer investigation was made of certain phenomena that can be found among the solutions of a simple nonlinear differential equation. (2) An arbitrary function generator and associated equipment were developed to be used in the solution of more difficult nonlinear equations such as those representing nonlinear servomechanisms. This function generator operates on the logarithmic principle and can be likened to an electronic slide rule. (3) The operation of the function generator was demonstrated by forming a nonlinear computer using the generator and other active and passive circuit elements. This computer was used to investigate the solutions to several modifications of Van der Pol's equation.https://thesis.library.caltech.edu/id/eprint/1644An Analysis of Servomechanisms Containing a Dependent Variable Nonlinearity
https://resolver.caltech.edu/CaltechETD:etd-04212003-105744
Authors: {'items': [{'id': 'Anderson-Jack-Steele', 'name': {'family': 'Anderson', 'given': 'Jack Steele'}, 'show_email': 'NO'}]}
Year: 1953
DOI: 10.7907/SQ9A-3G13
A method is developed for obtaining the transient response of an automatic control system containing a nonlinearity in which one dependent variable may be expressed as a unique function of another variable. This method involves obtaining a mathematical expression for the nonlinear characteristic by an expansion in Legendre polynomials, introducing this expression into the equations describing the control system behavior thus obtaining a nonlinear equation in a power series of a dependent variable and solving this nonlinear equation by means of an assumed infinite series solution technique. The rules governing application of the method are discussed.
A saturation type nonlinearity is used to illustrate the application of the method. A second order system is employed to illustrate the accuracy of the method and to present a numerical technique for solving the series of equations arising from the infinite series method of solving the nonlinear system equation. The stability of a fourth order missile control equation with a saturation limit on the control surface is investigated by the method.
https://thesis.library.caltech.edu/id/eprint/1431Electric Analog Computer Study of Supersonic Flutter of Elastic Delta Wings
https://resolver.caltech.edu/CaltechETD:etd-11252003-111202
Authors: {'items': [{'id': 'Basin-Michael-Abram', 'name': {'family': 'Basin', 'given': 'Michael Abram'}, 'show_email': 'NO'}]}
Year: 1954
DOI: 10.7907/FHFK-QR87
This thesis presents a method for the solution of the supersonic flutter problem for elastic delta wings with supersonic leading edges.
In Part I, the necessary aerodynamic equations are developed, first in integral form, and then in a power series expansion in order to obtain a practical expression for the pressure at a point on the wing due to the motion of the wing surface in a supersonic air stream.
Part II gives a method for computing the lifts on a partitioned wing, and sets up cell division criteria. These methods are then applied to a specific wing form.
Parts III and IV present the electrical analogs for the aerodynamic lifts, and for the elastic wing structure respectively. These analogs are then applied to the example of part II.
Part V presents the results of the actual flutter study performed on the above wing on the California Institute of Technology Electric Analog Computer.
Part VI contains the conclusions of the study, and recommendations and suggestions for further research.https://thesis.library.caltech.edu/id/eprint/4659An Investigation of the Influence of Fuselage Elasticity on the Pitching Motion of a Controlled Missile
https://resolver.caltech.edu/CaltechETD:etd-01142004-104429
Authors: {'items': [{'id': 'Upthegrove-Harry-Nelson', 'name': {'family': 'Upthegrove', 'given': 'Harry Nelson'}, 'show_email': 'NO'}]}
Year: 1954
DOI: 10.7907/B4FA-2254
This thesis describes an investigation of the effects of elastic deformation on the performance of a guided missile type of aircraft. An imaginary typical missile is selected, and an electric model is devised to represent the missile with its associated aerodynamic forces and control functions. The results of measurements on this model are presented in the form of oscillographs, and their significance is discussed.
It is determined that elasticity effects can rarely be ignored, and that improvement in missile and control system performance will probably require more rather than less consideration of elastic deformation of the aircraft structure.https://thesis.library.caltech.edu/id/eprint/159The Dielectric Recovery of Parallel Arcs
https://resolver.caltech.edu/CaltechETD:etd-12122003-092252
Authors: {'items': [{'id': 'Shennum-Robert-Herman', 'name': {'family': 'Shennum', 'given': 'Robert Herman'}, 'show_email': 'NO'}]}
Year: 1954
DOI: 10.7907/G0V3-BW26
A study of the factors which influence the stability of simultaneous, distributed arcs on power lines is reported. The types of arcs studied were through air between iron electrodes, flashover along a wood-air boundary between copper electrodes and arcs confined in expulsion-tube lightning arrestors.
Variations of line geometry influence line characteristic impedance, series and shunt impedances and equivalent length. The studies indicate that of these factors, only the series impedance between the simultaneous arcs is important unless the line length separating the two arcs is very short.
The analysis illustrates that it is possible to calculate approximate critical spacing of a test gap in terms of the geometry and material of arc electrodes and the potential at the gap. This potential was shown to depend upon both the power frequency conduction characteristics of the remainder of the circuit and upon the nature of the surge initiating the arc. The relative importance of these two factors was estimated.
The studies show that it is possible for arcs to rob one another after they have been independently initiated on power distribution systems. Consequently, with appropriate placement of lightning arrestors it is possible to protect a system so that even though a direct stroke of lightning may initiate an arc at a location remote from the arrestors, the arrestors can still rob the open arc of current and hence extinguish it.https://thesis.library.caltech.edu/id/eprint/4952A Study of the Effects of Damping on Normal Modes of Electrical and Mechanical Systems
https://resolver.caltech.edu/CaltechETD:etd-12042003-110450
Authors: {'items': [{'id': 'Crumb-Stephen-Franklin', 'name': {'family': 'Crumb', 'given': 'Stephen Franklin'}, 'show_email': 'NO'}]}
Year: 1955
DOI: 10.7907/YESX-K771
This thesis is a general investigation of some of the properties of free and forced vibrations in linear, non-conservative systems. Particular emphasis is placed upon the problems which arise in normal mode studies made on the electric analog computer at the California Institute of Technology.
In Part I the major problems are defined, and limitations of the study are discussed. Part II is a review of the basic theory of normal modes included primarily to establish familiarity with the notation to be used later. In Part III, modifications of normal mode concepts, as applied to damped systems, are examined. "Small damping" criteria are discussed, and a set of theorems of small damping is presented.
In Part IV a series of normal mode analog circuits for damped systems are developed. Part V is a study of uniform damping, generalizing and extending some of the work of Rayleigh, Bode, and Guillemin. It is shown that for any type of uniform damping, all of the basic normal mode concepts are preserved.
In Part VI the theory of mode separation in uniformly damped systems is considered. Criteria for determining mode frequencies and mode parameters are developed. A multiple drive method of exciting normal modes is proposed. In Part VII, some of the methods of Part VI are extended to non-uniformly damped systems. Equivalent orthogonal systems are proposed to approximate the behavior of systems with moderately non-uniform damping. A quantitative measure of non-uniformity is presented.
In Part VIII, numerical examples and experimental results in support of the theory are presented. Concluding remarks are made in Part IX.
https://thesis.library.caltech.edu/id/eprint/4767An Electronic Digital Polynomial Root Extractor
https://resolver.caltech.edu/CaltechETD:etd-03242004-145118
Authors: {'items': [{'id': 'Johnson-Robert-Royce', 'name': {'family': 'Johnson', 'given': 'Robert Royce'}, 'show_email': 'NO'}]}
Year: 1956
DOI: 10.7907/XSCJ-XT27
Many mathematical techniques exist for factoring algebraic polynomials. Most require much computation and programming and are practical methods only with large machine computers. A special purpose electronic digital computer designed to factor polynomials of high degree is described. The mathematical method is an adaptation of a Taylor serial approximation used to connect the problem and its formulation with a special machine implementation. The computer uses a small rotating magnetic drum, about 200 germanium diodes, and 20 logical flip-flops. Unique features of the system are the simple algebraic logical design techniques and the ease of programming. The result is a small, simple, and useful computer.https://thesis.library.caltech.edu/id/eprint/1092The Measurement of Gas Produced by Electrical Power Arcs in Insulating Oil
https://resolver.caltech.edu/CaltechETD:etd-07142004-113646
Authors: {'items': [{'id': 'Kuo-Tsao-Hwa', 'name': {'family': 'Kuo', 'given': 'Tsao Hwa'}, 'show_email': 'NO'}]}
Year: 1957
DOI: 10.7907/EMS6-8Z18
This thesis presents a study of the factors which influence the gas rate (volume of gas produced per unit arc energy) in insulating oil. Such data are fundamental to the design of oil circuit breakers. It is found that the gas rate varies with electrode length, location of arc on the surface, duration of current and material of arcing surface. The gas rate is higher for the longer gap length (3/4 inch) than for the shorter gap length (1/4 inch). An arc on the center of the surface generally gives a lower gas rate while that on the edge a higher gas rate. For arcs occurring on the center, copper surface has a higher gas rate while steel and silver elkalite (alloy used in practice) surfaces have approximately the same lower rate. A high gas rate results in the short duration arc of 1/2 cycle and nearly equal low gas rates are obtained for durations of 2 cycles and 3 1/2 cycles. No appreciable change in gas rate is observed as the magnitude of current peak varies from 400 to 800 amperes.
With the method used in this research, steel surfaces have the desired character that arcs always occur on the center. For 600 amperes and 2 cycle duration, the gas rate of flat steel surfaces referred to 760 mm and 25[degrees]C is 74, 133 and 175 cubic centimeters per kilowatt second at 1/4, 1/2 and 3/4 inch gap lengths respectively.
Because the gas rate for silver elkalite and steel surfaces are about the same with arcs on the center, the effect of current magnitude is small and the arc duration in most circuit breaker operation about 2 cycles, the total volume of gas produced in actual operation of a circuit breaker can be calculated approximately by correlating the above figures with field test records of arc energy and electrode travel.https://thesis.library.caltech.edu/id/eprint/2877The Logical Design of a Serial General-Purpose Computer with Micro-Program Capabilities
https://resolver.caltech.edu/CaltechETD:etd-01202006-130420
Authors: {'items': [{'id': 'Tanaka-Richard-Isamu', 'name': {'family': 'Tanaka', 'given': 'Richard Isamu'}, 'show_email': 'NO'}]}
Year: 1958
DOI: 10.7907/TE7V-ZX24
This thesis presents the detailed logical design of a serial, general-purpose digital computer with micro-program capabilities. The machine is a medium-speed computer with a magnetic drum memory, and includes a modest magnetic core memory for storing micro-orders.
The micro-program feature relates to the command structure of the machine. Micro-orders enable the programmer to specify many of the elementary internal operations; these micro-orders are arranged into micro-routines which synthesize computation sequences equivalent to commands in more conventional computers. Therefore, effectively, the machine has an extremely long command list.
Included are descriptions of the internal organization of the machine, and the detailed logical equations which are directly applicable to the construction of the computer. A few micro-routines, for both usual and uncommon commands, are also shown, to illustrate possible applications of the machine.https://thesis.library.caltech.edu/id/eprint/254An Investigation of the Use of Feedback Control to Raise the Flutter Speed of an Aeroelastic System
https://resolver.caltech.edu/CaltechETD:etd-10082004-111321
Authors: {'items': [{'id': 'Braham-Harold-Stanley', 'name': {'family': 'Braham', 'given': 'Harold Stanley'}, 'show_email': 'NO'}]}
Year: 1958
DOI: 10.7907/5EVF-4945
This study deals with the possibility of raising the flutter speed of an aeroelastic system by feedback control. Emphasis is given to feedback control through a control surface driven by a powered actuator. It is shown that only for certain aeroelastic systems is it possible to significantly raise the flutter speed by this type of feedback control.
In addition, a second type of feedback control is considered, a jet reaction torquer acting on the wing itself. A marked increase in flutter speed is possible in essentially all aeroelastic systems with this form of control. Despite the fact that this type of controller requires more power than the control surface actuator, its use may be practical in many cases.
The question of the reliability of the feedback control used to raise flutter speed is becoming less important in this era of missiles where the entire system is controlled by automatic devices.
Part I discusses the basic problem. Parts II and III develop a general feedback theory. This theory, when applied to a specific aeroelastic system, permits ready determination of whether an increase in flutter speed is possible for that system. Parts IV and V consider numerical investigations of many systems using a control surface actuator. Analyses are made by both the general feedback theory and by analog computer, showing similar numerical results. Part VI considers the jet reaction torquer and the increase in flutter speed that it can achieve. Conclusions are presented in Part VII.https://thesis.library.caltech.edu/id/eprint/3982A Digital-Computer-Programmed Topological Method of Coordinate Selection for Numerical Computations in an Electrical Network
https://resolver.caltech.edu/CaltechTHESIS:08192011-102613198
Authors: {'items': [{'id': 'Lock-Kenneth', 'name': {'family': 'Lock', 'given': 'Kenneth'}, 'show_email': 'NO'}]}
Year: 1962
DOI: 10.7907/1KC2-ZF72
In this thesis an algorithm is developed for setting up the
differential equations and initial conditions of an electrical network of arbitrarily connected capacitors, resistors, inductors, multiwinding ideal transformers, and ideal voltage and current sources that topologically represents a large class of systems. The algorithm
formulates the equation in a set of coordinates such that all matrices to be inverted are nonsingular. The topological description of the circuit is used to select a nonsingular set of coordinates which enables the computation of the transient responses and the short circuit
admittances to a set of arbitrarily chosen ports of a network. Transformers are accounted for by appropriately selecting a set of dependent variables from the set of transformer linear equations. The algorithm
for selecting a nonsingular set of coordinates, being mainly symbol manipulations, is coded in LISP. It is also shown that the same method may be applied to systems with nonlinear parameter matrices.https://thesis.library.caltech.edu/id/eprint/6604Deposited Loops Coupling Magnetic Films as Fast Computer Elements
https://resolver.caltech.edu/CaltechETD:etd-10302002-083025
Authors: {'items': [{'id': 'Reardon-Bernard-Charles', 'name': {'family': 'Reardon', 'given': 'Bernard Charles'}, 'show_email': 'NO'}]}
Year: 1964
DOI: 10.7907/1FZC-JY74
NOTE: Text or symbols not renderable in plain ASCII are indicated by [...]. Abstract is included in .pdf document.
An experimental study of the fast switching properties of the film-loop assemblies indicated that a line charge model for a rectangular magnetic film represents the external flux distribution of the film with good ([...]10%) accuracy. The flux linkage of a deposited loop equals 100% of the film flux. A 20% decrease in film flux was measured at both ends of the film easy axis and a quadratic flux distribution is consistent with the experimental results. Eddy current and circulating loop current fields affect the nanosecond range switching of magnetic films. Eddy current fields increase rapidly with the fraction of magnetic film perimeter covered by the loop conductors. Capacitive loop current fields cause a small increase in film switching time. Resistive loop loading can slow film switching significantly. The bias and drive field properties of deposited loops can be predicted from the loop dimensions. The deposition of complete film assemblies was effected without opening the vacuum of the system. The film switching test equipment incorporated a new method for the integration of fast switching signals and had a response time of 0.8 nsec. A theoretical study of the coupling loop circuitry indicated that the loop attenuation and the spatial distribution of loop induced voltage distort the film output signal. These effects are small, but not negligible, for a loop with conductor separation of 50,000 A and conductor thicknesses of 4000 A, enclosing a 1 cm[superscript 2] film switching in 1 nsec.
The use of integrated deposited circuitry containing magnetic films is feasible for fast computers. A set of logic elements suitable for a deposited configuration has reasonable fan-in and fan-out potentialities. Theoretical miniaturization limits for deposited film logic circuitry depend chiefly on the resistivity of the coupling loops. Operation of such circuitry at low temperature reduces the limit of film size to about 14 microns. Miniaturization alleviates the effect of loop currents and attenuation on film switching signals, leading to low power, short switching time circuitry.https://thesis.library.caltech.edu/id/eprint/4311The Application of Tunnel Diodes to Switching and Logical Circuits
https://resolver.caltech.edu/CaltechETD:etd-10222002-085219
Authors: {'items': [{'id': "O'Regan-Patrick-Gerard", 'name': {'family': "O'Regan", 'given': 'Patrick Gerard'}, 'show_email': 'NO'}]}
Year: 1964
DOI: 10.7907/ZPA2-X230
No abstract submitted.https://thesis.library.caltech.edu/id/eprint/4200A Threshold Gate Feed-Forward Switching Net Algorithm
https://resolver.caltech.edu/CaltechETD:etd-09202002-135451
Authors: {'items': [{'id': 'Hughes-Gordon-Frierson', 'name': {'family': 'Hughes', 'given': 'Gordon Frierson'}, 'show_email': 'NO'}]}
Year: 1964
DOI: 10.7907/K8X1-1697
A general algorithm is presented for the efficient computation of feed-forward nets of general threshold gates which realize given bi-valued switching functions. A simplified version of the algorithm is presented for the case of symmetric threshold nets which realize symmetric switching functions.
These algorithms produce near-minimal gate nets, and the results of a digital computer program for the general algorithm are presented to illustrate the degree of efficiency and minimality obtained in practice.
Both algorithms are proved to give a minimal one-gate net if one exists for a given switching function; a necessary criterion is given for the symmetric algorithm to produce a minimal two-gate net if one exists; and two-gate minimality is also demonstrated for the general algorithm, for a certain class of two-gate switching functions.
The case of partially defined switching functions is also treated.https://thesis.library.caltech.edu/id/eprint/3649A computer-aided investigation of motion detection units in the fly
https://resolver.caltech.edu/CaltechTHESIS:08202015-141724571
Authors: {'items': [{'id': 'Dill-J-C', 'name': {'family': 'Dill', 'given': 'John Cedric'}}]}
Year: 1970
DOI: 10.7907/06J6-SK55
<p>This work deals with two related areas: processing of visual information in the central nervous system, and the application of computer systems to research in neurophysiology.</p>
<p>Certain classes of interneurons in the brain and optic lobes of the blowfly <u>Calliphora</u> <u>phaenicia</u> were previously shown to be sensitive to the direction of motion of visual stimuli. These units were identified by visual field, preferred direction of motion, and anatomical location from which recorded. The present work is addressed to the questions: (1) is there interaction between pairs of these units, and (2) if such relationships can be found, what is their nature. To answer these questions, it is essential to record from two or more units simultaneously, and to use more than a single recording electrode if recording points are to be chosen independently. Accordingly, such techniques were developed and are described.</p>
<p>One must also have practical, convenient means for analyzing the large volumes of data so obtained. It is shown that use of an appropriately designed computer system is a profitable approach to this problem. Both hardware and software requirements for a suitable system are discussed and an approach to computer-aided data analysis developed. A description is given of members of a collection of application programs developed for analysis of neuro-physiological data and operated in the environment of and with support from an appropriate computer system. In particular, techniques developed for classification of multiple units recorded on the same electrode are illustrated as are methods for convenient graphical manipulation of data via a computer-driven display.</p>
<p>By means of multiple electrode techniques and the computer-aided data acquisition and analysis system, the path followed by one of the motion detection units was traced from open optic lobe through the brain and into the opposite lobe. It is further shown that this unit and its mirror image in the opposite lobe have a mutually inhibitory relationship. This relationship is investigated. The existence of interaction between other pairs of units is also shown. For pairs of units responding to motion in the same direction, the relationship is of an excitatory nature; for those responding to motion in opposed directions, it is inhibitory. </p>
<p>Experience gained from use of the computer system is discussed and a critical review of the current system is given. The most useful features of the system were found to be the fast response, the ability to go from one analysis technique to another rapidly and conveniently, and the interactive nature of the display system. The shortcomings of the system were problems in real-time use and the programming barrier—the fact that building new analysis techniques requires a high degree of programming knowledge and skill. It is concluded that computer system of the kind discussed will play an increasingly important role in studies of the central nervous system. </p>
https://thesis.library.caltech.edu/id/eprint/9108Formal Methods in the Foundations of Science
https://resolver.caltech.edu/CaltechTHESIS:08252015-104226727
Authors: {'items': [{'id': 'Randall-David-Lawrence', 'name': {'family': 'Randall', 'given': 'David Lawrence'}}]}
Year: 1970
DOI: 10.7907/KB5H-4F11
<p>The intent of this study is to provide formal apparatus which facilitates the investigation of problems in the methodology of science. The introduction contains several examples of such problems and motivates the subsequent formalism.</p>
<p>A general definition of a formal language is presented, and this definition is used to characterize an individual’s view of the world around him. A notion of empirical observation is developed which is independent of language. The interplay of formal language and observation is taken as the central theme. The process of science is conceived as the finding of that formal language that best expresses the available experimental evidence.</p>
<p>To characterize the manner in which a formal language imposes structure on its universe of discourse, the fundamental concepts of elements and states of a formal language are introduced. Using these, the notion of a basis for a formal language is developed as a collection of minimal states distinguishable within the language. The relation of these concepts to those of model theory is discussed.</p>
<p>An a priori probability defined on sets of observations is postulated as a reflection of an individual’s ontology. This probability, in conjunction with a formal language and a basis for that language, induces a subjective probability describing an individual’s conceptual view of admissible configurations of the universe. As a function of this subjective probability, and consequently of language, a measure of the informativeness of empirical observations is introduced and is shown to be intuitively plausible – particularly in the case of scientific experimentation.</p>
<p>The developed formalism is then systematically applied to the general problems presented in the introduction. The relationship of scientific theories to empirical observations is discussed and the need for certain tacit, unstatable knowledge is shown to be necessary to fully comprehend the meaning of realistic theories. The idea that many common concepts can be specified only by drawing on knowledge obtained from an infinite number of observations is presented, and the problems of reductionism are examined in this context. </p>
<p>A definition of when one formal language can be considered to be more expressive than another is presented, and the change in the informativeness of an observation as language changes is investigated. In this regard it is shown that the information inherent in an observation may decrease for a more expressive language. </p>
<p>The general problem of induction and its relation to the scientific method are discussed. Two hypotheses concerning an individual’s selection of an optimal language for a particular domain of discourse are presented and specific examples from the introduction are examined. </p>
https://thesis.library.caltech.edu/id/eprint/9116Information Processing by the First Optic Ganglion in Dipterans
https://resolver.caltech.edu/CaltechTHESIS:11212017-105124061
Authors: {'items': [{'id': 'Arnett-David-Woods', 'name': {'family': 'Arnett', 'given': 'David Woods'}}]}
Year: 1971
DOI: 10.7907/VSZT-PE65
<p>Information processing properties of the dipteran first optic
ganglion were studied by observing and analyzing the discharge behavior
of two units in the intermediate chiasma and the slow potential
behavior of two units in the first optic ganglion. Both types of
chiasma units (on-off, on-maintained) were centripetal and corresponded
to second order units in the first optic ganglion. The on-off
unit was characterized by a transient discharge following the onset
and cessation of a light pulse presented anywhere in the receptive
field which had an elliptical configuration with average major and
minor half directional sensitivity angles of 5.5 and 4.3 degrees,
respectively. The receptive field of the on-maintained unit was
composed of three roughly circular regions arranged adjacently
along a line, and stimulation of the center region elicited a sustained
dis charge while stimulation of either adjacent region elicited an
off discharge. The average half directional sensitivity angle of the
center region was 2.5 degrees which compared well with the
acceptance angle of the photo receptors. The orientation of the major
axes of the on-off and on-maintained unit receptive fields was always
that of the medio-lateral axis of the compound eye.</p>
<p>The on and off regions of the on-maintained unit receptive
field were antagonistic, for stimulation of either off region inhibited
or suppressed a discharge resulting from region stimulation.
Furthermore, the off response was inhibited by region stimulation
if the cessation of on region stimulation preceded, by not more than
200 msec., the cessation of off region stimulation.</p>
<p>The discharge patterns of both units were independent of the
stimulus spectral wavelength, and all units studied, of both type,
possessed the same spectral sensitivity which was characterized by
two peaks of approximately equal maximum sensitivity centered at
350 mµ and 485 mµ. It was concluded that these units belonged to the
system served by photoreceptors 1-6 and that a photopigment
with two absorption peaks was responsible for the observed spectral
sensitivities. Neither unit displayed any sensitivity to the plane of
polarized light.</p>
<p>Positive slow potentials were recorded from the first optic
ganglion with fine micropipettes, and they were believed to originate
from the photo receptor axons. A hyperpolarizing slow potential,
most likely of intracellular origin, was also recorded, but its
properties were not studied in detail.</p>
<p>Based on their information processing properties, the on-off
and on-maintained units were identified with the two type I monopolar
neurons of each cartridge. A model was developed which adequately
explained the behavior of both types of units, and its structure was
compared with the known anatomical structure of the first optic
ganglion.</p>
https://thesis.library.caltech.edu/id/eprint/10566Nonlinear dynamic transfer functions for certain retinal neuronal systems
https://resolver.caltech.edu/CaltechTHESIS:02072014-142214067
Authors: {'items': [{'id': 'Marmarelis-P-Z', 'name': {'family': 'Marmarelis', 'given': 'Panos Zissis'}, 'show_email': 'NO'}]}
Year: 1972
DOI: 10.7907/TBKA-2Q74
<p>The applicability of the white-noise method to the identification
of a nonlinear system is investigated. Subsequently, the method
is applied to certain vertebrate retinal neuronal systems and nonlinear,
dynamic transfer functions are derived which describe quantitatively
the information transformations starting with the light-pattern stimulus
and culminating in the ganglion response which constitutes the visually-derived
input to the brain. The retina of the catfish, Ictalurus
punctatus, is used for the experiments.</p>
<p>The Wiener formulation of the white-noise theory is shown to be
impractical and difficult to apply to a physical system. A different
formulation based on crosscorrelation techniques is shown to be applicable
to a wide range of physical systems provided certain considerations
are taken into account. These considerations include the time-invariancy
of the system, an optimum choice of the white-noise input bandwidth,
nonlinearities that allow a representation in terms of a small number
of characterizing kernels, the memory of the system and the temporal
length of the characterizing experiment. Error analysis of the kernel
estimates is made taking into account various sources of error such
as noise at the input and output, bandwidth of white-noise input and
the truncation of the gaussian by the apparatus.</p>
<p>Nonlinear transfer functions are obtained, as sets of kernels,
for several neuronal systems: Light → Receptors, Light → Horizontal,
Horizontal → Ganglion, Light → Ganglion and Light → ERG. The derived
models can predict, with reasonable accuracy, the system response to
any input. Comparison of model and physical system performance showed
close agreement for a great number of tests, the most stringent of
which is comparison of their responses to a white-noise input. Other
tests include step and sine responses and power spectra.</p>
<p>Many functional traits are revealed by these models. Some
are: (a) the receptor and horizontal cell systems are nearly linear
(small signal) with certain "small" nonlinearities, and become faster
(latency-wise and frequency-response-wise) at higher intensity levels,
(b) all ganglion systems are nonlinear (half-wave rectification), (c)
the receptive field center to ganglion system is slower (latency-wise
and frequency-response-wise) than the periphery to ganglion system,
(d) the lateral (eccentric) ganglion systems are just as fast (latency
and frequency response) as the concentric ones, (e) (bipolar response)
= (input from receptors) - (input from horizontal cell), (f) receptive
field center and periphery exert an antagonistic influence on the
ganglion response, (g) implications about the origin of ERG, and many
others.</p>
<p>An analytical solution is obtained for the spatial distribution
of potential in the S-space, which fits very well experimental data.
Different synaptic mechanisms of excitation for the external and
internal horizontal cells are implied.</p>https://thesis.library.caltech.edu/id/eprint/8066Error Recovery in Robot Systems
https://resolver.caltech.edu/CaltechTHESIS:09102018-150127337
Authors: {'items': [{'id': 'Srinivas-Sankaran', 'name': {'family': 'Srinivas', 'given': 'Sankaran'}}]}
Year: 1977
DOI: 10.7907/P7QY-KN31
<p>This dissertation addresses itself to the problem faced by a robot in recovering from failures during execution of a task. Failures occur partly because sensory information is inaccurate, partly because effectors do not always perform as expected, and partly because the domain in which the robot operates cannot be characterized exactly. Robot systems with automated planners have traditionally dealt with the problem of error recovery by merely replanning to achieve the desired goal, without attempting to characterize the failure in any way whatsoever.</p>
<p>The central idea in this thesis is that planning recovery from failures has its own special techniques, distinct from those used in conventional planning systems. Two viewpoints, looking at the past for an explanation of the failure, and looking at the current situation to attempt a characterization of the failure state, provide powerful heuristics for error recovery. This thesis suggests that these heuristics can be formalized as failure reason analysis and multiple outcome analysis, and that knowledge relevant for such analysis can be provided through a failure reason model and a multiple outcome model associated with each action.</p>
<p>The failure reason model about why actions provides a means for representing fail, like bumping into an object to be grasped because of servoing errors or because of inaccurate information about the location of the object. The model also provides knowledge required for distinguishing between the different reasons for failure. Finally, it includes recommendations of corrective actions to be taken if failure is attributed to a specific reason. This model in used in failure reason analysis in building a failure tree representing possible explanations of the failure. The explanations represented in the tree are then used in planning recovery.</p>
<p>The multiple outcome model provides a way of representing the possible outcomes of an action, like bumping onto the object or bumping onto the ground in the immediate vicinity of the object, ignoring the fact that these outcomes could be the result of several different reasons. Knowledge required to distinguish between different outcomes is provided as part of the model. In cases where the immediately available information is inadequate to identify the outcome of an action, the multiple outcome model provides a basis for executing actions to serve as information gathering steps. The novel feature here is that information gathering is directed by specific expectations about the state of the world.</p>
<p>A computer implementation of a program called MEND has provided a medium for exploring the above idea. MEND has been designed to automate recovery from failures in simple manipulation tasks to be performed by the JPL robot, but the techniques used in MEND have greater generality. A first implementation of MEND established the basis of this investigation. A second version, which has been designed to correct some limitations of the first version, has not yet been fully implemented and integrated with the JPL robot system.</p>
<p>The techniques of planning recovery from failures through failure reason analysis and multiple outcome analysis are contributions to the subject of robotics. More importantly, however, the problem of error recovery is recognized to be a member of a larger class of problems involving knowledge representation and common sense reasoning, both of which are core topics in the study of artificial intelligence. The solution presented in this thesis makes some new contributions to these core topics.</p>https://thesis.library.caltech.edu/id/eprint/11180Temporal-spatial function analysis of the class-IIa1 motion detection cell of the fly Calliphora phaenicia
https://resolver.caltech.edu/CaltechETD:etd-10312006-134932
Authors: {'items': [{'id': 'Jin-M-Y', 'name': {'family': 'Jin', 'given': 'Michael Yih-Hwa'}, 'show_email': 'NO'}]}
Year: 1981
DOI: 10.7907/ccc0-ws70
The physiology of the horizontal motion detection cell class-IIa1 of the fly Calliphora phaenicia was studied by extracellular recording. The receptive field was treated as a multi-input single output system such that Volterra-Wiener functional formalism could be used to describe the input-output relation. The present work has three objectives: (1) To study the functional of a subunit called the basic motion detection unit. Interpretation of the kernel's features was obtained by comparing it with the results of transient type stimuli. (2) To investigate the organization of the nonlinear spatial interactions in relation to the hexagonal array of visual elements along the horizontal and vertical lines of symmetry. Two sets of experiments were done to find the difference in the organizations of this interaction under light and dark adapted conditions. (3) To investigate the binocular interaction between two class-IIa1 cells each located in different lobes.
The basic motion detection unit could be modeled by two first and second order self kernels, which are associated with their two channels, and a cross kernel. The modeling response of a single channel is dominated by the contribution from the second order self kernels because the receptive field is of on-off type. The cross kernel predicts the multiplication-like directionally selective motion response. Both white noise and transient type stimuli confirmed that an excitatory effect as well as an inhibitory effect exist when stimulus patterns move in the forward direction and backward direction, respectively. The cross kernel model can be thought of as a correlation model which has delay type linear filters (instead of differential and integration type filters). An inhibition model with four types of inhibitory interactions was developed which adequately explained the function of this basic motion detection unit.
Horizontally and vertically aligned eight-stripe white noise patterns were used to investigate the receptive field organization. Under dark adapted conditions the interaction was limited to six adjacent columns and rows. The horizontal cross kernels reveal directionally selective characteristics and have significant 'weights' for those describing the interactions between one column and the nearest four adjacent columns. The vertical cross kernels show a mutually excitatory effect, which could be the evidence of neural pooling of photoreceptors. Under light adapted conditions two major differences appeared. They are: (1) Only two horizontal cross kernels, which describe the interactions between one column and the nearest two columns, have significant 'weights'. The interactions outside this range show reversed directionally selective characteristics. (2) All horizontal and vertical cross kernels have negative diagonal components which indicate a mutually inhibitory effect. This could be a nonlinear sensitivity control mechanism. All the above results were studied with monocular preparation.
The binocular interaction was studied by stimulus patterns located both within the binocular region and outside it. The results confirmed that a mutually inhibitory relationship exists between the class-IIa1 cell and its mirror image in the opposite lobe.https://thesis.library.caltech.edu/id/eprint/4341A study of horizontal cell-photoreceptor interaction in the frog retina using a randomly modulated stimulus
https://resolver.caltech.edu/CaltechETD:etd-10042006-110912
Authors: {'items': [{'id': 'Melton-R-B', 'name': {'family': 'Melton', 'given': 'Ronald Benjamin'}, 'show_email': 'NO'}]}
Year: 1981
DOI: 10.7907/6dzj-fd89
The intracellular responses of frog horizontal cells and photoreceptors to conventional and randomly modulated stimuli were recorded by intracellular probing using glass micropipettes. The stimuli were designed to test for photoreceptor response compression by surround illumination. The results indicated that the response of photoreceptors is reduced by stimulation of the area surrounding the cell's receptive field. It appeared that the effect is generated by negative feedback from horizontal cells to both rods and cones. These findings are in good agreement with the results of earlier studies by other investigators who established that horizontal cells feed back to cone photoreceptors in many vertebrate species.https://thesis.library.caltech.edu/id/eprint/3917