Browsing by Author "VanLandingham, Hugh F."
Now showing 1 - 20 of 153
Results Per Page
Sort Options
- Adaptation and Installation of a Robust State Estimation Package in the Eef UtilityChapman, Michael Addison (Virginia Tech, 1999-02-08)Robust estimation methods have been successfully applied to the problem of power system state estimation in a real-time environment. The Schweppe-type GM-estimator with the Huber psi-function (SHGM) has been fully installed in conjunction with a topology processor in the EEF utility, headquartered in Fribourg, Switzerland. Some basic concepts of maximum likelihood estimation and robust analysis are reviewed, and applied to the development of the SHGM-estimator. The algorithms used by the topology processor and state estimator are presented, and the superior performance of the SHGM-estimator over the classic weighted least squares estimator is demonstrated on the EEF network. The measurement configuration of the EEF network has been evaluated, and suggestions for its reinforcement have been proposed.
- Adaptive Control using IIR Lattice FiltersHevey, Stephen J. (Virginia Tech, 1998-04-24)This work is a study of a hybrid adaptive controller that blends fixed feedback control and adaptive feedback control techniques. This type of adaptive controller removes the requirement that information about the disturbance is known apriori. Additionally, the control structure is implemented in such a way that as long as the adaptive controller is stable during adaptation, the system consisting of the controller and plant remain stable. The objective is to design and implement an adaptive controller that damps the structural vibrations induced in a multi-modal structure. The adaptive controller utilizes an adaptive infinite impulse response lattice filter for improved damping over the fixed feedback controller alone. An adaptive finite impulse response LMS filter is also implemented for comparison of the ability for both algorithms to reject harmonic, narrow bandwidth and wide bandwidth disturbances. It is demonstrated that the lattice filter algorithm performs slightly better than the LMS filter algorithm in all three disturbance cases. The lattice filter also requires less than half the order of the LMS filter to get the same performance.
- Adaptive optimal control of AC/DC systemsRostamkolai, Niusha (Virginia Polytechnic Institute and State University, 1986)The dissertation presents a new control strategy for two terminal HVDC systems embedded in an AC network. The control is based upon real-time measurements performed on the AC/DC system. Use is made of a technique for high speed accurate measurement of positive sequence voltages and currents, first developed in the field of computer relaying. The real-time measurements provides a term in the control law to compensate for inaccuracies following departure from the operating point. The control criterion is to damp out the electromechanical oscillations following a disturbance. The main contribution of the dissertation is to describe a new optimal controller formulation which contains a measurement based component. Optimal controllers are commonly constructed using linearized equations of the system around the operating point. In DC systems this approach is of a very limited value because of a highly nonlinear nature of the system. With the controller developed in this dissertation, it becomes possible to describe the system as a nonlinear dynamic system. The approximation resulting from the usual linearization of the system equations is thus avoided, and leads to a better controller design. The control technique is illustrated with a small AC/DC system. However, the equations formulated are sufficiently general, so that the technique can be applied to a larger system. Simulation results are included to represent the effectiveness of the developed controller.
- Adaptive Predictive Feedback Techniques for Vibration ControlEure, Kenneth W. II (Virginia Tech, 1998-02-03)In this dissertation, adaptive predictive feedback control is used to suppress plate vibrations. The adaptive predictive controller consists of an on-line identification technique coupled with a control scheme. Various system identification techniques are investigated and implemented including batch least squares, projection algorithm, and recursive least squares. The control algorithms used include Generalized Predictive Control and Deadbeat Predictive Control. This dissertation combines system identification and control to regulate broadband disturbances in modally-dense structures. As it is assumed that the system to be regulated is unknown or time varying, the control schemes presented in this work have the ability to identify and regulate a plant with only an initial estimate of the system order. In addition, theoretical development and experimental results presented in this work confirm the fact that an adaptive controller operating in the presence of disturbances will automatically incorporate an internal noise model of the disturbance perturbing the plant if the system model order is chosen sufficiently large. It is also shown that the adaptive controller has the ability to track changes in the disturbance spectrum as well as track a time varying plant under certain conditions. This work presents a broadband multi-input multi-output control scheme which utilizes both the DSP processor and the PC processor in order to handle the computational demand of broadband regulation of a modally-dense plant. Also, the system identification technique and the control algorithm may be combined to produce a direct adaptive control scheme which estimates the control parameters directly from input and output data. Experimental results for various control techniques are presented using an acoustic plant, a rectangular plate with clamped boundary conditions, and a time varying plate.
- Adaptive Self-Tuning Neuro Wavelet Network ControllersLekutai, Gaviphat (Virginia Tech, 1997-03-31)Single layer feed forward neural networks with hidden nodes of adaptive wavelet functions (wavenets) have been successfully demonstrated to have potential in many applications. Yet applications in the process control area have not been investigated. In this paper an application to a self-tuning design method for an unknown nonlinear system is presented. Different types of frame wavelet functions are integrated for their simplicity, availability, and capability of constructing adaptive controllers. Infinite impulse response (IIR) recurrent structures are combined in cascade to the network to provide a double local structure resulting in improved speed of learning. In particular, neuro-based controllers assume a certain model structure to approximate the system dynamics of the "unknown" plant and generate the control signal. The capability of neuro-controllers to self-tuning of an unknown nonlinear plants is then illustrated through design examples. Simulation results demonstrate that the self-tuning design methods are directly applicable for a large class of nonlinear control systems.
- Aeroelastic modeling and flutter control in aircraft with low aspect ratio composite wingsMorris, Russell A. (Virginia Tech, 1996)A comprehensive study including modeling and control of aeroelastic instabilities in free flying aircraft with flexible wings has been completed. The structural model of the wing consists of a trapezoidal composite plate rigidly attached to a fuselage with rigid-body degrees of freedom. Both quasi-steady and quasi-static aerodynamic strip theories were used to analyze several different flutter mechanisms for a variety of low aspect ratio wing configurations. The most critical flutter mechanism was found to be body-freedom flutter, a coupling of aircraft pitching and wing bending motions, for wings in a forward-sweep configuration. In addition, a modal approximation to the flutter eigenvalue problem was used to substantially reduce computation cost, making the resulting model very attractive for use in larger multiobjective design packages. Composite ply angle tailoring was investigated as a passive method of increasing the body-freedom flutter airspeed of an aircraft model. In addition, wing mounted piezoelectric sensor and induced-strain actuator patches were used in conjunction with active feedback control laws to increase the airspeed at which body-freedom flutter occurs. Two control laws were tested, coupled and independent modal position feedback, to delay frequency coalescence and thus increase the flutter airspeed.
- Aircraft cruise performance optimization using chattering controlsBhardwaj, Pradeep (Virginia Tech, 1986-07-05)Aircraft Cruise Performance is examined by using energy-state modelling to investigate fuel-range optimal trajectories. Chattering controls are considered appropriate when the hodograph is non-convex. Classical steady-state cruise, simple chattering-cruise and the extended chattering-cruise models are studied as constrained parameter-optimization problems. The term "extended chattering" refers to vehicle system modelling extended to maintain vertical equilibrium only on the average. Numerical solution is obtained using a variable-metric gradient-protection algorithm and computational results are presented for three different aircraft. This study shows that simple chattering cruise for certain specific energies can result in substantial fuel savings over classical steady-state cruise. However extended chattering cruise results in only marginal fuel savings when compared to simple chattering cruise.
- Analysis and design of a novel controller architecture and design methodology for speed control of switched reluctance motorsJackson, Terry W. (Virginia Tech, 1996-07-05)This paper presents a novel controller architecture and speed control design methodology suitable for low cost, low performance switched reluctance motor drives. By utilizing inexpensive components in a simple, compact architecture, a low cost controller is developed which achieves a performance level similar to many high performance controllers. A speed control design methodology is established and analyzed based on the linearized small signal model of the switched reluctance motor. This unique control methodology is simple and provides a starting point for further research into speed/current controller parameter design for switched reluctance motors. The analysis, design and realization of the speed controller are presented. The derivation of the design methodology for speed controlled, switched reluctance motor drives is discussed, along with computer simulations for verification. Experimental results utilizing the proposed architecture and design methodology verify the control design and performance capabilities of the speed controller system.
- Application of Bifurcation Theory to Subsynchronous Resonance in Power SystemsHarb, Ahmad M. (Virginia Tech, 1996-12-16)A bifurcation analysis is used to investigate the complex dynamics of two heavily loaded single-machine-infinite-busbar power systems modeling the characteristics of the BOARDMAN generator with respect to the rest of the North-Western American Power System and the CHOLLA# generator with respect to the SOWARO station. In the BOARDMAN system, we show that there are three Hopf bifurcations at practical compensation values, while in the CHOLLA#4 system, we show that there is only one Hopf bifurcation. The results show that as the compensation level increases, the operating condition loses stability with a complex conjugate pair of eigenvalues of the Jacobian matrix crossing transversely from the left- to the right-half of the complex plane, signifying a Hopf bifurcation. As a result, the power system oscillates subsynchronously with a small limit-cycle attractor. As the compensation level increases, the limit cycle grows and then loses stability via a secondary Hopf bifurcation, resulting in the creation of a two-period quasiperiodic subsynchronous oscillation, a two-torus attractor. On further increases of the compensation level, the quasiperiodic attractor collides with its basin boundary, resulting in the destruction of the attractor and its basin boundary in a bluesky catastrophe. Consequently, there are no bounded motions. When a damper winding is placed either along the q-axis, or d-axis, or both axes of the BOARDMAN system and the machine saturation is considered in the CHOLLA#4 system, the study shows that, there is only one Hopf bifurcation and it occurs at a much lower level of compensation, indicating that the damper windings and the machine saturation destabilize the system by inducing subsynchronous resonance. Finally, we investigate the effect of linear and nonlinear controllers on mitigating subsynchronous resonance in the CHOLLA#4 system . The study shows that the linear controller increases the compensation level at which subsynchronous resonance occurs and the nonlinear controller does not affect the location and type of the Hopf bifurcation, but it reduces the amplitude of the limit cycle born as a result of the Hopf bifurcation.
- Application of the systems engineering approach to the conversion of ocean surveillance vessels into hydrographic survey, buoy tending, and general oceanography missions for the National Oceanic and Atmospheric AdministrationIzenson, Shawn M. (Virginia Tech, 1993-12-14)The T-AGOS 13 Class is a suitable platform to meet all mission requirements for both NOAA Conversion designs. The T-AGOS 13 class vessels cannot, however, match the existing NOAA medium endurance charting ship (Fairweather class) in both speed and number of accommodations.
- Applications of phasor measurements to the real-time monitoring of a power systemBarber, David Edward (Virginia Tech, 1994-03-05)This thesis discusses applications of phasor measurement units to power system monitoring and synchronous generator modeling. Adjustments to a previously developed PMU placement algorithm are described which observe generator and tie line flows explicitly and reduces the number of PMUs required for a system, still observing the major dynamic components of a system. This adjusted methodology leaves some buses unobserved. A method for estimating the state of the unobserved region is developed based on using constant admittance or constant current load models. These models are accurate for a small neighborhood around the operating point when they were calculated. To determine the maximum error expected for any given system estimate, an equation relating the maximum error in the voltages to the maximum change in load power is derived. Once the issue of power system monitoring has been presented, the application of PMUs to the synchronous generator modeling is explored. This thesis deals with the on-line identification of the generator transient model using a recursive version of the generalized least squares algorithm. Simulations have been performed to demonstrate the validity and difficulties with these methods.
- An approach to a robust speaker recognition systemTran, Michael (Virginia Tech, 1994)This dissertation presents a design of a robust, automatic speaker recognition (ASR) system. The ASR system is designed to work with both text-independent and text-dependent speaker recognition. Several speaker spectral features are studied to determine their contribution in term of accuracy to the system. A new algorithm is designed to label a speaker voice as either male-type voice or female-type voice. Following this division, the processing time of the speaker identification for the ASR system will be reduced by about half. Rectangular window, Hamming window, first order preemphasis filter, and many proposed spectral distances are also investigated. The principal components analysis is used to achieve high degree of female-type and male-type separation as well as the speaker recognition accuracy. Spectral features are combined to improve the recognition performance of the system. In addition, many other system components such as speech endpoint detection, automatic noise thresholds, etc. are required to build correctly in order to achieve high speaker recognition accuracy. Multi-stage decision process is used both to improve and to speed up the decision if certain criteria are met. Finally, TIMIT acoustic continuous speech corpus is used to evaluate the speaker recognition performance and the robustness of the system.
- Architecture design and simulation for distributed learning classifier systemsGaff, Douglas G. (Virginia Tech, 1995-05-05)In this thesis, we introduce the Distributed Learning Classifier System (DLCS) as a novel extension of J. H. Holland's standard learning classifier system. While the standard LCS offers effective real-time control and learning, one of its limitations is that it does not provide a mechanism for allowing communication between LCS agents in a multiple-agent scenario. Often multiple-agents are used to solve large tasks collectively by subdividing the task into smaller parts. Multiple agents can also be used to solve a task in parallel so that a solution can be arrived at more rapidly. With the DLCS, we introduce mechanisms that satisfy both of these cases, while still providing compatible operation with the LCS. We introduce three types of messages that can be passed between DLCS agents. The first, the classifier message, allows agents to share learned information with one another, thereby helping agents benefit from each other's successes. The second, the action message, allows agents to "talk" to one another. The third, the bucket brigade algorithm payoff message, extends the chain rewarding payoff scheme of the standard LCS to multiple DLCS agents. Finally, we present some simulation results for both the standard LCS and the DLCS. Our LCS simulations examine some of the important aspects of learning classifier system operation, as well as illustrate some of the shortcomings. The DCLS simulations justify the distributed architecture and suggest future directions for achieving learning among multiple agents.
- Artificial Intelligence Applications in the Diagnosis of Power Transformer Incipient FaultsWang, Zhenyuan (Virginia Tech, 2000-08-08)This dissertation is a systematic study of artificial intelligence (AI) applications for the diagnosis of power transformer incipient fault. The AI techniques include artificial neural networks (ANN, or briefly neural networks - NN), expert systems, fuzzy systems and multivariate regression. The fault diagnosis is based on dissolved gas-in-oil analysis (DGA). A literature review showed that the conventional fault diagnosis methods, i.e. the ratio methods (Rogers, Dornenburg and IEC) and the key gas method, have limitations such as the "no decision" problem. Various AI techniques may help solve the problems and present a better solution. Based on the IEC 599 standard and industrial experiences, a knowledge-based inference engine for fault detection was developed. Using historical transformer failure data from an industrial partner, a multi-layer perceptron (MLP) modular neural network was identified as the best choice among several neural network architectures. Subsequently, the concept of a hybrid diagnosis was proposed and implemented, resulting in a combined neural network and expert system tool (the ANNEPS system) for power transformer incipient diagnosis. The abnormal condition screening process, as well as the principle and algorithms of combining the outputs of knowledge based and neural network based diagnosis, were proposed and implemented in the ANNEPS. Methods of fuzzy logic based transformer oil/paper insulation condition assessment, and estimation of oil sampling interval and maintenance recommendations, were also proposed and implemented. Several methods of power transformer incipient fault location were investigated, and a 7Ã 21Ã 5 MLP network was identified as the best choice. Several methods for on-load tap changer (OLTC) coking diagnosis were also investigated, and a MLP based modular network was identified as the best choice. Logistic regression analysis was identified as a good auditor in neural network input pattern selection processes. The above results can help developing better power transformer maintenance strategies, and serve as the basis of on-line DGA transformer monitors.
- Atomic Clock Augmentation For Receivers Using the Global Positioning SystemKline, Paul A. (Virginia Tech, 1997-02-07)For receivers using the Global Positioning System (GPS), it is standard procedure to treat the receiver clock bias from GPS time as an unknown. This requires four range measurements to the satellites in order to solve for three dimensional position and clock offset. If the receiver clock could be synchronized with GPS time, the extra range measurement would not be necessary. To achieve this synchronization, a stable frequency reference must be incorporated into the GPS user set. This concept is known as clock aiding or clock augmentation of GPS receivers. Clock augmentation increases the availability of the navigation function because only three GPS satellites are required. Also, it is shown that clock augmentation improves vertical accuracy by reducing the vertical dilution of precision (VDOP), which is a unitless multiplier that translates range measurement error into vertical position error. This improvement in vertical accuracy is particularly beneficial for applications involving final approach and landing of aircraft using GPS, because GPS typically provides better horizontal accuracy than vertical accuracy. The benefits of atomic clock augmentation are limited by factors that cause a loss of synchronization either between the receiver and GPS time, or between ground station and airborne receivers processing GPS data in differential mode (DGPS). Among the error sources that cause a clock offset are antenna rotation, hardware drifts due to temperature variations, and relativistic effects for GPS receivers on moving platforms. Antenna rotation and temperature effects are addressed and supported by experimental data. It is shown that two particular relativity terms thought to be missing from GPS receiver algorithms are not evident in data collected during a flight test experiment. Upon addressing the error sources, the dissertation concludes with analysis of DGPS data collected during a flight test at the Federal Aviation Administration (FAA) Tech Center in Atlantic City, during which external rubidium oscillators were used by airborne (Boeing 757-B) and ground station GPS receivers. A new method of clock modeling is introduced, and this clock model is used to demonstrate the improvement in vertical accuracy, as well as three-satellite navigation.
- Automatic Ultrasonic Headway Control for a Scaled Robotic CarHenry, Richard Douglas (Virginia Tech, 2001-12-18)Intelligent Transportation Systems and supporting technologies have been an active area of research for some time. Human drivers exhibit slower response times and errors in judgment that can have serious adverse affects on traffic flow. These types of errors can be reduced or eliminated from the driving experience by introducing computer control systems into the automotive arena. The purpose of this research was to develop a scale model platform for the rapid prototyping and testing of ITS systems and technologies. Specifically, this body of work was concerned with the development of an automatic headway control system that utilized ultrasonic sensors. This control system was intended to automatically maintain headway distance in an effort to create an adaptive cruise control system for this scale model vehicle. Implementation of such systems could conceivably reduce driver fatigue by removing the burden of maintaining safe following distance from the driver. System dynamics of car-like robots with nonholonomic constraints were employed in this research to create a controller for an autonomous path following vehicle. The application of a working kinematic model describing car-like robotic systems allowed the development of a simple first order controller, as well as a sliding mode controller. Following the development and simulation of these two control laws, the system was applied to the FLASH project scale model vehicle to assess the practical use of the system on a mock highway. A satisfactory result is produced after testing was completed, and the application of such systems to scale model platforms is feasible.
- Autonomous tactile object exploration and estimation using simple sensorsHollinger, James G. (Virginia Tech, 1994-05-05)In order for robots to become more useful they must be able to adapt and operate in foreign or unpredictable environments. The goal of this thesis is to present an algorithm that will enable a robot to autonomously explore its environment by touch and then estimate the shape of objects it encounters. To demonstrate the feasibility and functionality of such an algorithm, it was fully implemented on a MERLIN 6540 industrial robot. A unique compliant end-effector (consisting of a trackball mounted to a force/torque sensor on a sliding mechanism) and a fuzzy logic force controller were developed to overcome the difficulties inherent in force control on a stepper motor robot. A Kalman filter based quadric shape estimator was then used to describe the objects encountered in the MERLIN's workspace. The minimization of a cost function based on the shape estimator's uncertainty guided the robot along an exploration trajectory designed to produce the fastest converging shape estimate. Results of various exploration trials using autonomous and preprogrammed trajectories are presented. In addition to shape estimates, surface curvature measurements were also obtained. The unique end-effector that provided compliance for the force controller was also able to measure the arc length traversed on the object's surface. Arc length combined with surface orientation makes it possible to determine local surface curvature.
- A backprogagation neutral network in an address block classifiction systemGrzech, Matthew Phillip (Virginia Tech, 1991-05-15)The U.S. Postal Service (USPS) is investing heavily in research and development of automated mail handling systems. A major component in these systems is the use of Optical Character Recognition (OCR) to read the destination address and ZIP Code, and then bar code the mail piece. High speed sorting equipment can then sort the mail using the bar code. Current USPS OCR/automated mail handling systems only process letter mail (no automated address-reading systems exist for nonletter mail). Moreover, these OCR systems only capture and read a restricted field-of-view image. Letters can be rejected by these OCR systems because of nonstandard address location (outside the field-of-view), skewed address lines, or handwritten addresses. Current research is working toward building OCR systems capable of processing all forms of mail which include letters, flats, and irregular parcel and pieces (IPPs). These systems must scan an entire mail image for the destination address block which can assume any orientation. For nonletter mail, such as magazines, this is an exceedingly difficult task, since the entire face of up to 11 by 14 inches must be searched, and the address block must be chosen from all the other extraneous nonaddress information. This paper details an experimental address block location system developed at MITRE. The" system uses a backpropagation neural network trained to discriminate the frequency characteristics of address blocks from other candidates. The current system is trained on magazine flat mail.
- A Bayesian Network Approach to the Self-organization and Learning in Intelligent AgentsSahin, Ferat (Virginia Tech, 2000-08-25)A Bayesian network approach to self-organization and learning is introduced for use with intelligent agents. Bayesian networks, with the help of influence diagrams, are employed to create a decision-theoretic intelligent agent. Influence diagrams combine both Bayesian networks and utility theory. In this research, an intelligent agent is modeled by its belief, preference, and capabilities attributes. Each agent is assumed to have its own belief about its environment. The belief aspect of the intelligent agent is accomplished by a Bayesian network. The goal of an intelligent agent is said to be the preference of the agent and is represented with a utility function in the decision theoretic intelligent agent. Capabilities are represented with a set of possible actions of the decision-theoretic intelligent agent. Influence diagrams have utility nodes and decision nodes to handle the preference and capabilities of the decision-theoretic intelligent agent, respectively. Learning is accomplished by Bayesian networks in the decision-theoretic intelligent agent. Bayesian network learning methods are discussed intensively in this paper. Because intelligent agents will explore and learn the environment, the learning algorithm should be implemented online. None of the existent Bayesian network learning algorithms has online learning. Thus, an online Bayesian network learning method is proposed to allow the intelligent agent learn during its exploration. Self-organization of the intelligent agents is accomplished because each agent models other agents by observing their behavior. Agents have belief, not only about environment, but also about other agents. Therefore, an agent takes its decisions according to the model of the environment and the model of the other agents. Even though each agent acts independently, they take the other agents behaviors into account to make a decision. This permits the agents to organize themselves for a common task. To test the proposed intelligent agent's learning and self-organizing abilities, Windows application software is written to simulate multi-agent systems. The software, IntelliAgent, lets the user design decision-theoretic intelligent agents both manually and automatically. The software can also be used for knowledge discovery by employing Bayesian network learning a database. Additionally, we have explored a well-known herding problem to obtain sound results for our intelligent agent design. In the problem, a dog tries to herd a sheep to a certain location, i.e. a pen. The sheep tries to avoid the dog by retreating from the dog. The herding problem is simulated using the IntelliAgent software. Simulations provided good results in terms of the dog's learning ability and its ability to organize its actions according to the sheep's (other agent) behavior. In summary, a decision-theoretic approach is applied to the self-organization and learning problems in intelligent agents. Software was written to simulate the learning and self-organization abilities of the proposed agent design. A user manual for the software and the simulation results are presented. This research is supported by the Office of Naval Research with the grant number N00014-98-1-0779. Their financial support is greatly appreciated.
- Biologically Inspired Modular Neural NetworksAzam, Farooq (Virginia Tech, 2000-05-19)This dissertation explores the modular learning in artificial neural networks that mainly driven by the inspiration from the neurobiological basis of the human learning. The presented modularization approaches to the neural network design and learning are inspired by the engineering, complexity, psychological and neurobiological aspects. The main theme of this dissertation is to explore the organization and functioning of the brain to discover new structural and learning inspirations that can be subsequently utilized to design artificial neural network. The artificial neural networks are touted to be a neurobiologicaly inspired paradigm that emulate the functioning of the vertebrate brain. The brain is a highly structured entity with localized regions of neurons specialized in performing specific tasks. On the other hand, the mainstream monolithic feed-forward neural networks are generally unstructured black boxes which is their major performance limiting characteristic. The non explicit structure and monolithic nature of the current mainstream artificial neural networks results in lack of the capability of systematic incorporation of functional or task-specific a priori knowledge in the artificial neural network design process. The problem caused by these limitations are discussed in detail in this dissertation and remedial solutions are presented that are driven by the functioning of the brain and its structural organization. Also, this dissertation presents an in depth study of the currently available modular neural network architectures along with highlighting their shortcomings and investigates new modular artificial neural network models in order to overcome pointed out shortcomings. The resulting proposed modular neural network models have greater accuracy, generalization, comprehensible simplified neural structure, ease of training and more user confidence. These benefits are readily obvious for certain problems, depending upon availability and usage of available a priori knowledge about the problems. The modular neural network models presented in this dissertation exploit the capabilities of the principle of divide and conquer in the design and learning of the modular artificial neural networks. The strategy of divide and conquer solves a complex computational problem by dividing it into simpler sub-problems and then combining the individual solutions to the sub-problems into a solution to the original problem. The divisions of a task considered in this dissertation are the automatic decomposition of the mappings to be learned, decompositions of the artificial neural networks to minimize harmful interaction during the learning process, and explicit decomposition of the application task into sub-tasks that are learned separately. The versatility and capabilities of the new proposed modular neural networks are demonstrated by the experimental results. A comparison of the current modular neural network design techniques with the ones introduced in this dissertation, is also presented for reference. The results presented in this dissertation lay a solid foundation for design and learning of the artificial neural networks that have sound neurobiological basis that leads to superior design techniques. Areas of the future research are also presented.