Browsing by Subject "Expert systems"
Now showing 1 - 4 of 4
Results Per Page
Sort Options
Item Concurrency modeling extensions to the Fusion development methodology(Texas Tech University, 1997-05) Wenzel, Peter W.The "Fusion" software development methodology is a self-claimed second-generation full-coverage development method for object-oriented software covering the traditional analysis, design, and implementation phases as well as providing management tools for software development. Fusion's deficiency is its lack of support for concurrency modeling which is essential in the problem domains of all real-time systems. With this one exception. Fusion is an excellent example of a fully integrated object-oriented development methodology, combining the best of several first-generation object-oriented analysis and design (OOAD) methods. The Fusion development methodology may be extended by integrating concurrency modeling into the method, making it more suitable for real-time problem domains. The goals of this thesis are threefold: (1) identify the requirements for modeling concurrency in object-oriented systems, (2) propose extensions to the Fusion object-oriented method for modeling concurrency, and (3) demonstrate the proposed concurrency modeling extensions via a case study. The thesis identifies basic object-oriented concurrency modeling requirements by examining existing concurrency modeling techniques. These requirements are then used to form highly integrated concurrency modeling extensions to the Fusion object-oriented development methodology. Finally, the Fusion concurrency modeling extensions are demonstrated using the telecommunications real-time problem domain of cellular digital packet data (CDPD).Item Control of ball and beam with neural networks(Texas Tech University, 1996-05) Eaton, Paul H.The ball-and-beam problem is a benchmark for testing new control algorithms. In the Worid Congress On Neural Networks, 1994, Prof Lotfi Zadeh proposed a more difficult version which he claimed required a fuzzy logic controller. This experiment uses a beam, partially covered with a sticky substance, increasing the difficulty of predicting the ball's motion. We complicated the problem even more by not using any information concerning the ball's velocity. Although it is common to use the first differences of the ball's consecutive positions as a measure of velocity and explicit input to the controller, we preferred to exploit recurrent neural networks inputting only consecutive positions instead. We have used truncated backpropagation through time with the Node-Decoupled Extended Kalman Filter (NDEKF) algorithm to update the weights in the networks. The neurocontroller uses a form of approximate dynamic programming called an adaptive critic design. A hierarchy of such designs exists. Our system uses Dual Heuristic Programming (DHP), an upper-level design. To our best knowledge, our results are the first use of DHP to control a physical system. It is also the first system we know of to meet Zadeh's challenge.Item Evolution of expert systems(Texas Tech University, 1993-12) Culebro, Joaquín Marcos PalaciosExpert systems are computer programs for providing expertise emulative of that which might be expected from human experts in solving complex problems for which analytical solutions are not available. Evolution of an expert system refers to the initial development of the system and its continuing modification in order to improve its performance. Any modifications made to an expert system have the potential of producing undesirable logical errors and side-effects that are difficult to find or prevent. Although much research has focused on facilitating the evolution of expert systems, most of the limitations still exist. This dissertation proposes an approach for structuring and evolving expert systems for applications in which the provision of the desired expertise is beyond the reach of either analytical or traditional heuristic approaches, but in which the knowledge domain is causally connected and the relevant causality can be expressed in procedural form. The research vehicle used is that of a hypothetical manufacturing system in which products of different types use some of the same workstations, and some of the product types loop back to workstations that they have previously used. The expertise sought is that of scheduling starts of products into the first stage of production so as to yield a stream of output that satisfies a user-specified balance among a variety of business performance measures including timeliness of production output.Item Neural networks and evolutionary computation for real-time quality control(Texas Tech University, 1997-05) Patro, SanjuktaQuality control in general and automated quality control in particular are assuming major importance in modem society as technological SNStems are becoming increasingly complex and highly interconnected. Traditional statistical process control techniques are inadequate to address control problems in automated processes because of the high degree of data correlation characterized by such processes. Classical process control methods depend on simplifying assumptions of plant linearity and time-invariance to make the problem analytically tractable. They are therefore limited in effectiveness of the control of complex, nonlinear, multivariable processes. This dissertation attempts to overcome some of the limitations and shortcomings of traditional quality control methods through the integration of two technologies, neural networks and evolutionary computation. An autonomous control system prototype has been developed to control (maintain quality variables within desired limits) a process by providing high level adaptation to changes in the plant, environment, and control objectives. This technology utilizes memory and learning techniques to overcome limitations of traditional control methods, namely data autocorrelation, requirements of simplifying assumptions, and requirements of a priori information about the process. The robustness and applicability of this integrated technology is demonstrated though results obtained from tests involving simulated processes of varying degrees of complexity.