Off-campus WSU users: To download campus access dissertations, please use the following link to log into our proxy server with your WSU access ID and password, then click the "Off-campus Download" button below.
Non-WSU users: Please talk to your librarian about requesting this dissertation through interlibrary loan.
Access Type
WSU Access
Date of Award
January 2021
Degree Type
Dissertation
Degree Name
Ph.D.
Department
Electrical and Computer Engineering
First Advisor
HARPREET SINGH
Abstract
With the advancement of internet and web technologies, there is an increasing interest in the development of intelligent systems [92]. One of the major components of the intelligent systems is the fuzzy logic, which has generated huge interest in a variety of applications. During the last several decades, there have been a number of offshoots of fuzzy logic, and these are applied in all major disciplines of engineering and science. The development of neural network [73] and neuro-fuzzy logic [93] contributed significantly to the design of a large number of real-life applications. During the last few years, the hierarchical fuzzy logic [72][76][78][79] has come out as one of the major contributors in this venture. Hierarchical fuzzy logic is the representation of a fuzzy logic, where fuzzy systems are connected in the form of a hierarchy. The advantage of such a representation is the reduction of the complexity and dimensionality of the systems. This dissertation aims at developing algorithms for hierarchical fuzzy logic and their applications in the areas of engineering and science. In this work, the algorithm for the multi-input multi-output systems using hierarchical fuzzy logic has been developed. The procedure was explained with the help of case studies. The simulation of results of hierarchical fuzzy logic are given. Image processing has become one of the major research areas because of its applications in different fields such as engineering, bio-medical, defense etc. With the expansion of social media, the applications of image processing have reached significant heights. A large number of researchers have contributed to the area of advanced image processing using deep and convolution neural networks [73]. Deep neural networks however do not consider uncertainty and imprecision, which is a major constraint for the intelligent systems [92]. Hierarchical fuzzy logic [72][76][78][79] considers uncertainty and imprecision behavior, and acts as the backbone of intelligent systems [92]. An algorithm especially suitable for image processing applications, keeping into account uncertainty and imprecision has been developed in this dissertation. The algorithm is tested with the help of real-life image datasets. These datasets, such as MNIST, YaleB etc., are available on social websites for research purposes only. The algorithm has been developed to handle these large image datasets with the help of hierarchical fuzzy systems. The results then compared to CNN model extracted from MatConvNet (MATLAB Toolbox). We further modified these image datasets and inserted noise such as gaussian noise. We further validate the accuracy of the algorithm on these noisy image datasets. Both the multi-input multi-output systems and the image processing systems described in this work are in fact a representation of classification of images and data. In general, the classification of data has been described so that it has only one specific output. In this dissertation, the classification in the form of multi-layered architecture has been described. Survivability has been a topic of research especially in the area of defense and security for the last several years. A large number of models has been proposed from time to time for survivability and its applications. Onion model for survivability is considered one of the most significant models for defense applications. In this dissertation, the hierarchical fuzzy representation for the survivability onion model has been given and discussed. This model consists of several layers, given in the form of onion. This has led to the new development of the multi-layer multi-input multi-output hierarchical model for the survivability. The problem of survivability is very complex in nature. It can involve different types of data and information such as non-linear behavior representation, image processing and video processing applications, human assessment and interpretation etc. Because of the complexity and the non-availability of data, the algorithms presented and developed are yet to be implemented. It is hoped that the work done in this dissertation in the area of multi-input multi-output hierarchical systems will result in a large number of forthcoming applications, especially in the area of survivability for defense and security.
Recommended Citation
Kamthan, Shashank, "Hierarchical Intelligent Systems And Their Applications To Survivability" (2021). Wayne State University Dissertations. 3419.
https://digitalcommons.wayne.edu/oa_dissertations/3419