Indiana University-Purdue University Indianapolis (IUPUI) / Depth estimation is increasingly becoming more important in computer vision. The requirement for autonomous systems to gauge their surroundings is of the utmost importance in order to avoid obstacles, preventing damage to itself and/or other systems or people. Depth measuring/estimation systems that use multiple cameras from multiple views can be expensive and extremely complex. And as these autonomous systems decrease in size and available power, the supporting sensors required to estimate depth must also shrink in size and power consumption. This research will concentrate on a single passive method known as Depth from Defocus (DfD), which uses an in-focus and out-of-focus image to infer the depth of objects in a scene. The major contribution of this research is the introduction of a new Deep Learning (DL) architecture to process the the in-focus and out-of-focus images to produce a depth map for the scene improving both speed and performance over a range of lighting conditions. Compared to the previous state-of-the-art multi-label graph cuts algorithms applied to the synthetically blurred dataset the DfD-Net produced a 34.30% improvement in the average Normalized Root Mean Square Error (NRMSE). Similarly the DfD-Net architecture produced a 76.69% improvement in the average Normalized Mean Absolute Error (NMAE). Only the Structural Similarity Index (SSIM) had a small average decrease of 2.68% when compared to the graph cuts algorithm. This slight reduction in the SSIM value is a result of the SSIM metric penalizing images that appear to be noisy. In some instances the DfD-Net output is mottled, which is interpreted as noise by the SSIM metric. This research introduces two methods of deep learning architecture optimization. The first method employs the use of a variant of the Particle Swarm Optimization (PSO) algorithm to improve the performance of the DfD-Net architecture. The PSO algorithm was able to find a combination of the number of convolutional filters, the size of the filters, the activation layers used, the use of a batch normalization layer between filters and the size of the input image used during training to produce a network architecture that resulted in an average NRMSE that was approximately 6.25% better than the baseline DfD-Net average NRMSE. This optimized architecture also resulted in an average NMAE that was 5.25% better than the baseline DfD-Net average NMAE. Only the SSIM metric did not see a gain in performance, dropping by 0.26% when compared to the baseline DfD-Net average SSIM value. The second method illustrates the use of a Self Organizing Map clustering method to reduce the number convolutional filters in the DfD-Net to reduce the overall run time of the architecture while still retaining the network performance exhibited prior to the reduction. This method produces a reduced DfD-Net architecture that has a run time decrease of between 14.91% and 44.85% depending on the hardware architecture that is running the network. The final reduced DfD-Net resulted in a network architecture that had an overall decrease in the average NRMSE value of approximately 3.4% when compared to the baseline, unaltered DfD-Net, mean NRMSE value. The NMAE and the SSIM results for the reduced architecture were 0.65% and 0.13% below the baseline results respectively. This illustrates that reducing the network architecture complexity does not necessarily reduce the reduction in performance. Finally, this research introduced a new, real world dataset that was captured using a camera and a voltage controlled microfluidic lens to capture the visual data and a 2-D scanning LIDAR to capture the ground truth data. The visual data consists of images captured at seven different exposure times and 17 discrete voltage steps per exposure time. The objects in this dataset were divided into four repeating scene patterns in which the same surfaces were used. These scenes were located between 1.5 and 2.5 meters from the camera and LIDAR. This was done so any of the deep learning algorithms tested would see the same texture at multiple depths and multiple blurs. The DfD-Net architecture was employed in two separate tests using the real world dataset. The first test was the synthetic blurring of the real world dataset and assessing the performance of the DfD-Net trained on the Middlebury dataset. The results of the real world dataset for the scenes that were between 1.5 and 2.2 meters from the camera the DfD-Net trained on the Middlebury dataset produced an average NRMSE, NMAE and SSIM value that exceeded the test results of the DfD-Net tested on the Middlebury test set. The second test conducted was the training and testing solely on the real world dataset. Analysis of the camera and lens behavior led to an optimal lens voltage step configuration of 141 and 129. Using this configuration, training the DfD-Net resulted in an average NRMSE, NMAE and SSIM of 0.0660, 0.0517 and 0.8028 with a standard deviation of 0.0173, 0.0186 and 0.0641 respectively.

Indiana University-Purdue University Indianapolis (IUPUI) / The electrical disturbances in the power system have threatened the stability of the system. In the first step, it is necessary to detect these electrical disturbances or events. In the next step, a proper control should apply to the system to decrease the consequences of the disturbances. One-shot control is one of the effective methods for stabilizing the events. In this method, a proper amount of loads are increased or decreased to the electrical system. Determining the amounts of loads, and the location for shedding is crucial. Moreover, some control combinations are more effective for some events and less effective for some others. Therefore, this project is completed in two different sections. First, finding the effective control combinations, second, finding an algorithm for applying different control combinations to different contingencies in real-time. To find effective control combinations, sensitivity analysis is employed to locate the most effective loads in the system. Then to find the control combination commands, gradient descent, and PSO algorithm are used in this project. In the next step, a pattern recognition method is used to apply the appropriate control combination for every event. The decision tree is selected as the pattern recognition method. The three most effective control combinations found by sensitivity analysis and the PSO method are used in the remainder of this study. A decision tree is trained for each of the three control combinations, and their outputs are combined into an algorithm for selecting the best control in real-time. Finally, the algorithm is evaluated using a test set of contingencies. The final results reveal a 30\% improvement in comparison to the previous studies.

No description available.

No description available.

The feedforward neural network (NN) is a mathematical model capable of representing any non-linear relationship between input and output data. It has been succesfully applied to a wide variety of classification and function approximation problems. Various neural network training algorithms were developed, including the particle swarm optimiser (PSO), which was shown to outperform the standard back propagation training algorithm on a selection of problems. However, it was usually assumed that the environment in which a NN operates is static. Such an assumption is often not valid for real life problems, and the training algorithms have to be adapted accordingly. Various dynamic versions of the PSO have already been developed. This work investigates the applicability of dynamic PSO algorithms to NN training in dynamic environments, and compares the performance of dynamic PSO algorithms to the performance of back propagation. Three popular dynamic PSO variants are considered. The extent of adaptive properties of back propagation and dynamic PSO under different kinds of dynamic environments is determined. Dynamic PSO is shown to be a viable alternative to back propagation, especially under the environments exhibiting infrequent gradual changes. Copyright 2011, University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria. No part of this work may be reproduced or transmitted in any form or by any means, without the prior written permission of the University of Pretoria. Please cite as follows: Rakitianskaia, A 2011, Using particle swarm optimisation to train feedforward neural networks in dynamic environments, MSc dissertation, University of Pretoria, Pretoria, viewed yymmdd < http://upetd.up.ac.za/thesis/available/etd-02132012-233212 / > C12/4/406/gm / Dissertation (MSc)--University of Pretoria, 2011. / Computer Science / Unrestricted

Indiana University-Purdue University Indianapolis (IUPUI) / Plug-in Hybrid Electric Vehicles (PHEV) are new generation Hybrid Electric Vehicles (HEV) with larger battery capacity compared to Hybrid Electric Vehicles. They can store electrical energy from a domestic power supply and can drive the vehicle alone in Electric Vehicle (EV) mode. According to the U.S. Department of Transportation 80 % of the American driving public on average drives under 50 miles per day. A PHEV vehicle that can drive up to 50 miles by making maximum use of cheaper electrical energy from a domestic supply can significantly reduce the conventional fuel consumption. This may also help in improving the environment as PHEVs emit less harmful gases. However, the Energy Management System (EMS) of PHEVs would have to be very different from existing EMSs of HEVs. In this thesis, three different Energy Management Systems have been designed specifically for PHEVs using simulated study. For most of the EMS development mathematical vehicle models for powersplit drivetrain configuration are built and later on the results are tested on advanced vehicle modeling tools like ADVISOR or PSAT. The main objective of the study is to design EMSs to reduce fuel consumption by the vehicle. These EMSs are compared with existing EMSs which show overall improvement. x In this thesis the final EMS is designed in three intermediate steps. First, a simple rule based EMS was designed to improve the fuel economy for parametric study. Second, an optimized EMS was designed with the main objective to improve fuel economy of the vehicle. Here Particle Swarm Optimization (PSO) technique is used to obtain the optimum parameter values. This EMS has provided optimum parameters which result in optimum blended mode operation of the vehicle. Finally, to obtain optimum charge depletion and charge sustaining mode operation of the vehicle an advanced PSO EMS is designed which provides optimal results for the vehicle to operate in charge depletion and charge sustaining modes. Furthermore, to implement the developed advanced PSO EMS in real-time a possible real time implementation technique is designed using neural networks. This neural network implementation provides sub-optimal results as compared to advanced PSO EMS results but it can be implemented in real time in a vehicle. These EMSs can be used to obtain optimal results for the vehicle driving conditions such that fuel economy is improved. Moreover, the optimal designed EMS can also be implemented in real-time using the neural network procedure described.

The rapidly increasing demands for high data-rate applications and the growth of wireless devices connected to the internet overcrowded the radio frequency spectrum. This necessitates researchers to examine higher frequencies for wireless communication. Recently, visible light communication (VLC) has received significant attention as a viable solution to complement the RF technologies, thanks to the abundant unregulated/unlicensed spectrum it occupies while utilizing the existing lighting infrastructure. However, due to the physical properties of light, the signal cannot penetrate through obstacles, and the VLC system heavily relies on the existence of a line-of-sight (LoS) link between VLC transmitters and receivers. Optical reconfigurable intelligent surfaces (RISs) are recently proposed with the ability to dynamically control the wireless channel, which offers opportunities to enhance the VLC system performance by exploiting the non-LoS components of the VLC link. In this thesis, we highlight the recent developments in optical RISs and the various reflection characteristics they provide for the incident optical beams. Then, we investigate RIS-assisted VLC systems for both indoor and outdoor setups. Firstly, in indoor VLC systems, we study multi-user RIS-assisted VLC systems while considering specular and diffuse reflecting RISs. As the channel gain varies significantly between users in VLC systems, a large gap in performance is observed between users. We aim to maximize the VLC system achievable sum rate while ensuring network fairness. We formulate multi-objective optimization problems for both specular and diffuse reflecting RISs and propose a solution utilizing low complex genetic algorithm (GA) and particle swarm optimization (PSO). We highlight the advantages provided by the proposed algorithms in terms of achievable sum rate and network fairness performance. In addition, we assess the link outage ratio for specular reflecting RISs and assess the gains provided by diffuse RISs while considering an environment with mobile users. Secondly, in the context of outdoor VLC systems, we provide an overview of outdoor RIS-assisted VLC systems. In particular, we highlight the benefits of optical RISs to mitigate LoS blockage and VLC transceivers misalignment. More specifically, we focus on RIS-assisted unmanned aerial vehicles (UAVs)-based VLC, RIS-assisted vehicular VLC, and RIS-assisted streetlight-based communication. In addition, we highlight the use of RISs to support VLC outdoor-to-indoor communications.

Dynamic pricing is a difficult problem for machine learning. The environment is noisy, dynamic and has a measurable cost associated with exploration that necessitates that learning be done in short-time horizons. These short-time horizons force the learning algorithms to make pricing decisions based on scarce data. In this work, various machine learning algorithms are compared in the context of dynamic pricing. These algorithms include the Kalman filter, artificial neural networks, particle swarm optimization and genetic algorithms. The majority of these algorithms have been modified to handle the pricing problem. The results show that these adaptations allow the learning algorithms to handle the noisy dynamic conditions and to learn quickly.

Particle swarm optimization (PSO) has previously been parallelized primarily by distributing the computation corresponding to particles across multiple processors. In this thesis we present a speculative approach to the parallelization of PSO that we refer to as SEPSO. In our approach, we refactor PSO such that the computation needed for iteration t+1 can be done concurrently with the computation needed for iteration t. Thus we can perform two iterations of PSO at once. Even with some amount of wasted computation, we show that this approach to parallelization in PSO often outperforms the standard parallelization of simply adding particles to the swarm. SEPSO produces results that are exactly equivalent to PSO; this is not a new algorithm or variant, only a new method of parallelization. However, given this new parallelization model we can relax the requirement of exactly reproducing PSO in an attempt to produce better results. We present several such relaxations, including keeping the best speculative position evaluated instead of the one corresponding to the standard behavior of PSO, and speculating several iterations ahead instead of just one. We show that these methods dramatically improve the performance of parallel PSO in many cases, giving speed ups of up to six times compared to previous parallelization techniques.

Function optimization---finding the minimum or maximum of a given function---is an extremely challenging problem with applications in physics, economics, machine learning, engineering, and many other fields. While optimization is an active area of research, only a portion of this work acknowledges parallel computation, which is now widely available. Today, anyone with a modest budget can buy a cluster with hundreds of cores, pay for access to a supercomputer with thousands of processors, or at least purchase a laptop with 8 cores. Thus, an algorithm that works well in serial but cannot be parallelized is needlessly inefficient in real-life computationalenvironments.We address these issues in three connected threads of development: a high-level programming framework that makes it possible to create flexible and efficient implementations of optimization algorithms; improvements to an existing algorithm, Particle Swarm Optimization, to make it take better advantage of parallel resources; and a statistical model designed to efficiently use available information in parallel optimization by inferring search directions. Each of these is an essential step toward effective parallel optimization. First, without a suitable high-level programming model, expediency leads to purely serial development with parallel issues only an afterthought. Second, PSO has proven effective for optimization and is an excellent candidate to consider for efficient parallel implementations. Third, a model for inference of search directions is useful for understanding communication in the context of parallel optimization and provides a flexible base for continuing optimization research.