Electrical and Computer Engineering

Providing ancillary services for future smart microgrid can be a challenging task because of lack of conventional automatic generation control (AGC) and spinning reserves, and expensive storage devices. In addition, strong motivation to increase the penetration of renewable energy in power systems, particularly at the distribution level, introduces new challenges for frequency and voltage regulation. Thus, increased attention has been focused on demand response (DR), especially in the smart grid environment, where two-way communication and customer participation are part of. This paper presents a comprehensive central DR algorithm for frequency regulation, while minimizing the amount of manipulated load, in a smart microgrid. Simulation studies have been carried out on an IEEE 13-bus standard distribution system operating as a microgrid with and without variable wind generation. Simulation results show that the proposed comprehensive DR control strategy provides frequency (and consequently voltage) regulation as well as minimizing the amount of manipulated responsive loads in the absence/presence of wind power generation.

Although pattern recognition in both the visual and auditory domains is an extremely old problem with numerous intertwined research threads over the years, there remains a need for innovation in this field. Continued work on incremental improvements to the conventional techniques appears unlikely to produce the significant discoveries and breakthroughs that will lead to dramatic improvement in automatic detection and classification performance. Clearly, new ideas and techniques are desperately needed.

In this paper we introduce the use of map-seeking circuits for auditory pattern detection and classification. A map-seeking circuit is a signal processing structure used to detect a desired feature in a mixture by iteratively transforming, superposing, and comparing the composite mixture with a pattern template. The result is a mapping between the template and the position of the matched feature in the mixture. The iterative detection process is inspired by the neural connections in the human visual system. A particularly important feature of map-seeking classification is that the search operates on an additive superposition of allowable transformations of the desired feature vector, giving a linear increase in computation with increasing image complexity, rather than a brute-force feature detection that increases in computation geometrically.

Th is p aper d escribes a n R F so urce lo ca liz atio n a nd tracking system using an 8-element uniform circular array. The system is capable of precise RF source localization, fast target recognition and discrimination, fast spectral sensing, real-time processing for location tracking, and the suppression of jamming and multi-path signals. Existing systems and technology for precision location and tracking of non-cooperative RF emitters are costly and time consuming to develop and deploy, and some rely on satellite communication which is not always available and is susceptible to jamming. This novel RF source localization and tracking system offers strategic, fast, and low-cost advantages in the next generation of RF mapping, sensor networks, and mobile wireless systems since it is capable of automatic target localization and tracking, dynamic spectrum access, and directional power reception/radiation without relying on satellite communication.

A high speed focus control microelectromechanical systems (MEMS) mirror with a step response time of 100 µsec and small-displacement bandwidth of 25 kHz is reported for a 3 mm diameter, electrostatically actuated SU-8 membrane mirror. The dominant effect limiting the mirror bandwidth is viscous air damping, and the innovation we describe is the use of a perforated counter-electrode backplate that facilitates air flow underneath the membrane. We have adopted a model, originally developed for a MEMS microphone, to engineer the damping characteristics and design the air hole patterns. Cryogenic deep silicon etching creates through-wafer perforations in the backplate, and fabricated devices achieve wide-bandwidth actuation. The design approach, fabrication process, and dynamic characterization of the MEMS mirrors are shown. Finally, the focus control mirror is used in a confocal microscope for fast axial focus scanning to provide x-z cross-sectioned in vivo images. [2012-0341] Index Terms-Cryogenic deep reactive ion etching (DRIE), high-speed focus control, MEMS deformable mirrors, SU-8.

A high speed focus control microelectromechanical systems (MEMS) mirror with a step response time of 100 µsec and small-displacement bandwidth of 25 kHz is reported for a 3 mm diameter, electrostatically actuated SU-8 membrane mirror. The dominant effect limiting the mirror bandwidth is viscous air damping, and the innovation we describe is the use of a perforated counter-electrode backplate that facilitates air flow underneath the membrane. We have adopted a model, originally developed for a MEMS microphone, to engineer the damping characteristics and design the air hole patterns. Cryogenic deep silicon etching creates through-wafer perforations in the backplate, and fabricated devices achieve wide-bandwidth actuation. The design approach, fabrication process, and dynamic characterization of the MEMS mirrors are shown. Finally, the focus control mirror is used in a confocal microscope for fast axial focus scanning to provide x-z cross-sectioned in vivo images. [2012-0341] Index Terms-Cryogenic deep reactive ion etching (DRIE), high-speed focus control, MEMS deformable mirrors, SU-8.

Potential rise of global temperature due to climate change has huge impact on rice productivity and above all on food security. The variation of temperature along with diverse climatic phenomena like cyclone, drought and changing rainfall patterns cause significant loss in food grain production every year. Although weather prediction and meteorology is a very complex and imprecise science, recent research activities with Artificial Neural Network (ANN) have shown that it has powerful pattern classification and pattern recognition capabilities which can be used as a tool to get a reasonable accurate prediction of weather patterns. In this paper, an ANN model based on Multilayer Perceptron concept has been developed and trained using backpropagation learning algorithm to estimate the Daily Mean, Maximum and Minimum temperature of Dhaka, capital of Bangladesh..The result shows that Neural Network can be used for temperature prediction successfully.

the opportunity and invaluable guidance throughout this research. His dynamism, vision, sincerity and motivation have deeply inspired me. It was a great privilege and honor to carry out my work under his guidance. I am extremely grateful for what he has offered me not only as a mentor but also as a human being. I would also like to thank him for his patience, empathy, and most importantly, understanding me as a person. My Special thanks go to Prof. Steven Shaw for his support in gunshot recordings. I am extremely grateful to Angelo Borzino, for his support in the interpretation of recorded signals. I am extending my gratitude to Tyler Davis, for his assistance during the whole period. Finally, my thanks go to my parents and all my friends and staff of this department, who has directly and indirectly, supported me to complete the research work.

Acoustic gunshot signals consist of a high amplitude and short duration impulsive sound known as the muzzle blast and the shock wave. This experiment involved documenting gunshot muzzle blast sounds produced by eight commonly used firearms including Remington 870, Colt 45, Glock 19 with 9mm ammunition, Glock 23, Sig 239, AR15, 22LR, and Ruger SP 101 with 357 magnum and 38 special ammunition. An elevated microphone bracket (3m above the ground) was built to achieve a quasianechoic environment for the duration of the muzzle blast. Twelve microphones (GRAS 40DP) were mounted on the bracket in a semi-circular arc to observe the azimuthal variation of the muzzle blast. Signals were recorded using LabVIEW. Similarities and differences among waveforms are presented.

For the purposes of audio forensics research we have obtained multi-channel acoustical recordings of gunshots under controlled conditions for several firearms. The recordings are made using an elevated platform and an elevated spatial array of microphones to provide quasi-anechoic directional recordings of the muzzle blast. The consistency and repeatability of gunshot sounds are relevant to many areas of forensic analysis. This paper includes a description of the recording process and a summary comparison of the acoustical waveforms obtained from ten successive shots from the same firearm by an experienced marksman. Practical examples and applications are presented.

A generalized indoor light sensor can provide information to build and monitor indoor lighting arrangement that is aesthetically pleasing and conforming to the requirements set forth by the inhabitants. However, the identification of the surrounding lighting environment from the sensed parameters has some limitations and challenges. Till-to-date, classifiers are designed to identify only a single source, even in a multi-source environment. Classification based only on sensed values can be imperfect, as multi-type sources can share common parameters, or readings from a single source can fluctuate over time. The classification performances are mostly evaluated in controlled environments. In this work, we use a customised Bluetooth Low Energy (BLE) based light sensor that can sense and advertise major lighting parameters as instructed. Based on sensed parameters and adopting several Machine Learning (ML) and Neural Network (NN) based models off-board, we try to identify the singular and mixed presence of the four dissimilar types of sources: Incandescent, LED, CFL, and Sunlight in indoor surroundings. Off-board identification can get challenging where packet loss scenario is common. For that, we study how IoT devices with superior computational capability can utilise dimensional reduction techniques to minimize the required on-air traffic for classification. We then test classifiers with all those approaches both in controlled environments and real-world testbeds. The result shows that our best model can detect lighting environments with an accuracy of up to 98.22% in the controlled scenario and 83.33% in real-world testbeds.