2016 Electrical Engineering Abstracts

Poster Number: ECE-01

Title: Wireless Technologies and Spike Sorting Algorithms for Brain Machine Interfaces

Authors: Sylmarie Dávila-Montero; Ehsan Ashoori; Yuning Yang; Andrew Mason

Abstract: In Brain Machine Interfaces, the recording of brain activity can be performed by connecting intercortical electrodes with wires to external computers, but to avoid risk of infection and mobility limitations, it is highly desirable to have wireless power and data telemetry. Moreover, with the desire of wireless technologies, improvements in the areas of power-area efficiency of spike sorting algorithms implemented in neural signal processors (NSPs) and tracking of drifts in chronic neural signals are of special interest. This research thrust aims to improve three areas: (1) wireless power and data telemetry, (2) power-area efficiency of spike sorting algorithms, and (3) methods for tracking drifts in neural signal recorded overtime. Firstly, a new approach that consists of using concentric helical coils is presented. This approach benefits from a relatively high coupling coefficient which is resistant to coils displacements while it makes a comfortable situation for the subject. Wireless power transmission as well as data telemetry using OOK modulation is accomplished in this project. Secondly, streaming digital hardware architectures have been developed with the goal of implementing spike sorting algorithms on chip, enabling significant power and area savings without loss of sorting performance compared to existing NSPs. Statistical analysis on neural signals was performed to optimize design parameters and minimize hardware cost. Lastly, in the design of algorithms to track the drifts of neural spikes, clustering methods for spike sorting algorithms are under study to determine specific parameters that could make them track migration and appearance of new neurons in the recording.

This work was supported in part by NIOSH, MSU_SPG

 

Poster Number: ECE-02

Title: Recommendation for Cold-Start Users/Items

Authors: Iman Barjasteh; Rana Forsati; Farzan Masrour; Abdol-Hossein Esfahanian; Hayder Radha

Abstract: A major challenge in collaborative filtering based recom- mender systems is how to provide recommendations when rating data is sparse or entirely missing for a subset of users or items, commonly known as the cold-start problem. In recent years, there has been considerable interest in devel- oping new solutions that address the cold-start problem. These solutions are mainly based on the idea of exploit- ing other sources of information to compensate for the lack of rating data. In this paper, we propose a novel algorith- mic framework based on matrix factorization that simulta- neously exploits the similarity information among users and items to alleviate the cold-start problem. In contrast to ex- isting methods, the proposed algorithm decouples the follow- ing two aspects of the cold-start problem: (a) the completion of a rating sub-matrix, which is generated by excluding cold- start users and items from the original rating matrix; and (b) the transduction of knowledge from existing ratings to cold-start items/users using side information. This crucial difference significantly boosts the performance when appro- priate side information is incorporated. We provide theo- retical guarantees on the estimation error of the proposed two-stage algorithm based on the richness of similarity in- formation in capturing the rating data. to the best of our knowledge, this is the first algorithm that addresses the cold- start problem with provable guarantees. We also conduct thorough experiments on synthetic and real datasets that demonstrate the effectiveness of the proposed algorithm and highlights the usefulness of auxiliary information in dealing with both cold-start users and items.

 

Poster Number: ECE-03

Title: Dynamic Modeling of Robotic Fish Caudal Fin with Electrorheological Fluid-Enabled Tunable Stiffness

Authors: Sanaz Bazaz Behbahani; Xiaobo Tan

Abstract: In this study, we investigate a robotic fish actuated by a flexible caudal fin, which is filled with electrorheological (ER) fluid and thus enables tunable stiffness. This feature can be used in optimizing the robotic fish speed or maneuverability in different operating regimes. Lighthill’s large amplitude elongated-body theory is used to calculate the hydrodynamic force on the caudal fin, and Hamilton’s principle is used to derive the dynamic equations of motion of the caudal fin. The dynamic equations are then discritized using the finite element method, to obtain an approximate numerical solution. In particular, simulation is conducted to understand the influence of the applied electric field on the stiffness and thrust performance of the caudal fin. Experiments are conducted on a prototyped ER fluid filled flexible beam to validate the proposed dynamic model. First, the effective Young's modulus of the beam is calculated in air, by measuring the vibration of the beam and extracting natural frequency and damping ratio. The natural frequency of the beam is measured in water as well. Finally, the flexible beam is oscillated in water using a servo.

This work was supported in part by National Science Foundation (Grant DBI 0939454, IIS 1319602, IIP 1343413, CCF 1331852, and ECCS 1446793).

 

Poster Number: ECE-04

Title: Creating and Curating Reproducible Research Artifacts in the Internet-of-Things Era

Authors: Sam Boling; Andrew J. Mason

Abstract: Abstract: Reproducible research has been recognized as a growing concern in most areas of science. to achieve widespread adoption of repeatable, transparent research practices, some commentators have identified a need for better software for authoring reproducible publications. Complicating this goal, scientific investigations increasingly involve interdisciplinary teams, sophisticated workflows for acquiring and analyzing data, and huge datasets that rely on considerable metadata to interpret. Computational scientists have begun to adopt tools for managing the complex histories of their data and procedures, but software which simultaneously allows researchers to specify experiments, remotely control equipment, and capture and organize data remains immature. This project demonstrates a software architecture for programmable remote control of custom and commercial lab equipment, automatic annotation and queryable storage of data sets, and provenance-aware specification and refinement of experiment and analysis procedures. The design consists of a suite of small, single-purpose software services which may be connected and controlled from a web browser, notably including a graphical programming tool, an abstraction layer for interfacing with commercial hardware and custom embedded systems, and a hybrid document/table database for persistent storage of annotated experimental data. The software implementation embraces modern web technologies and best practices to produce a modular, user-extensible framework that is well-suited for helping to integrate computer-controlled research labs with the emerging Internet of Things.

This work was supported in part by NIH grant R01ES022302

 

Poster Number: ECE-05

Title: Injection Molding Terahertz Passive Components

Authors: Jennifer Byford; Zachary Purtill; Premjeet Chahal

Abstract: A new fabrication process for passive terahertz components is introduced. to improve upon prior work in dielectric block machining, micromachining, and injection molding to create terahertz passive components we introduce the use of 3D printed injection molds as a low cost alternative for fabrication. This new process increases the amount of available materials to use in devices, is easy, and cost effective. Molds are 3D printed on commercially available printers using polylactic acid (PLA), a biodegradable aliphatic polyester and VeroWhite. An injection molding machine is used to melt low density polyethylene (LDPE) or high density polyethylene (HDPE) pellets and fill the molds. Sample passive THz components are designed in ANSYS HFSS and fabricated using the new process including a probe, ridge waveguide, and photonic crystal filter. Samples are then measured using a frequency domain terahertz system and compared to their expected performance from simulation. Future work to improve the process is also considered.

This work was supported in part by GAANN Grant

 

Poster Number: ECE-06

Title: Nonlinear Model Predictive Control of a Tail-Actuated Robotic Fish

Authors: Maria Castaño; Xiaobo Tan

Abstract: Oceanic sustainability has been a growing global concern due to the increase of potential threats to the integrity of aquatic ecosystems. As a result, there is been an increase of interest on the use autonomous aquatic robots to monitor such environments. In recent years, underwater robots that propel and maneuver themselves as real fish, often called robotic fish, have emerged as mobile sensing platforms for the monitoring of freshwater and marine environments.These robots achieve locomotion via actively controlled fins, and actuation is often achieved via oscillatory inputs. Given the unpredictability of the environments these robots charter, path planning and accurate trajectory control is of importance for mission success and to achieve high energy efficiency. In this work, we propose a nonlinear model predictive control (NMPC) for tracking of an optimal planned path. In this design, we use the the bias, and amplitude of the tail-beat as the input to be determined by the NMPC. The effectiveness of the proposed approach is demonstrated via simulation.

This work was supported in part by National Science Foundation (IIS 1319602, CCF 1331852, ECCS 1446793)

 

Poster Number: ECE-07

Title: Study of the Effect of Temperature and Pocket Holder Depth on Single Crystal Diamond Growth via Microwave Plasma Assisted CVD

Authors: Amanda Charris; Shreya Nad; Jes Asmussen

Abstract: Single crystal diamond (SCD) substrates were successfully synthesized via microwave plasma assisted chemical vapor deposition (MPACVD) in a 2.45 GHz microwave plasma CVD reactor at an experimental pressure of 240 torr (320mbar), H2 flow rate 400 sccm, and 5% CH4/H2 methane concentration. The SCD seed was placed in pocket substrate holder as discussed in [1]. Input power variations from 1.8 kW to 3.3 kW were performed in order to keep the temperature constant between 1000°C and 1020°C. Growth times of 10 hours, 24 hours, 48 hours, and 72 hours was conducted. The growth rate varied between 26 and 30 μm/h depending on the input discharge power density. Thick (1.1 mm – 1.7mm) diamond crystals were deposited on 3.5mm x 3.5mm x 1.4mm HPHT type Ib, (100)-oriented single crystal diamond seed substrates. The surface morphologies exhibited smooth and flat top surface without the formation of a polycrystalline diamond rim. By modifying the pocket holder depths, substrate temperatures, and power densities, different crystal shapes were synthesized. Consequently, under these different growth conditions different faces on CVD crystals were observed and were investigated using scanning electron microscopy (SEM) and single crystal x-ray diffraction. The effect of the temperature, pocket holder depth, and power level variations on the crystal growth rate, dimensions and quality are discussed. [1] S. Nad, Y. Gu, and J. Asmussen, Diam. Relat. Mater., vol. 60, pp. 26–34, 2015.

 

Poster Number: ECE-08

Title: Natural Language Based Robotic Programming

Authors: Yu Cheng; Jiatong Bao; Yunyi Jia; Zhihui Deng; Haichu Chen; Lixin Dong; Ning Xi

Abstract: There are multiple ways to program a new skill for a robotic system. Natural language (NL) programming is a very promising method due to its versatility, ease of use for humans, and the lack of need for extensive training. Since NL instructions given by human users can not be understood by the robots directly, the linguistic input has to be transformed into a formal representation which tries to capture the intent of the users and removes the ambiguity of NL. Current formal representations implicitly require instructors to give step by step NL action plans, then map the instructions into formal representation. This requires the instructor to be familiar with both the task and the robot such that the instructor is able to decompose the task into action plans with respect to the robot being used, which is not suitable for industrial applications, such as assembly tasks. Also, it is inconvenient and increases the burden for untrained users. to overcome this issue, we propose to use Dependency Relation Matrix (DRM) to model the assembly jobs tasked with the NL instructions as a group of spatial relations. The proposed approach is more reasonable, convenient, and suitable for untrained users to describe the task and program the skill for the robot. Additionally, DRM can be used to decompose the given task as a sequence of subtasks and symbolically parameterize each subtask. In this paper, we describe the DRM framework, describe how to decompose and parameterize the tasks, and illustrate the utility of this approach with experiment results.

This work was supported in part by National Science Foundation

 

Poster Number: ECE-10

Title: Distributed Time Difference of Arrival Localization of a Moving Target

Authors: Osama Ennasr; Xiaobo Tan

Abstract: Localization and tracking of a moving target has been established as a key problem in wireless sensor networks, with many algorithms being proposed in this area. In particular, time-difference of arrival (TDOA) localization is considered to be a cost effective and accurate localization technique. However, traditional TDOA algorithms rely on a central node that produces an estimate of the target's location by gathering measurements from all other nodes in the network. In this work, we solve the problem by distributing the estimation among all agents in the network, avoiding problems posed by the centralized approach, such as single-node failure. Each agent in the network runs its own extended Kalman filter (EKF) in order to estimate the target's position, while a neighbor-based averaging procedure ensures that all agents agree on the same estimation. This approach does not require each node to fully observe the process, i.e. some nodes in the network may have an insufficient number of neighbors to accurately estimate the target's position. A bound on the maximum estimation error is established analytically, with numerical examples illustrating the performance of the proposed algorithm.

This work was supported in part by National Science Foundation (IIS 1319602, CCF 1331852, ECCS 1446793)

 

Poster Number: ECE-11

Title: Waveguide Verification Standards for the Characterization of Magnetic Materials

Authors: Jonathan Frasch; Edward Rothwell

Abstract: Measuring unknown materials requires care in the calibration and setup of the measurement system. Without a proper procedure, the level of error in the measured and derived property values becomes uncertain. Many magnetic materials are manufactured and can have a variable composition. A surrogate material with predictable and consistent measurement values can provide a measure of the quality of the calibration and give useful information about the error of the system. We have developed a class of surrogate samples with consistent properties, which are made from fully insertable metallic segments. The size and position of these segments are determined through the use of evolutionary algorithms, which search the possible designs for those that best match a given profile for electrical permittivity and magnetic permeability when using the Nicolson-Ross-Weir methodology for extraction. Analysis of these designs is carried out using a mode-matching method for sequential changes of dimension of the waveguide. Results for an X-Band surrogate have already been obtained and further work is in progress for additional frequency bands.

This work was supported in part by Livingston Tool & Mfg . Co.

 

Poster Number: ECE-12

Title: Mobile Node Location Tracking Using LED Optical-Based Simultaneous Localization and Communication

Authors: Jason N. Greenberg; Xiaobo Tan

Abstract: Localization and communication are both essential functionalities of any practical sensor network. Being able to achieve both functionalities through a single technique, i.e. Simultaneous Localization And Communication (SLAC), would greatly reduce the complexities of system implementation. In this poster presentation implementation details of a location tracking algorithm using an LED-based optical communication system for location identification and a Kalman Filter for position estimation is discussed. The intended application of this method is for use in small low-powered underwater robotic sensor network applications where the current acoustic based-technologies are not feasible since they are often too bulky and require too much energy to be practical on these robots. Furthermore acoustic technologies are limited to low data rates and typically suffer from propagation delays which make them unsuitable to a data hungry network which is densely connected. This poster presents simulation and experimental results of this tracking technique applied to a terrestrial network of mobile and stationary robots as proof of concept. Future work will explore the effectiveness of this technique in an aquatic setting.

This work was supported in part by NSF

 

Poster Number: ECE-13

Title: A Generative Kriging Surrogate Model for Constrained and Unconstrained Multiobjective Optimization

Authors: Rayan Hussein; Kalyanmoy Deb

Abstract: Surrogate models are eective in reducing the computational time required for solving optimization problems. However, there have been a lukewarm interest in nding multiple trade-o solutions for multi-objective optimization problems using surrogate models. The literature on surrogate modeling for constrained optimization problems is also rare. The difficulty lies in the requirement of building and solving multiple surrogate models, one for each Pareto-optimal solution. In this paper, we rst provide a brief introduction of the past studies and suggest a computationally fast, Kriging-based, and generative procedure for nding multiple near Pareto-optimal solutions in a systematic manner. The expected improvement metric is maximized using a real-parameter genetic algorithm for nding new solutions for high-delity evaluations. The approach is computationally fast due to the interlinking of building multiple surrogate models and in its systematic sequencing methodology for assisting one model with another. In standard two and three-objective test problems with and without constraints, our proposed methodology takes only a few hundreds of high-delity solution evaluations to nd a widely distributed near Pareto-optimal solutions compared to the standard EMO methods requiring tens of thousands of high-delity solution evaluations. The framework is generic and can be extended to utilize other surrogate modeling methods easily.

 

Poster Number: ECE-14

Title: Critical Evaluation of Shunt and Series Compensation Schemes for Hybrid Matrix Converter

Authors: Ameer Janabi; Bingsen Wang

Abstract: This poster is focused on conditioning the voltage and current waveform quality of a hybrid matrix converter that consists of a conventional nine-switch matrix converter and a back-to-back voltage source converter. Upon critical evaluation of the existing methods for shunt and series compensation, the fundamental limitations for achieving superior results have been identified. A new strategy based on power averaging for obtaining the reference compensating current and voltage has been proposed. The effectiveness of the proposed method has been evidenced by the simulation results for both the shunt and series compensation with concurrent presence of the harmonic components in voltages and currents.

 

Poster Number: ECE-15

Title: Lightweight Linear Electric Machine Design for Free Piston Engine

Authors: William Jensen; Shanelle Foster

Abstract: Electric vehicles are restricted in their driving range by the limited energy storage capacity of current battery technology. Extending the driving range can be achieved by utilizing an internal combustion engine (ICE) to charge the battery that powers the drivetrain. to improve on this system, a free piston engine (FPE) equipped with a linear generator can replace the ICE. A FPE allows for flexible fuel options and the elimination of the crankshaft, which can lead to cleaner electric energy generation. Lightweighting the linear generator enables the FPE to operate at a higher speed and can possibly lead to improved vehicle efficiency. This project presents a linear machine design that allows for the freedom to use any material to support the moving secondary. Aluminum is chosen to replace the back iron in the moving secondary of the linear generator because it is lighter, cheaper, and easier to couple to the combustion piston. A Halbach array of permanent magnets is incorporated in the design of the proposed machine to maintain the main flux path in the secondary. Finite Element (FE) simulation results demonstrate that by replacing the back iron in the secondary with aluminum, the rated speed of the machine increases without significant degradation in performance. Simulation results are also compared with a commercially available linear machine to verify the advantage of designing a lightweight machine with a Halbach array. Analytical and FE thermal analysis of the proposed machine are used to evaluate steady-state operating temperatures at rated current.

 

Poster Number: ECE-16

Title: Monolithic Multichannel GaN LED Arrays

Authors: Wasif Khan; Wen Li

Abstract: Establishment of a reliable bi directional communication channel between the nervous system of a freely behaving vertebrate and the external environment is important for the treatment of neurological diseases and events like spinal cord injury, stroke and traumatic major amputations. Optogenetics, which uses light stimulation to control the excitation and inhibition of action potentials in genetically modified neurons. However, conventional methods like tethered optical fiber impedes the subject from moving freely and poor spatial resolution of other methods limits their functionality. A light-stimulating interface with sufficient light intensity that also allows free behavior of the experiment subject is imperative. In response to meet these challenges, monolithic multichannel micro LED arrays have been fabricated using less expensive reactive ion etching. These LED arrays could be surgically implanted on the cortex of an animal subject to stimulate light in vivo without hindering the animal’s free behavior. Annealing was performed to improve the performance of the LED’s. These arrays demonstrated sufficient light intensity that is required for genetically modified neuron activation. The fabricated LED’s showed lower heat dissipation, which reduces the risk of impairing the neurons. The results of the RIE etching process, electrical and optical performance of the fabricated LED arrays were also characterized.

 

Poster Number: ECE-17

Title: Investigation, Optimization and Demagnetization Effect of the Use of New Ferrite Magnets for Design of Spoke Type and PMASynRM Motors

Authors: Cristian A. Lopez; Elias G. Strangas

Abstract: Reducing the amount of rare earth materials used in interior permanent magnet (IPM) machines makes for a lower cost and rare earth material non-dependency. Using new to the market ferrite magnets could make this possible. A study of the effect of the use of new ferrites in certain performance measures could be done to compare to the performance of currently used ferrite magnets and also determine the effect of temperature in demagnetization. Two types of motor designs, the spoke type ferrite and the permanent magnet synchronous reluctance motor (PMASynRM) were studied and optimized to make a fair comparison. A clear increase of 13% - 22% in torque performance was found when using the new ferrites with the optimization still to be done.

 

Poster Number: ECE-18

Title: Large-Signal RF Model Extraction of GaN HEMT Power Amplifiers

Authors: Nicholas C. Miller; John D. Albrecht

Abstract: Power amplifier circuit design relies heavily on device parameterization to capture non-linear behavior. Extracting high-fidelity models often involves extensive electrical characterization that must be repeated after any change in the underlying device design, materials, or fabrication. If the high-speed device is operating at a fundamental frequency of 35 GHz, it becomes exceedingly difficult to measure device responses at the harmonics. Due to the non-linear nature of the devices, large-signal single-tone stimuli will generate non-negligible signals at harmonics of the fundamental frequency. These large voltage inputs also yield hot electron effects, implying that proper treatment of electronic band structure becomes paramount. Furthermore, external matching circuits, e.g., high-Q bandpass filters, must be included to properly emulate transistors in power amplifier systems. A predictive simulation is an attractive alternative to fabrication and measurements for high-frequency, large-signal characterizations of electronic devices that would otherwise be prohibitively expensive. Moreover, compact models, which accurately predict the underlying device performance, could be imported into circuit design software, e.g. Keysight's Advanced Design System (ADS,) and used for circuit design with state-of-the-art transistors.

This work was supported in part by The SMART Scholarship Program

 

Poster Number: ECE-19

Title: Embedded Passive RF Tags Towards Intrinsically Locatable Buried Plastic Materials

Authors: Mohd Ifwat Mohd Ghazali; Saranraj Karuppuswami; Premjeet Chahal

Abstract: This paper demonstrates the use of passive harmonic tags as markers for buried plastic pipes. The tag design is based on a double slot antenna that can be buried in the plastic pipes during manufacturing. Design and measurement results are presented for two tag designs: one with a metal backing and one without. The tags are embedded in a plastic casing representing the walls of the plastic pipes. The tag operates at a fundamental frequency (fo) of 2.5 GHz with a return signal (2fo) of 5 GHz. Use of harmonic tags eliminates the clutter from the surface of the ground and thus enhances signal to noise (S/N) ratio of the return signal without the use of filters. The antenna is linearly polarized and polarization can be used as a marker for the direction of the buried pipes. The tag design is compact and can also readily be interrogated using simple hand-held radar units.

This work was supported in part by Department of Transportation, Midland Research Institute for Value Chain Creation

 

Poster Number: ECE-20

Title: Enhancing the Safe and Efficient High Pressure Microwave Plasma Assisted CVD Operating Regime for SCD Synthesis Using Continuous Wave and Pulsed Microwave Excitation

Authors: Matthias Muehle; Jes Asmussen; Michael F. Becker; Thomas Schuelke

Abstract: Achieving single crystalline diamond (SCD) wafer sizes above 1” requires serious growth effort. Diamond is not increasing its lateral dimensions during the growth process. Thus there are 3 main concepts to increase SCD dimensions: (1) mosaic growth, (2) flipped side approach, and (3) flipped seed approach. The first two concepts have been realized. The built-up of internal stress between the individual clones or on the flipped side makes these approaches impractical. The two major concerns with the superior flipped side approach are a significantly higher amount of SCD post processing and far more total growth needed. While we made good progress in mastering the first issue, the problem of significantly increasing the growth rate still has to be addressed. The development of new growth reactors allowed enhancing the safe and efficient growth window resulting in deposition rates up to 75 um/hour. This reactor was limited in maximum pressure (300 torr) due the stability of the plasma. We equipped a reactor with a power supply, switchable between continuous and pulsed excitation. We propose a series of different reactor performance curve experiments while expanding the pressure range to 400 torr using a continuous wave discharge. The substrate temperature as function of the input power will be recorded. Photography and optical emission spectroscopy we will be recorded for plasma characterization. Variations of different parameters of a pulsed microwave discharge, such as pulsing frequency, duty cycle and the role of Pavg vs. Ppeak by varying Pmin and Pmax, are performed.

 

Poster Number: ECE-21

Title: Portable Electrochemical Malaria Detection System for Affordable In-Field Measurement

Authors: Sina Parsnejad; Tung-yi Lin; Linlin Tu; Andrew Mason

Abstract: Malaria is one of the most important infectious diseases in the 21th century with an estimated 216 million infections and 655 thousand deaths reported in 2010. Despite efforts, this disease is the number one cause of morbidity and mortality in children under 5. Diagnosis of Malaria is difficult and currently the only viable method is to detect Malaria parasites is visual analysis using as microscopes and image processors. However, these methods are expensive, stationary and require trained professionals to implement and, hence, fail to perform well in poor rural areas and remote locations. In order to address these issues, we are developing a portable electrochemical detection system. The system is composed of an inexpensive, discardable, electrochemical biosensor and an attachable electrochemical instrumentation module that is connected to and powered by a smartphone. The smartphone processes and stores data and provides a user interface to display test results and control the detection system. The electrochemical instrumentation module utilizes state-of-the-art analog chips and a low power digital microcontroller to realize a miniaturized and low power sensor data acquisition sytem that can adapte intelligently and autonomously to various measurement conditions. together, the microfluidic biosensor and miniaturized electrochemical instrumentation module enable aportable and affordable system that does not require trained users and can operate in rural areas.

This work was supported in part by NIH

 

Poster Number: ECE-22

Title: Soft Pneumatic Bending Actuator with integrated CNT-Based Strain Sensors

Authors: Thassyo Pinto; Le Cai; Chuan Wang; Xiaobo Tan

Abstract: Soft robotics is a recent trend in engineering that seeks to create machines that are soft, compliant, and capable of withstanding damage, wear and high stress. Soft actuators are the major elements of soft robots, and their elastomeric substrate enables generation of sophisticated motion with simple controls. Although several fabrication methods, material selection, and structure design have been investigated for the construction of soft bending actuators, limited attention has been paid to the integration of distributed sensors for performing localized measurement. Carbon nanotubes (CNTs) are molecular-scale tubes of carbon atoms with remarkable mechanical and electronic properties, showing potential application in sensing devices. In this work, we present the design, fabrication, and testing of a new type of CNT-based sensor array for measuring strain along the bottom layer of a soft pneumatic bending actuator. Simulation and experimentation were performed in order to analyze the soft actuator deformation during bending. The results demonstrate the promise of the proposed soft actuators with integrated strain sensing, which lays groundwork for a myriad of applications in grasping, manipulation, and bio-inspired locomotion.

This work was supported in part by CAPES Foundation (BEX-13404-13-0); National Science Foundation (DBI-0939454)

 

Poster Number: ECE-23

Title: Computational Investigation Using Subregion Finite Element Method for Solving the Inverse Problem in Eddy Current NDE

Authors: Mohammad R. Rawashdeh; Lalita Udpa; S. Ratnajeevan H. Hoole

Abstract: Eddy Current Testing (ECT) plays a key role in detecting cracks and corrosion in conductors and ferromagnetic materials. A computational model for ECT is valuable for optimizing the parameters of the test procedure and maximize the detection probability of defects. Finite Element Method (FEM) is widely used in industry for developing simulation models. This paper presents an efficient mathematical technique in which the subregion method can be used in conjunction with finite element method to simulate an Eddy Current Test and generate defect signals. The proposed study is particularly useful in the solution of inverse problems where the objective is to detect and reconstruct multiple defects, given eddy current test data. The underlying idea in this method is to separate the defect region from the rest of the geometry so that only this subregion is considered in the iterative process. Several examples of defects are presented along with tangible results and improved processing time demonstrate the power of subregion finite element method as an effective method in solving inverse problems in Nondestructive Evaluation (NDE).

 

Poster Number: ECE-24

Title: Cupula-Inspired IPMC Sensor: Fabrication, Simulation and Sensor Characterization

Authors: Montassar Aidi Sharif; Hong Lei; Xiaobo Tan

Abstract: Ionic polymer metal composites (IPMCs) have inherent sensing and actuation capabilities. IPMCs are functional materials that produce a current or a voltage under an applied mechanical load and produce a mechanical displacement under an applied voltage. IPMCs have applications in multiple fields such as biomedical engineering and underwater robotics. This work presents the mechanoelectrical model of an IPMC when it is used as a flow sensor in mimicking the fish lateral line system. We consider both an IPMC alone and an IPMC encapsulated with a cupula-like structure. COMSOL 5.1 has been used to implement a 2D finite element model to simulate the IPMC and the cupula-based IPMC in underwater environments. Experiments involving IPMC sensors under both AC and DC flow conditions are conducted to validate the FE model.

 

Poster Number: ECE-25

Title: Experimental Implementation of Extended Kalman Filter-Based Optical Beam Tracking with a Single Receiver

Authors: Pratap Bhanu Solanki; Xiaobo Tan

Abstract: For an efficient free-space optical communication, the receiver should have a close-to-line-of-sight (LOS) link with the transmitter. Maintaining an alignment for LOS is a difficult task due to the constant movement of the underlying mobile platform or unwanted disturbances. The previously proposed Extended Kalman Filter-based algorithm uses light intensity measurements from single receiver photo-diode and a scanning technique to estimate the relative orientation between the receiver and the transmitter, which is used to adjust the receiver’s orientation accordingly. This work focuses on the experimental implementation of the algorithms, and evaluates the estimation and control performance through an extensive set of simulation and experiments. The proposed algorithms are found to be effective in general, and their limitations are further explored in this work.

This work was supported in part by National Science Foundation (IIS 1319602, ECCS 1446793)

 

Poster Number: ECE-26

Title: Modeling and Assessment of PV Solar Plants for Composite System Reliability Considering Radiation Variability and Component Availability

Authors: Samer Sulaeman; Joydeep Mitra

Abstract: This paper presents a method to model the output power of large PV systems in composite system reliability assessment. Grid level PV systems are usually constructed from a large number of power electronic components and PV panels. Modeling of these systems in power system reliability is a complex task due to the dependency of the output power on the intermittent source (solar) and the availability of a large number of system components. An analytical method to construct a capacity outage probability and frequency table (COPAFT) that captures both the intermittency of the input source and component failures is proposed to model PV systems. The intermittency of the input source and components availabilities are modeled separately and then convolved to construct a single COPAFT. The resulting COPAFT forms a multi-state reliability model of the entire solar facility. The proposed method reduces the complexity of modeling and evaluating large PV systems in composite system reliability assessment. The method is demonstrated on IEEE RTS. Considering the PV farm location with a view to enhance system reliability, a sensitivity study was conducted to measure the effect of the location of the PV farm on overall system reliability. The results confirm that connecting PV farms to the buses that are at a high risk enhances the overall system reliability.

 

Poster Number: ECE-27

Title: Locally Linear Manifold Model for Gap-Filling Algorithms of Hyperspectral Imagery

Authors: Suha Suliman; Hayder Radha

Abstract: Landsat 7 Enhanced Thematic Mapper Plus (ETM+) Scan Line Corrector (SLC) device, which corrects for the satellite motion, has failed since May 2003 resulting in a loss of about 22% of the data. Thus, each scan overlaps in the middle of each scene instead of performing a regular parallel scans. Many data recovery approaches were implemented to fill in the gaps and most of them have limitation in terms of low accuracy at certain areas and high computing time. to improve the reconstruction of Landsat 7 SLC-off images, Locally Linear Manifold (LLM) model is proposed for filling gaps in hyperspectral imagery. In this approach, each spectral band is modeled as a non- linear locally affine manifold that can be learned from the matching bands at different time instances. Moreover, each band is divided into small overlapping spatial patches. In particular, each patch is considered to b e a linear combination (approximately on an affine space) of a set of corresponding patches from the same location that are adjacent in time or from the same season of the year. Fill patches are selected from Landsat 5 Thematic Mapper (TM) pro ducts of the year 1984 through 2011 which have similar spatial and radiometric resolution as Landsat 7 pro ducts. Using this approach, the gap-filling process involves feasible point on the learned manifold to approximate the missing pixels. The proposed LLM framework is compared to some existing single-source (Average and Inverse Distance Weight (IDW)) and multi-source (Lo cal Linear Histogram Matching (LLHM ) and Adaptive Window Linear Histogram Matching (AWLHM)) gap-filling methodologies. We analyze the effectiveness of the proposed LLM approach through simulation examples with known ground-truth. It is shown that the LLM-model driven approach outperforms all existing recovery methods considered in this study. The superiority of LLM is illustrated by providing better reconstructed images with higher accuracy even over heterogeneous landscape. Moreover, it is relatively simple to realize algorithmically and it needs much less computing time when compared to the state of-the-art AWLHM approach.

 

Poster Number: ECE-28

Title: VO2-Based MEMS Mirror

Authors: David Torres; Tongyu Wang; Sarah Dooley; Xiaobo Tan; Huikai Xie; Nelson Sepúlveda

Abstract: We present the fabrication of the first microelectromechanical systems (MEMS) mirror devices with integrated vanadium dioxide (VO2), where the actuation mechanism is mainly due to the solid-solid phase transition of VO2. The device consists of four actuators that control the movement of a platform with a reflective layer, where movement of individual actuators would tilt the platform (tilt movement) and synchronous movement of the actuators would increase/decrease the elevation of the platform (piston movement). The present VO2-based MEMS mirror device is operated electro-thermally through integrated resistive heaters, and its behavior is characterized across the phase transition of VO2, which occurs at a temperature of approximately 68 degree Celsius and spans about 10 degree Celsius. The maximum vertical displacement (piston actuation) of the mirror platform is 75 μm, and it occurs for an input voltage of 1.1 V. This translates to an average power consumption of 6.5 mW per mirror actuator, and a total power consumption of 26.1 mW for the entire device.

This work was supported in part by This work was supported in part by the National Science Foundation under Grant ECCS 1306311 and Grant CMMI 1301243. Device development was made possible by a cooperative research and development agreement (CRADA No. 15-075-RY-01) between AFRLs Sensors Dir

 

Poster Number: ECE-29

Title: Transformation of Functional Connectivity Brain Networks to Signals

Authors: Marisel Villafañe-Delgado; Selin Aviyente

Abstract: Functional connectivity networks in the brain exhibit complex network characteristics, such as small-worldness and scale-free.  The transformation of networks to time series brings an alternative for the characterization of network’s structure.  Previously proposed methods are limited to binary graphs and hence cannot be applied to functional connectivity networks.  In this work, it is proposed to employ the resistance distance matrix of weighted graphs as the distance matrix for transforming networks to signals based on classical multidimensional scaling.  By using properties of the resistance distance, we present a framework for obtaining information about the network’s structure based on the signals and then reconstructing the original network from those signals.  Finally, the proposed method is applied to functional connectivity networks based on electroencephalographic data.

This work was supported in part by NSF GRFP

 

Poster Number: ECE-30

Title: Causality Analysis of fMRI Data Based on the Directed Information Theory Framework

Authors: Zhe Wang; Ahmed Alahmadi; David C. Zhu; Tongtong Li

Abstract: Here, we conduct fMRI based causality analysis in brain connectivity by exploiting the directed information (DI) theory framework. More specifically, we provide the detailed procedure on how to calculate the DI for two finite time series. The two major steps involved here are optimal bin size selection for data digitization, and probability estimation. Also, we demonstrate the applicability of DI based causality analysis using experimental fMRI data, and compare the results with that of the Granger Causality (GC) analysis. Our analysis indicates that GC analysis is effective in detecting linear or nearly linear causal relationship, but may have difficulty in capturing nonlinear causal relationships. On the other hand, DI based causality analysis is more effective in capturing both linear and non-linear causal relationships.

 

Poster Number: ECE-31

Title: Lab-on-CMOS Platforms for Highly Integrated Microfluidic Biosensor Arrays

Authors: Heyu Yin; Lin Li; Andrew J. Mason

Abstract: Monolithic microsystems based on complementary metal-oxide-semiconductor (CMOS) microelectronics offer great promises to analyze biological and chemical processes via high throughput platforms that physically interface miniaturized sensors with fluid samples. The “lab-on-CMOS” concept pioneered at MSU seeks to develop an integrated microsystem platform that incorporates a biointerface array, CMOS instrumentation circuits and microfluidic sample handling devices into a compact, cost-effective, continuous-use, analytical system suitable for biological research and biomedical applications. The lab-on-CMOS platform we are currently developing utilizes a CMOS electrochemical instrumentation chip with an array of sensor electrodes directly on the CMOS chip to eliminate wiring constraints and minimize measurement noise. An epoxy carrier is employed to extend the surface area of the sensor-electronics chip for subsequent integration with multi-channel microfluidics. A screen-printed planar metallization technique for lab-on-CMOS that overcomes challenges associated with traditional thin film metallization has been developed. Utilizing the epoxy chip-in-carrier packaging approach with screen-printed metallization, electrical interconnects are shown to reliably resolve up to 10μm step height differences between the CMOS chip and the surrounding carrier. The metallization process was also shown to be compatible with subsequent microfluidic integration to complete a lab-on-CMOS device platform that is well suited for high throughput biosensor measurement.

This work was supported in part by NSF_ECCS-1307939

 

Poster Number: ECE-32

Title: Wearable Electrochemical Gas Sensor Array for Personal Air Pollution Exposure Assessment

Authors: Heyu Yin; Sina Parsnejad; Hao Wan; Sam Boling Ehsan Ashoorie; Andrew J. Mason

Abstract: Exposure to air pollution consistently ranks among the leading causes of illness and mortality globally, and the growing potential impact of airborne pollutants and explosive gases on human health and occupational safety has escalated the demand for sensors to monitor hazardous gases. Unfortunately, current preventative measures and treatments for air toxins are ineffective due in large part of our inability to properly characterize and quantify acute exposure to air pollutants. to overcome these challenges, a wearable autonomous multi-gas sensor system capable of real-time environmental monitoring could provide immediate feedback to warn the wearer of imminent environmental threats as well as a record of individual exposure that would aid the development of new treatment approaches. We present a miniaturized wearable system for the pollutant gas monitoring that seeks to achieve this goal by synergistically integrating sensors, electronics, and data analysis algorithms into an autonomous wearable system. Electrochemical sensors featuring room temperature ionic liquid electrolytes are utilized for low-power operation, high sensitivity and selectivity, and long life with low maintenance. Micro-fabricated electrode structures enable miniaturization and rapid response. A custom multi-mode electrochemical instrumentation circuit combines all needed signal condition while minimizing system cost, size and power consumption. Embedded sensor array signal processing algorithms enable gas classification and concentration estimation of a real-world mixture of gas analytes within wearable system for acute exposure assessment.

This work was supported in part by NIH_R01ES022302

 

Poster Number: ECE-33

Title: Fully-Printed Stretchable Conductor and Strain Gauge

Authors: Suoming Zhang; Le Cai; Wei Li; Jinshui Miao; tongyu Wang; Nelson Sepúlveda; Junghoon Yeom; Chuan Wang

Abstract: This paper exploits the strategy to make the device fabricated by the non-stretchable material Silver Nanoparticles (AgNPs) stretchable by investigating the geometry design using a fully printed process as the application of stretchable conductor and strain gauge. We had printed the AgNPs pattern onto the elastomer substrate and showed the stretchability of the device could be tuned by changing the radius of the serpentine structure. The device with smaller radius was more sensitive to the strain, resulting a high gauge factor of 1000000, demonstrated as a stain gauge to detect the finger motion of human beings, while the device with larger radius was more stretchable (more than 25%), used as a stretchable conductor for driving the LED. The printed strategy and demonstrated application would have broad prospects in stretchable electronics.