Recent Submissions

  • Retrodirective phased array antenna for nanosatellites

    Long, Justin W.; Thorsen, Denise; Kegege, Obadiah; Hawkins, Joseph; Mayer, Charles (2019-12)
    This thesis presents a S-band phased array antenna for CubeSat applications. Existing state-of the-art high gain antenna systems are not well suited to the majority of CubeSats, those that fall within the 1U (10 cm x 10 cm x 10 cm) to 3U (10 cm x 10 cm x 30 cm) size ranges and in Low Earth Orbit (LEO). The system presented in this thesis is designed specifically to meet the needs of those satellites. This system is designed to fit on the 1U face (10 cm x 10 cm) of a CubeSat and requires no deployables. The use of beamforming and retrodirective algorithms reduces the pointing requirements of the antenna, easing the strict requirements that high gain antennas typically force on a CubeSat mission. Additionally, this design minimizes volume and uses low cost Commercial-off-the-Shelf (COTS) parts. This thesis discusses the theoretical background of phased array theory and retrodirective algorithms. Analysis are presented that show the characteristics and advantages of retrodirective phased antenna systems. Preliminary trade studies and design analyses show the feasibility and expected performance of a system utilizing existing COTS parts. The preliminary analysis shows that an antenna system can be achieved with ≥8.5 dBi of gain, 27dB of transmitted signal gain, 20% Power Added Efficiency (PAE) within a 1 W to 2 W power output, and an 80° effective beamwidth. Simulation results show an example antenna array that achieves 8.14 dBi of gain and an 82° effective beamwidth. Testing results on a prototype of the front-end electronics show that with minimal calibration, the beamforming and scanning error can be reduced to 5°. The power consumption and signal gain of the electronics is also verified through testing. The CubeSat Communications Platform, a CubeSat mission funded through the Air Force Research Laboratory is in Phase A design to demonstrate this antenna system, along with other experimental payloads. This thesis includes a discussion of interface control, mission requirements, operations, and a recommended experiment sequence to test and verify the antenna system on orbit.
  • Classification and signal processing of radio backscatter from meteors

    Klemm, Jared; Thorsen, Denise; Bossert, Katrina; Collins, Richard; Mayer, Charlie (2019-12)
    Ground-based radar systems are routinely used to detect the trails of ionized particles that are formed by meteoroids falling through Earth's atmosphere. The most common use for these meteor radar systems is for atmospheric wind studies of the mesosphere and lower thermosphere (80-100 km altitude). Because these meteor trails are embedded in the background winds of the middle atmosphere, atmospheric winds in that region can be measured by observing the radial velocities of the trails. There has also been a considerable amount of research over the last few decades into estimation of neutral atmospheric temperatures using the measured decay time of meteor trails. Several methods exist for estimating atmospheric temperature using meteor radar observations, but there are limitations to these approaches. This thesis focuses on examining aspects of meteor radar signal and data processing, specifically interferometry and echo classification. Interferometry using the measured signal phase differences between antennas allows for the location of meteor trails to be unambiguously determined. Classification schemes are used to identify which echoes can be modeled as underdense meteors, overdense meteors, or other potentially non-meteor echoes. Finally, based on the proposed classification scheme, this thesis examines several temperature estimation methods for both underdense and overdense echoes and discusses the current issues in this area. Preliminary results from a newly installed meteor radar at Poker Flat Research Range are also presented.
  • A novel virtual reality-based system for remote environmental monitoring and control using an activity modulated wireless sensor network

    Montz, Benjamin; Raskovic, Dejan; Mayer, Charles E.; Thorsen, Denise (2019-08)
    The ability to monitor and control a home environment remotely has improved considerably in recent years due to improvements in the computational power, reduction in physical size, reduced implementation cost, and widespread use of both wireless sensor networks and smart home systems. This thesis presents a remote environment management system that integrated a custom wireless sensor network that monitored environmental factors in multiple locations, a smart system that controlled those factors, and a virtual reality system that functioned as a remote interface with the environment. The resulting system enabled a user to efficiently interact with a distant environment using an immersive virtual reality experience. The user was able to interact with the remote environments by issuing voice commands, performing hand gestures, and interacting with virtual objects. This type of system has applications in many fields ranging from healthcare to the industrial sector. The case study system that was designed in this thesis monitored and controlled the environments of several rooms in a home. A novel approach to modulating the activity of the wireless sensor network was implemented in this system. The rate at which the sensor nodes collected and transmitted data was modulated based on the visibility of the virtual objects called VSNs. These virtual sensor nodes displayed the sensor node measurements in virtual reality. This method was expanded upon using a motion prediction algorithm that was used to predict if the virtual sensor nodes were going to be visible to the user. This prediction was then used to modulate the activity of the wireless sensor network. These activity modulation algorithms were used to reduce the power consumption of the wireless sensor network and thus increasing its operational lifespan, while simultaneously reducing unnecessary RF signals in the environment that can interfere with the operation of other wireless systems. These algorithms would be crucial for systems monitoring complex sensor-rich environments where reducing the data transmitted and extending the system's lifespan was paramount, such as managing the environments of many rooms in a large industrial park or controlling the environments of spacecraft from Mission Control on Earth.
  • Intelligent traffic monitoring and control system

    Ch, Nabil Al Nahin; Raskovic, Dejan; Thorsen, Denise; Hatfield, Michael (2019-08)
    This thesis presents an intelligent system for monitoring and controlling traffic by sensing vehicles' attributes and using communication between vehicles and roadside infrastructures. The goal of this system is to improve the safety of the commuters and help the drivers in making better decisions by providing them with additional information about the traffic conditions. A prototype system consisting of a roadside unit (RSU) and an on-board unit (OBU) was developed to test the functionalities of the proposed system. The RSU consists of sensors for detecting vehicles and estimating their attributes and a radio for communicating with the OBU. The OBU also has a radio for communication purpose. Afterward, a vehicle was used to test the functionalities of the system and the communication between OBU and RSU was evaluated by emulating the presence of a vehicle. A protocol for exchanging messages between the RSU and the OBU was developed to support effective communication. The efficiency of the communication process was further improved by varying the transmission range of different messages. A format for the message was proposed to convey all the necessary information efficiently. The process of collecting vehicle data, processing them and extracting useful information from the data was discussed here along with some limitations of the proposed system.
  • Modeling and analysis of geothermal organic rankine cycle turbines coupled with asynchronous generators as a primary power source in islanded microgrids

    Green, Nathan; Wies, Richard; Huang, Daisy; Shirazi, Mariko (2019-05)
    Local renewable resources, such as geothermal hot springs, are being explored as prime electric power and heat sources in remote permanently islanded microgrids, and in some cases these renewable resources have already been implemented. In these types of remote areas, diesel electric generation is typically the prime source of power, even in areas where alternative resources are readily available, despite the high fuel cost due to transportation. This thesis shows that geothermal hot springs, when locally available, can provide primary power for these remote microgrids with temperatures as low as 20°C below the boiling point of water. The geothermal heat can be converted to electrical energy using an organic Rankine cycle turbine in combination with a self-excited induction generator. A steady-state energy balance model has been developed using MATLAB® and Simulink® for simulating greenfield and brownfield geothermal microgrids at Pilgrim Hot Springs, Alaska and Bergstagir, Iceland, respectively, to demonstrate viability of this microgrid design. The results of the simulations have shown that modest loads can be primarily powered off of these low temperature geothermal organic Rankine cycles over long time scales. As expected, more power is available during colder months when sink temperatures are lower, thus increasing the temperature differential. More research is needed to examine system response over shorter time scale transients, which are beyond the scope of this work.
  • FRAM based low power systems for low duty cycle applications

    Gossel, Cody A.; Raskovic, Dejan; Thorsen, Denise; Sonwalkar, Vikas (2019-05)
    Ferro-Electric Random Access Memory (FRAM) is a leap forward in non-volatile data storage technology for embedded systems. It allows for persistent storage without any power consumption, fulfilling the same role as flash memory. FRAM, however, provides several major advantages over flash memory, which can be leveraged to substantially reduce sleep current in a device. In applications where most of the time is spent sleeping these reductions can have a large impact on the average current. With careful design sleep currents as low as 72 nA have been demonstrated. A lower current consumption allows for more flexibility in deploying the device; smaller batteries or alternative power sources can be considered, and operating life can be extended. FRAM is not appropriate for every situation and there are some considerations to obtain the maximum benefit from its use. An MSP430FR2311 microcontroller is used to measure the performance of the FRAM and how to structure a program to achieve the lowest power consumption. Clock speed and instruction caching in particular have a large effect on the power consumption and tests are performed to quantify their effect. Two case studies are considered, a feedback control system and a data logger. Both cases involve large amounts of data writes and allow for the effects of the FRAM to be easily observed. Expected battery life is determined for each case when the sample rate is varied, suggesting that average operating current for the two solutions will nearly converge when the sampling period exceeds 1000 s. For sampling periods on the order of one second operating current can be reduced from 15.4 μA to 730 nA by utilizing FRAM in lieu of flash.
  • A two-layer energy-efficient wireless sensor network for precision agriculture applications

    Ntarugera, Osiris V.; Raskovic, Dejan; Mayer, Charles E.; Hatfield, Michael (2018-12)
    The agriculture industry has benefited from the recent technological evolution; for example, farmers now use satellite images to monitor large fields. The use of technology in agriculture, generally referred to as Precision Agriculture, has attracted a lot of research interest from electrical engineers. One particular area of Precision Agriculture is the application of embedded systems in monitoring large crop fields. Sensor nodes are placed at various locations in the field where they measure different parameters, such as temperature and soil moisture. The collected measurements are sent to a central hub outside of the field where they can be further processed and displayed for the farmers to make appropriate decisions. From the farmers' perspective, this kind of wireless sensor network (WSN) is a cost-effective solution that allows them to gather accurate information about their crops in real time and significantly improve production. To scientists, it provides invaluable information that can help them improve farming processes or even develop new crop varieties. From the embedded systems stand-point however, such a network poses several challenges, mainly battery life and network lifetime. Battery life is a serious challenge because nodes are scattered in the field and it would be labor intensive and expensive to replace their batteries. It is important to keep nodes alive because dead nodes not only fail to collect data but they also fail to relay packets from other active nodes. Radio communication draws most of the node's battery in WSN, so most energy saving techniques revolve around careful management of the radio. In this study, we focus on routing protocols that maximize the lifetime of the network. Most researchers have suggested various routing schemes to minimize battery consumption by finding the shortest path to a hub; however, when looking at the network as a whole, this approach may not be ideal. We present a lifetime-maximizing routing scheme that uses a cost function to distribute the traffic load among all nodes and to spare those with low remaining energy. The cost function being essential to our algorithm, we evaluate the impact of different types of cost function on the network lifetime. Lastly, we evaluate the impact of link quality in the cost function. Simulation results show that the power cost function has the best performance and that link quality can improve network lifetime. Another major contribution of this research is the design of a test framework that can be used to evaluate other routing protocols. In order to evaluate our routing protocol, we created a WSN simulation in Castalia. The simulation and the routing protocol are highly parametric and with minor modifications, users can experiment with new protocols or variations of ours. Using our platform can save users a lot of time and trouble, especially those unfamiliar with simulation tools, hence allowing them to focus their efforts on their protocol.
  • Design of a multichannel outdoor data logger for precise temperature measurements

    Dolgikh, Kirill; Raskovic, Dejan; Romanovsky, Vladimir; Thorsen, Denise (2018-12)
    In this thesis, we present the prototype of a multichannel data logger (the Logger) for precise temperature measurements. Its intended application is to take soil (permafrost) temperature measurements with a thermistor probe or thermistor string. However, its hardware and firmware architectures are quite flexible, so it can be used in other applications. The Logger has 16 channels. In addition to the internal memory, it supports microSD-cards of up to 2 GB, which allows it to store up to 41.9 million measurements for each channel. The Logger's estimated battery life is 8 years when making measurements once per hour. It has a radio transmitter, which will allow it to download data wirelessly and potentially participate in a wireless sensor network once the appropriate firmware is developed. Currently, only the communication protocol with the radio is implemented, while the radio-to-radio protocol is under development. The Logger is small - only 6 x 1 x 1 inches and the final product will be even smaller. Components are rated down to -40° C and the Logger successfully passed testing at -30° C. After the extensive testing to ensure performance it has been shown that the Logger outperforms the Campbell Scientific, Inc. CR1000 logger and exceeds the design requirements. Measured temperature resolution of the Logger is below 2.5 mK in the entire temperature range. The Logger's equivalent temperature accuracy, which was determined using a known resistive input, is below 10 mK within -25° C to 40° C and below 20 mK elsewhere. The developed calibration technique provides the equivalent accuracy below 0.3 mK within -40° C to 40° C. To provide an accuracy of ±0.01° C when making temperature measurements with thermistors, the Logger should be calibrated against a thermometer that has been calibrated as a secondary standard, which will be done in the future.
  • Cubesat Attitude Control Utilizing Low-Power Magnetic Torquers & A Magnetometer

    Mentch, Donald B.; Thorsen, Denise (2011)
    The CubeSat Project has lowered development time and costs associated with university satellite missions that conform to their 10 centimeter cube design specification. Providing attitude control to a spacecraft, of such small volume, with a very limited power budget has been a challenge around the world. This work describes the development of an attitude control system based on a very low-power magnetic torquer used in conjunction with a magnetometer. This will be the first flight use of this torquer which is composed of a hard magnetic material wrapped inside of a solenoid. By discharging a capacitor through the solenoid, the magnetic dipole moment of this permanent magnet can be reversed. The completed attitude control system will make the first use of the low-power magnetic torquer to arrest satellite tip-off rates. It will then make the first known use of a dual axis magnetic dipole moment bias algorithm to achieve three-axis attitude alignment. The complete system is standalone for high inclination orbits, and will align the spacecraft to within 5 degrees of ram, nadir, and local vertical, without any requirement for attitude determination. The system arrests tip-off rates of up to 5� per second (in all 3 axes) for a satellite in a 600 kilometer polar orbit expending 0.56 milliwatts of power. Once in the proper alignment, it utilizes 0.028 milliwatts to maintain it. The system will function for low inclination orbits with the addition of a gravity boom. The system utilizes the magnetometer to calculate spacecraft body rates. This is the only known use of a magnetometer to directly measure spacecraft body rates without prior knowledge of spacecraft attitude.
  • Role Of Antennas And The Propagation Channel On The Performance Of An Ultra Wide Band (Uwb) Communication System

    Venkatasubramanian, Arun; Sonwalkar, Vikas (2007)
    The objectives of this dissertation are to experimentally and numerically quantify the effect of antennas and the propagation channel on the performance of an Ultra Wide Band (UWB) receiver. This work has led to the following new results: (1) the variation in the time duration of the impulse response of the oval dipole in the vertical plane is within 5% up to an angle theta = 60° off the broadside direction (theta = 90°); at larger angles a factor of six elongation in the time duration of the impulse response along the antenna axis (theta = 0°) is observed, (2) for an axial ratio of 0.5, the oval dipole has a Voltage Standing Wave Ratio (VSWR) of 2:1 (~11% reflection coefficient) in a 3.1 GHz bandwidth with a lower cut off frequency of 2.8 GHz; for an axial ratio of 2.0 this scales to 0.5 GHz bandwidth with a lower cut off frequency of 1.75 GHz, (3) a new theoretical model has been developed for UWB pulse propagation over the ground which takes into account the geometrical properties of the propagation channel (such as the heights of the transmitter (h1) and the receiver (h2) over the ground) and the nature of the radiated UWB pulse (such as pulse duration (taup) and cycle time (tauc)), (4) an improvement in bit error rate by up to a factor of 100 can be achieved for a matched filter receiver by careful orientation of the transmitting and the receiving oval dipole antennas used in the measurements presented in this dissertation.
  • Development Of Resonance Fluorescence Lidar For Studies Of The Aurora

    Su, Liguo; Collins, Richard L. (2007)
    In this thesis I present resonance fluorescence lidar studies of the middle and upper atmosphere. I focus on two specific applications; lidar measurements of heat fluxes in the mesosphere, and lidar measurements of auroral nitrogen ions in the thermosphere. In the heat flux study, I determine the limitations in state-of-the-art sodium Doppler wind-temperature lidar measurements. I conduct statistical analysis of current lidar measurements using analytical and Monte Carlo techniques and extend them to consider future measurements. I find that the expected biases for summertime flux measurements in polar regions will be larger than the geophysical values of the fluxes. In the nitrogen resonance lidar study, I conduct a simulation of the measurements under realistic auroral conditions and found that current lidar systems should be able to make statistically significant measurements of the nitrogen profile at a resolution of 6 km and 300 s. I develop a prototype nitrogen resonance lidar system operating at 390 nm. This lidar system is based on an existing dye laser-based iron resonance lidar system that operates at 372 nm. I designed and implemented a tuning control system that allows 1 pm resolution in the laser tuning. I made a set of field measurements and found that the performance of the prototype lidar was less than expected. I conduct an engineering analysis of the measurements and conclude that the lower than expected performance is due to the lasing characteristics of the dye laser.
  • Developing Computer Models To Study The Effect Of Outdoor Air Quality On Indoor Air For The Purpose Of Enhancing Indoor Air Quality

    Marsik, Tomas; Johnson, Ron (2007)
    People in developed countries spend the majority of their time indoors. Therefore, studying the effect of outdoor air quality on indoor air is of a great importance to human health. This thesis presents several dynamic computer models that were developed to study this effect. They estimate indoor pollutant levels based on outdoor levels, ventilation rate, and other factors. Also, an analysis method is presented that allows for quantifying the effect of outdoor air quality on indoor air at a given building based on measured real-time outdoor and indoor pollutant levels. An important part of this method is separating the measured indoor level into two components - a component caused by indoor sources and a component caused by pollutants penetrating from outdoors. This separation is accomplished using a dynamic model, which, unlike some other methods, also allows for processing transient samples and thus simplifies the needed measurements. Outdoor and indoor pollutant levels were measured at eight buildings in Fairbanks, Alaska and the developed method was used to analyze the data. The main focus was on fine particulate matter (PM2.5) and carbon monoxide (CO) - the pollutants of major concern in Fairbanks. The effective penetration efficiency for PM2.5 ranged from 0.16 to 0.69, and was close to unity for CO. The outdoor generated PM2.5 was responsible in average for about 67% of the indoor PM2.5 in residences, and close to 100% in office environments. These results imply that reducing outdoor pollution can have significant health benefits even for people spending the majority of their time indoors. An air-quality control algorithm for a Heating, Ventilation, and Air Conditioning (HVAC) system was developed and tested using one of the models. This algorithm was shown to reduce indoor PM2.5 levels by 65%. Another model was used to study various ventilation options for a typical Fairbanks home with respect to indoor air quality, energy consumption, overall economy, and environmental impact. Using a Heat Recovery Ventilator (HRV) with an additional filter was shown to be the best option. Another model was successfully used to address key factors for radon mitigation in a home located in a radon-prone area.
  • Speed -Sensorless Estimation And Position Control Of Induction Motors For Motion Control Applications

    Barut, Murat; Bogosyan, Seta; Hawkins, Joseph G.; Wies, Richard W.; Bracio, Boris (2006)
    High performance sensorless position control of induction motors (IMs) calls for estimation and control schemes which offer solutions to parameter uncertainties as well as to difficulties involved with accurate flux and velocity estimation at very low and zero speed. In this thesis, novel control and estimation methods have been developed to address these challenges. The proposed estimation algorithms are designed to minimize estimation error in both transient and steady-state over a wide velocity range, including very low and persistent zero speed operation. To this aim, initially single Extended Kalman Filter (EKF) algorithms are designed to estimate the flux, load torque, and velocity, as well as the rotor, Rr' or stator, Rs resistances. The temperature and frequency related variations of these parameters are well-known challenges in the estimation and control of IMs, and are subject to ongoing research. To further improve estimation and control performance in this thesis, a novel EKF approach is also developed which can achieve the simultaneous estimation of R r' and Rs for the first time in the sensorless IM control literature. The so-called Switching and Braided EKF algorithms are tested through experiments conducted under challenging parameter variations over a wide speed range, including under persistent operation at zero speed. Finally, in this thesis, a sensorless position control method is also designed using a new sliding mode controller (SMC) with reduced chattering. The results obtained with the proposed control and estimation schemes appear to be very compatible and many times superior to existing literature results for sensorless control of IMs in the very low and zero speed range. The developed estimation and control schemes could also be used with a variety of the sensorless speed and position control applications, which are challenged by a high number of parameter uncertainties.
  • Hybrid Electric Power Systems In Remote Arctic Villages: Economic And Environmental Analysis For Monitoring, Optimization, And Control

    Agrawal, Ashish N.; Wies, Richard (2006)
    The need for energy-efficient and reliable electric power in remote arctic communities of Alaska is a driving force for research in this work. Increasing oil prices, high transportation costs for fuels, and new environmental standards have forced many utilities to explore hybrid energy systems in an attempt to reduce the cost of electricity (COE). This research involves the development of a stand-alone hybrid power system model using MATLABRTM SimulinkRTM for synthesizing the power system data and performing the economic and environmental analysis of remote arctic power systems. The hybrid model consists of diesel electric generators (DEGs), a battery bank, a photovoltaic (PV) array, and wind turbine generators (WTGs). The economic part of the model is used to study the sensitivity analysis of fuel cost and the investment rate on the COE, the life cycle cost (LCC) of the system, and the payback time of the system. The environmental part of the model calculates the level of various pollutants including carbon dioxide (CO2), nitrogen oxides (NOx), and the particulate matter (PM10). The environmental analyses part of the model also calculates the avoided cost of various pollutants. The developed model was used to study the economics and environmental impacts of a stand-alone DEG system installed at the University of Alaska Fairbanks Energy Center, the wind-diesel-battery hybrid power system installed at Wales Village, Alaska, and the PV-diesel-battery hybrid power system installed at Lime Village, Alaska. The model was also used to predict the performance of a designed PV-wind-diesel-battery system for Kongiganak Village. The results obtained from the SimulinkRTM model were in close agreement with those predicted by the Hybrid Optimization Model for Electric Renewables (HOMER) software developed at National Renewable Energy Laboratory (NREL).
  • Modeling and exploring battery management strategies for use of LiCoO₂ lithium polymer cells in cold climates

    Thompson, Isaac D.; Wies, Richard; Raskovic, Dejan; Lawlor, Orion (2018-05)
    As the use of batteries to power vehicles becomes more common, a robust battery management system becomes necessary to monitor and maintain the batteries. Cold weather places a further burden on this system especially in small electric vehicles such as snowmobiles where it is desirable to use every bit of available energy from the battery cells. The problem with current battery management technology is that use of batteries in cold temperatures is often not addressed. The objective of this research was to develop an appropriate model of a lithium polymer cell with a cathode comprised of LiCoO₂ and develop an optimized charge/discharge method taking into account the effects of extreme cold weather and cell state of charge imbalance. A cell model was adapted and tuned that accurately captures the dynamics of a lithium polymer cell when discharged at temperatures below freezing. The model results were verified against cells discharged in an environmental chamber, which allowed accurate control of ambient temperature. Multiple scenarios were explored, looking at the effects of ambient temperature, cell initial temperature, internal heating, battery pack insulation, and how rapidly the cells were discharged. The results of the optimized battery management strategies showed improvements in the energy delivery capability of lithium polymer battery packs for small vehicles operating in extreme cold environments. In addition, this research extended the LiCoO₂ model down to -20 °C using validated data, showed that perceived cell capacity loss at low temperatures is primarily due to increased internal resistance, demonstrated that measured cell terminal voltage can rise under load at low temperatures, and showed that increasing the capacity of a battery pack has a better than linear gain in usable energy versus increased battery capacity. I.e., doubling battery pack capacity will more than double the useable range of the vehicle.
  • A novel low-cost autonomous 3D LIDAR system

    Dial, Ryker L.; Bogosyan, Seta; Hatfield, Michael; Lawlor, Orion (2018-05)
    To aid in humanity's efforts to colonize alien worlds, NASA's Robotic Mining Competition pits universities against one another to design autonomous mining robots that can extract the materials necessary for producing oxygen, water, fuel, and infrastructure. To mine autonomously on the uneven terrain, the robot must be able to produce a 3D map of its surroundings and navigate around obstacles. However, sensors that can be used for 3D mapping are typically expensive, have high computational requirements, and/or are designed primarily for indoor use. This thesis describes the creation of a novel low-cost 3D mapping system utilizing a pair of rotating LIDAR sensors, attached to a mobile testing platform. Also, the use of this system for 3D obstacle detection and navigation is shown. Finally, the use of deep learning to improve the scanning efficiency of the sensors is investigated.
  • A Sounding Rocket Attitude Determination Algorithm Suitable For Implementation Using Low Cost Sensors

    Charlton, Mark Christopher; Hawkins, Joseph G. (2003)
    The development of low-cost sensors has generated a corresponding movement to integrate them into many different applications. One such application is determining the rotational attitude of an object. Since many of these low-cost sensors are less accurate than their more expensive counterparts, their noisy measurements must be filtered to obtain optimum results. This work describes the development, testing, and evaluation of four filtering algorithms for the nonlinear sounding rocket attitude determination problem. Sun sensor, magnetometer, and rate sensor measurements are simulated. A quatenion formulation is used to avoid singularity problems associated with Euler angles and other three-parameter approaches. Prior to filtering, Gauss-Newton error minimization is used to reduce the six reference vector components to four quaternion components that minimize a quadratic error function. Two of the algorithms are based on the traditional extended Kalman filter (EKF) and two are based on the recently developed unscented Kalman filter (UKF). One of each incorporates rate measurements, while the others rely on differencing quaternions. All incorporate a simplified process model for state propagation allowing the algorithms to be applied to rockets with different physical characteristics, or even to other platforms. Simulated data are used to develop and test the algorithms, and each successfully estimates the attitude motion of the rocket, to varying degrees of accuracy. The UKF-based filter that incorporates rate sensor measurements demonstrates a clear performance advantage over both EKFs and the UKF without rate measurements. This is due to its superior mean and covariance propagation characteristics and the fact that differencing generates noisier rates than measuring. For one sample case, the "pointing accuracy" of the rocket spin axis is improved by approximately 39 percent over the EKF that uses rate measurements and by 40 percent over the UKF without rates. The performance of this UKF-based algorithm is evaluated under other-than-nominal conditions and proves robust with respect to data dropouts, motion other than predicted and over a wide range of sensor accuracies. This UKF-based algorithm provides a viable low cost alternative to the expensive attitude determination systems currently employed on sounding rockets.
  • Design And Implementation Of A Meteor Tracking Retrofit System For The Hf Radar At Kodiak Island, Alaska

    Parris, Richard Todd; Bristow, William (2003)
    The HF radar at Kodiak Island, Alaska, is part of the SuperDARN network of radars, and was originally designed to detect echoes from ionospheric field-aligned density irregularities. A new digital receiver has been implemented on the radar to allow provide the capabilities required for increased range resolution detection of meteor echoes. A meteor detection algorithm has also been implemented to detect meteor echoes with a range resolution of 4.5 km. The algorithm measures the velocity, range, and altitude of the echoes. This data can be used to derive information about the meteor region winds. The design and implementation of the receiver, the design and implementation of the meteor detection algorithm, and some meteor region wind data derived from the new system are presented. <p>
  • Small scale implementation of a robotic urban search and rescue network

    Kibler, Steven G. (2012-05)
    With the advancement of robotics technologies, it is now possible to use robots for high risk jobs that have historically been accomplished by humans. One such example is the use of robots for Urban Search and Rescue (USR): finding chemical spills, fires, or human survivors in disaster areas. With the ability to include inexpensive wireless transceivers, it is possible to network numerous robots as part of a swarm that can explore an area much more expeditiously than a single robot can. With the inclusion of wireless capabilities comes the necessity to create a protocol for the communication between robots. Also necessary is the creation of an exploration protocol that allows the network of robots to explore such a building or search area in as little time as possible yet as accurately as possible. This thesis covers the development of such a network of robots, starting with the hardware/software co-design, the individual robots' control mechanisms, and their mapping and communications protocols.
  • Ray tracing applications for high-frequency radar: characterizing artificial layers and background density perturbations in the ionosphere

    Theurer, Timothy E. (2012-08)
    In this thesis a numerical method of calculating ground-scattered power from the results of a ray tracing analysis is presented. The method is based on a conservation of energy approach and offers advantages over an alternative method derived from the radar equation. The improved numerical method is used to investigate two different physical phenomena by comparison with measured ground-scattered power observed by a high-frequency (HF) radar located in Kodiak, AK that is part of the Super Dual Auroral Radar Network (SuperDARN). First, the effects of artificial electron density layers on observed ground scatter is studied through a comparison of simulated and measured power profiles. The results demonstrate that the location and spatial dimensions of artificial layers may be estimated by a comparison of the location and amplitude of simulated and measured power enhancements. Second, a Monte-Carlo simulation method is used to characterize the temporal distribution of ground-scattered power. Random processes including background electron density perturbations, polarization, noise, and sample correlation are modeled in simulation and used to estimate statistical moment profiles. The simulated statistical moment profiles are compared to measured profiles as a means of model verification and to roughly approximate background electron density perturbations in the ionosphere.

View more