Master of Science, The Ohio State University, 2014, Mechanical Engineering

Battery life, cost and weight are some of the most important factors considered while designing battery packs for electrified vehicles. These factors directly affect the appeal of electric vehicles in the market. While, performance, cost and weight can be evaluated at the production and design stage, battery life is a dynamic parameter influenced by a multitude of factors and is hard to accurately predict, often leading to conservative designs with oversized and more expensive battery packs. Expensive batteries and complex, multi-factor aging phenomena ideally would require continuous tracking of the battery state of health. Battery capacity and internal resistance are commonly used to quantify battery state of health, as these metrics translate directly into range and power at the user level. While resistance growth is relatively easy to estimate in a vehicle, capacity fade requires measurements typically done at the laboratory level and conditions never encountered in a vehicle. This thesis aims to develop an algorithm capable of tracking in situ these two parameters throughout the life of battery. By far the most challenging aspects of battery state of health estimation is to only use information available in the vehicle during its normal use, and furthermore, suitable with available on-board computing resources for real-time implementation. To that effect, the `needs and wants' of an ideal in situ capacity estimator were clearly defined at the beginning of this work and algorithms that satisfy all the constraints were developed, tested and validated. This work leverages the experimental results of an aging campaign conducted in out laboratories on a total of 17 cells aged under a variety of realistic operating conditions. A sensitivity analysis of the output of the algorithm was then carried out to assess accuracy of the algorithms in the presence of parameter variations and sensor errors. Next, the separate capacity and resistance estimation algorithms we (open full item for complete abstract)

Committee: Yann Guezennec PhD (Advisor); Giorgio Rizzoni PhD (Committee Member) Subjects: Automotive Engineering; Energy; Mechanical Engineering

Master of Science in Engineering, University of Akron, 2016, Electrical Engineering

An observer-controller pair was designed to regulate the fuel flow rate and the flue-gas oxygen ratio of a combustion boiler. The structure of the observer was a proportional-integral state estimator. The designed controller was composed of a combination of two common controller structures: state-feedback with reference tracking and proportional-integral-derivative(PID). A discrete-time, linear state-space model of the combustion system was developed such that the linear controller and observer could be designed. This required establishing separate models pertaining to the combustion process, actuators, and sensors. The complete model of the combustion system incorporated all three models. The combustion model, which related the flue-gas oxygen ratio to the fuel and oxygen flow rates, was obtained using the mathematical formulas corresponding to combustion of natural gas. The actuators were modeled using measured fuel and oxygen flow rate data for various actuator signals, and fitting the data to a parametric model. The established nonlinear models for the combustion process and actuators required linearization about a specified operating point. The sensors model was then obtained using the predictive error identification technique based on batch input-output data. For the acquired model of the combustion system, a linear quadratic regulator was used to calculate the optimal state feedback gain. The classical controller gains were determined by tuning the gains and evaluating the simulation of the closed-loop response. Computer-aided simulations provided evidence that the controller and state estimator could regulate the desired set point in the presence of moderate disturbances. The observer-controller pair was implemented and verified on an experimental boiler system by means of an embedded system. Even in the presence of a disturbance resulting from a 50% blockage of the surface area of the air intake duct, the closed-loop system was capable of regulating t (open full item for complete abstract)

Committee: Nathan Ida Dr. (Advisor); Robert Veillette Dr. (Committee Member); Kye-Shin Lee Dr. (Committee Member) Subjects: Electrical Engineering; Engineering

Doctor of Philosophy, The Ohio State University, 2015, Mechanical Engineering

Vehicle electrification continues to be a key topic of interest for automotive manufacturers, in an effort to reduce the usage of fossil fuel energy and improve vehicle efficiency. Lithium-ion batteries are currently the technology of choice for hybrid and electric vehicles due to their decreasing cost and improved power and energy density over traditional lead-acid or nickel-metal-hydride batteries. In particular, batteries with composite electrodes have seen increased use in automotive applications due to their ability to balance energy density, power density, and cost by adjusting the amount of each material within the electrode. However, this improved performance introduces new challenges to ensure the battery pack operates safely, reliably, and durably. The vehicle's battery management system (BMS) is designed to meet these challenges, in part, by estimating the battery state of charge (SOC) and state of health (SOH). Knowledge of SOC allows the BMS to predict the available instantaneous power while ensuring the battery is operating within safe limits. As batteries age, they lose capacity and the ability to deliver power. Therefore, tracking the battery SOH is necessary to maintain accurate estimates of SOC and power throughout the battery life and give an accurate miles-to-empty metric to the driver. Recently, increased attention has been given to electrochemical models for SOC and SOH estimation, over traditional circuit models. Electrochemical models based on first principles have the potential to more accurately predict cell performance, as well as provide more information about the internal battery states. State of health estimation algorithms that do not use electrochemical-based models may have more difficulty maintaining an accurate battery model as the cell ages under varying degradation modes such as lithium consumption at the solid-electrolyte interface or active material dissolution. However, efforts to validate electrochemical model-based state (open full item for complete abstract)

Committee: Giorgio Rizzoni (Advisor); Canova Marcello (Committee Member); Guezennec Yann (Committee Member); Conlisk Terry (Committee Member) Subjects: Mechanical Engineering

PhD, University of Cincinnati, 2024, Engineering and Applied Science: Civil Engineering

Ramp metering is a freeway traffic management technique designed to regulate the vehicle flow merging onto the freeway, thereby improving the freeway throughput and minimizing disruptions. However, determining dynamic metering rate is always dependent upon accurate identification of mainline traffic and on-ramp arrival demand, all of which are not always accurately captured by traditional detection technologies (such as fixed-point loops). On the other hand, if an on-ramp experiences high arrival demand, queuing vehicles on the on-ramp may cause back propagation to the upstream local street, leading to traffic congestion across the local street network. To address these issues, this research proposes to create a coordinated control strategy to synchronize the ramp metering control schemes with the signalized control system at the upstream intersection. The implementation of such an integrated system necessitates robust intercommunication of detailed traffic state information between different roadway segments, thereby imposing more requirements on data acquisition. To solve this problem, recent research has begun to explore the potential of using data from Connected Vehicles (CVs) and/or Automated Vehicles (AVs), or simply termed CAVs in the research, to support the implementation of the integrated strategy. However, these previous studies often assume a 100% market penetration rate (MPR) of CAVs and global communication capabilities, which simplifies the analysis of CAV driving behavior and overlooks the complexities of traffic data dissemination. Hence, these optimal solutions cannot be guaranteed effective under a real-world mixed automated traffic environment, which should be a more common scenario in recent years. Moreover, due to the inherent complexity of this issue, existing research typically concentrates on either adaptive intersection signal control or ramp metering separately. The coordinated control between these two elements has not been consid (open full item for complete abstract)

Committee: Heng Wei Ph.D. (Committee Chair); John Ash Ph.D. (Committee Member); Zhixia Li Ph.D. (Committee Member); Xuefu Zhou Ph.D. (Committee Member) Subjects: Transportation

Doctor of Philosophy, The Ohio State University, 2023, Mechanical Engineering

Metal Powder Bed Fusion (PBF) is a type of additive manufacturing process that incrementally builds parts by fusing 2D slices of the geometry into layers of metal powder, using either a laser (L-PBF) or electron beam (E-PBF), and is among the emerging technologies of Industry 4.0. The predominant quality control methods for PBF are pre- and post-process tests of the part and materials, which are inefficient because they cannot prematurely halt malfunctioning builds as errors occur. Live (In-situ) monitoring of the PBF process for defects, in which the defects are oftentimes due to improper thermal management, and in-situ control of the PBF process to ensure good thermal management, are areas of active research. These efforts are currently dominated by constructing data-driven PBF thermal models and using the corresponding estimations to judge the current thermal state (process monitoring) and to decide correction factors (process control). Collecting the data for training these methods is costly and renders them inflexible with respect to changes in part design and processing conditions, because they do not offer guaranteed performance in environments that lay outside the scope of the training data. Since PBF exists to increase production flexibility, lessening this dependency on training data is essential. To address this challenge, we demonstrate the efficacy of applying training data-free algorithms to the in-situ PBF thermal process monitoring and control problems. Our process monitoring algorithm is the Ensemble Kalman Filter (EnKF), which is a type of state estimator that uses a particle swarm to generate self-tuned, approximately 2-norm optimal, model-based estimates of the relevant process signatures. Here, the signatures are all temperatures in the PBF build. Our control algorithm is Model Predictive Control (MPC), which uses model-based predictions of future process signatures (here, temperatures) to determine a sequence of process inputs that reg (open full item for complete abstract)

Committee: David Hoelzle (Advisor); Andrew Gillman (Committee Member); Andrea Serrani (Committee Member); Mrinal Kumar (Committee Member); Michael Groeber (Committee Member) Subjects: Mechanical Engineering

PhD, University of Cincinnati, 2022, Engineering and Applied Science: Aerospace Engineering

In this dissertation, the main focus is on, resilience and enhancement in the performance of autonomous multi-agent vehicles and transportation systems using machine learning techniques. Initially, we consider a vehicle platooning problem with a bi-directional controller. Here we study the effects of control based attack on string stability and solutions to detect, identify and mitigate the attack. Using Deep Convolutional Neural Networks (DCNN) with Gramian Angular Fields as pre-processing technique we detect and identify the ad- versarial vehicle in the platoon with only local information from the vehicles. Also, we provide mitigation strategy using Routh Hurwitz stability criterion. Next, we examine multiple ground robots performing cooperative localization using an Extended Infor- mation Filter (EIF). In this problem, we investigate the effects of Stealthy False Data Injection (SFDI) attack on the state estimator. Here, we provide various distributed deep learning strategies and specially One-Shot learning to detect these SFDI attacks and generalize it to n number of robots in the environment. In the next problem we solve a problem related to shoulder drop-offs on highways. Since shoulder drop-offs are one of the main reasons for accidents they have to be repaired and maintained by Department of Transportation which they currently to it manually and it is very challenging for visual eye. So, we automate this process using a low cost LIDAR called Livox and a camera to predict these drop-off levels using state of the art deep learning methods like PointNet, Vision Transformers and ConvNets. In the end we have solved problems related to the enhancement of state estimators and control algorithms for multi-agent vehicles. In the first problem, we have two Unmanned Aerial Vehicles (UAV) geo-localizing multiple ground targets using gimbal camera. The main focus here is to reduce the uncertainty and es (open full item for complete abstract)

Committee: Rajnikant Sharma Ph.D. (Committee Member); Kelly Cohen Ph.D. (Committee Member); Boyang Wang (Committee Member); Ali Minai Ph.D. (Committee Member); Ou Ma Ph.D. (Committee Member) Subjects: Artificial Intelligence

Doctor of Philosophy, The Ohio State University, 2021, Electrical and Computer Engineering

Traffic accidents result in a high number of fatalities each year. This brings up the importance of developing Autonomous Vehicles (AV) and Advanced Driver Assistance Systems (ADAS), due to their potential of increasing traffic safety by reducing vehicle crashes caused by driver errors. It could also be helpful to deploy the intelligent transportation systems (ITS) in different traffic scenarios to increase the efficiency of traffic flow and enlarge the traffic capacity. Planning and control of the autonomous vehicles, the two essential modules in autonomous driving, are still facing severe challenges in adapting to various traffic scenarios and complex environments. The planning and decision making of vehicles in urban traffic environment are still a big challenge for autonomous vehicles due to its complexity and uncertainties. Hence it is necessary to develop decision making and planning algorithms for vehicles in urban traffic, especially in intersections. Also, velocity profile planning for autonomous vehicles is also required based on various requirements according to the environment. Additionally, a convenient method for testing and validating the developed algorithms is also required. Hence a good simulation environment is important in the field of autonomous vehicles. This dissertation contributes to planning and decision making of autonomous vehicles in urban traffic scenarios as well as developing a way of generating realistic simulation environments as test beds to validate developed autonomous driving algorithms. Decision making methods and planning methods for autonomous shuttles and autonomous vehicles in urban traffic are proposed. A rule based decision maker working for last mile problem is introduced for an autonomous shuttle so that the autonomous shuttle can deal with typical traffic on designated routes. Then to deal with complex and uncertain urban traffic scenarios when the ego autonomous vehicles doesn't have full observability over other v (open full item for complete abstract)

Committee: Levent Guvenc (Advisor); Bilin Aksun-Guvenc (Committee Member); Keith Redmill A. (Committee Member); Vadim Utkin I. (Committee Member) Subjects: Electrical Engineering

Master of Science, The Ohio State University, 2021, Mechanical Engineering

Bipedal robots are becoming more popular in performing tasks in an environment that is designed for humans. For this purpose, most bipedal robots are equipped with various sensors to sense the robot's environment. From the measurements of the sensors, a perception system is implemented to translate and convert the raw data into a meaningful format corresponding to the tasks and also provide safety for humans, properties in the environment as well as the robot itself. This thesis presents the implementation of a perception system using various sensors available to a bipedal robot, Digit, to obtain objectively useful information of the environment as well as the state of the robot itself. Various methods of data processing were applied to available sensor measurements, then a mapping algorithm was implemented to generate a 3D model of the environment. Simultaneous localization and mapping (SLAM) algorithm was also implemented to perform mapping and provide odometry for localization in the absence of an external source of odometry. We found that performing SLAM using Light Detection and Ranging sensor (LiDAR) performs exceptionally well on the bipedal robot in closed indoor space. Additionally, state estimation is implemented with Invariant Extended Kalman filter using inertial measurement data and the assumption of contact points to predict the state of the robot over time. The performance of position estimation from Invariant Extended Kalman filter and odometry from LiDAR SLAM is compared with the default state estimator from Digit itself which are demonstrated through an experiment with ground truth reference.

Committee: Keith Redmill (Committee Member); Ayonga Hereid (Advisor) Subjects: Mechanical Engineering; Robotics

MS, University of Cincinnati, 2021, Engineering and Applied Science: Mechanical Engineering

This thesis aims to develop a modular sensor fusion framework based on Unscented Kalman Filter (UKF) that estimates the 6-DOF (Degree-of-Freedom) pose of a quadrotor UAV using the dynamics derived from Newton's laws of motion and localization systems such as visual-inertial odometry (VIO) system and GPU-IMU (GI) system for both indoor and outdoor environments. Micro Aerial Vehicles (MAVs), especially quadrotors, are gaining attention for applications such as package delivery, inspection, emergency response, and search and rescue missions. State estimation becomes very crucial for carrying out both remotely-controlled and autonomous operations. This problem, known as localization, has been explored in the literature using a wider range of sensors such as radars, lidars, cameras, IMUs, and global positioning systems (GPS). In outdoor environments, GPS provides a reliable source of information for carrying out localization. Onboard sensing means such as cameras, IMUs, radars, or lidars are used for indoor environments. The localization problem becomes challenging for indoor environments for several reasons: i) difficulty in processing information from these sensors; ii) most onboard sensors are prone to erroneous measurements; and iii) need specific environmental conditions to satisfy (such as the presence of unique features in the environment, adequate lighting). This thesis focuses on improving localization by incorporating the UAV dynamics into the estimation alongside various localization sensors. We used a monocular camera and an IMU as sensing devices for indoor localization while GPS and IMU for outdoor localization. In recent times, VIO has been explored using different approaches. However, few research works exploit the quadrotor Newtonian dynamics and the known thrust and torque inputs. Incorporating the information from the dynamics with known control inputs provide robust state estimation. Hence, this thesis aims to estimate the quadrotor UAV 6-DOF pose (open full item for complete abstract)

Committee: Manish Kumar Ph.D. (Committee Chair); David Thompson (Committee Member); Rajnikant Sharma Ph.D. (Committee Member) Subjects: Robots

MS, University of Cincinnati, 2021, Engineering and Applied Science: Mechanical Engineering

Monocular visual odometry has been an active area of research over the past decade owing to its variety of applications in the field of geology, autonomous driving, and mixed reality products. The fundamental challenge in working with monocular data is the absence of absolute depth information, making the odometry extraction process cumbersome. This thesis is focused on the development of a deep learning architecture to create a network for odometry estimation. The thesis particularly focuses on the transformer architecture to solve a complex vision problem. The intuition of the existing language translation network is modified to work on the comparison of images to create a novel data-driven relative state estimation methodology. This network is further modified to develop a novel architecture to create a physically consistent network that could handle images of varying aspect ratios and angles of view. This novel network addresses the pressing issue of such algorithms still being unusable for consumer-based generic purposes and develops a solution to bridge this gap by creating a data pipeline for varying camera models. The network is further tested on the KITTI dataset to ensure conformance with the real world. The results are plotted and a detailed analysis of the generalization of the network through feature representation is discussed.

Committee: Manish Kumar Ph.D. (Committee Chair); Rajnikant Sharma Ph.D. (Committee Member); David Thompson Ph.D. (Committee Member) Subjects: Mechanical Engineering

MS, University of Cincinnati, 2020, Engineering and Applied Science: Mechanical Engineering

Experimental modal analysis (EMA), which is an integral part of vibration analysis deals with finding the dynamic characteristics of a system namely the natural frequencies, damping and modal scaling. This information is crucial to the design of any structure as they would help predict the system response in its operating conditions. EMA is usually performed on an input output data model that is acquired from a structure. There are several methods which operate in the time and frequency domain to evaluate the modal parameters from a meaningful set of experimental data. The traditional polynomial based approaches use least squares methods to arrive at a good estimate of the numerator and the denominator matrix polynomials that can represent the frequency response functions. The modal parameters are then obtained from this mathematical fit of the experimental data. Another approach to this problem is the use of state space models used in controls domain. An nth order linear differential equation can be represented in the state space form with the defining system matrices A, B, C and D. The problem statement here is to fit the experimental data into its defining state space matrices of a suitable order since they would contain all the modal information in them. Numerical simulations for subspace identification (N4SID), developed by Van Overschee and de Moore, is one algorithm that can be used to build a state space model from measured input output data. This thesis work attempts to compare the above mentioned traditional polynomial based approaches to modal analysis with a state space based system identification approach using N4SID. Through these comparisons, the similarities in the description of a transfer function by these two methods are described. It also would serve as a starting point for its reader to compare more state space approaches with the traditional Unified Matrix Polynomial Approach (UMPA).

Committee: Randall Allemang Ph.D. (Committee Chair); Michael Mains M.S. (Committee Member); Allyn Phillips Ph.D. (Committee Member) Subjects: Mechanical Engineering

Master of Science in Mechanical Engineering, Cleveland State University, 2020, Washkewicz College of Engineering

Collegiate athletes rely on their muscles to compete in their respective sports. However, one injury requiring extended time out of competition could lead to muscle atrophy. As a result, athletes may learn to compensate for weakened muscle groups with stronger muscle groups; a change that may be almost undetectable. Consequently, compensating can add unnecessary stress to the musculoskeletal system, leading to reinjury. One way to combat this is by measuring muscle force. However, there are currently no methods to directly measure muscle force, so it must be solved for indirectly. This research aims to explore state estimation with trajectory optimization and a convolutional neural network. Both methods will be used to estimate the trajectories of the state variables and muscle force associated with forearm flexion. To serve as an input to both solution methods, artificial data was generated. This data contained measured trajectories for forearm position, angular velocity, muscle fiber length, muscle activation, and muscle force. In addition, the generated data included artificial sensor signals comprised of an electromyography (EMG) and inertial measurement unit (IMU). For testing, different signal to noise ratios were added to the generated sensor data. The trajectory optimization method was tested using different weight ratios. The results from this simulation study confirm that the tuning parameter should be chosen based on the noise levels present within the data. Moreover, this method of state estimation can accurately and precisely predict state variable trajectories at all noise levels. However, it struggles to predict muscle force when there is noise added to the data. A similar process was conducted to test the neural network; however, the batch size, was the tuning parameter selected for this method. The results from this portion of the simulation study conclude that the convolutional neural network was able to estimate the state variables precise (open full item for complete abstract)

Committee: Antonie van den Bogert (Advisor); Eric Schearer (Committee Member); Majid Rashidi (Committee Member) Subjects: Biomechanics; Mechanical Engineering

Doctor of Philosophy, Case Western Reserve University, 2021, EECS - System and Control Engineering

This dissertation formulates the problem of fault detection and diagnosis of DC electric power systems for the application of autonomous spacecraft. The ability to accurately identify and isolate failures in the electrical power system is critical to ensure the reliability of a spacecraft. This problem becomes more pronounced during deep space missions that lack the ability to monitor from ground control. The current state of electrical power system fault supervision is insufficient to guarantee highly reliable and robust operation. To solve this issue, a combination of model-based and rules-based techniques are used in a hierarchical framework to improve the diagnostic performance of the spacecraft electrical power system. Noise, disturbances, and modeling errors are considered in the design of the method. Practical considerations related to the hardware and software are discussed for the flight application. A wide array of failure types are simulated in a series of experiments to assess the functionality of the design. The experiments showed that the methods used improved the diagnostic capability of the autonomous system while taking into account the limitations attributed to flight software requirements. The significance of this study is to provide a framework capable of advanced diagnostics of an electrical power system with little to no interaction from a human operator.

Committee: Kenneth Loparo (Advisor); Vira Chankong (Committee Member); Kalmesh Mathur (Committee Member); Farhad Kaffashi (Committee Member) Subjects: Electrical Engineering; Energy

Master of Science, The Ohio State University, 2020, Computer Science and Engineering

In recent years, WiFi-based human-computer interaction has achieved significant progress in localization, fall detection, activity recognition applications since the innovation of CSI (Channel State Information). But WiFi sensing for fine-grained activity recognition like hand pose estimation is not yet discovered. In this study, we present a WiFi sensing system that only utilizes commercial off-the-shelf WiFi devices to capture human hand pose. To our knowledge, this is the first system that considers the application of hand pose estimation using CSI. We provide configuration details of data collection, data processing for CSI and image that can be reused for any other WiFi-based sensing research. And we propose a deep learning approach that achieves cross-modal learning from CSI to hand pose labels. Our system collects the CSI signals and 2D images in a time-synchronized manner. The 2D images are used to generate hand pose labels. And the CSI signals are collected from 3 x 3 transmitter and receiver antenna pairs and used as the input to our model. Our model includes 3 different learning targets. Experiment results show that CSI measurements have similar structures to digital images and popular network architecture for hand pose estimation in images can be applied to CSI measurements with slight modification.

Committee: Dong Xuan Dr. (Advisor); Wei-Lun Chao Dr. (Committee Member) Subjects: Computer Science

Doctor of Philosophy, The Ohio State University, 2019, Mechanical Engineering

Tremendous amount of research attention has been paid to the lithium-ion battery as it becomes the technology of choice for powertrain electrification due to its increasing power and energy densities as well as decreasing cost. Despite recent progress, however, battery aging still remains as the key challenge that prevents its wider applications in energy storage. While research in battery aging has been dominated by improving modeling precision and increasing estimation algorithm efficiency, the importance of data implemented for aging assessment has long been overlooked. In other words, there has been no definite answers to which is the favorable (both sensitive and practically existing) data profile for battery aging assessment and how much data from such profile is sufficient to assess battery degradation. Additionally, presently reported research has mostly focused on a certain battery operating condition without taking into account the impacts from various different usage patterns in practical electric vehicles, thus losing credibility in adapting to varying operating scenarios. To address these two critical issues, this study is proposed to identify the favorable data profile for lithium-ion battery aging assessment with consideration of usage patterns in electric vehicles. Both model-based and in-situ experimental approaches have been employed to identify the favorable data profile for aging assessment. For the model-based approach, the electrolyte enhanced single particle model (ESPM) which has been simplified from the porous model while still retain parameters with physical meanings has been selected and parameterized at the beginning of life. The 1C CCCV charging with sufficiently wide SOC range is determined as the favorable data profile through open-loop identification based on aging characterization. Improved EKF for full state estimation of both electrodes is designed to perform the close-loop identification of favorable data profile for aging asses (open full item for complete abstract)

Committee: Mrinal Kumar (Advisor); Ann Co (Committee Member); Lei Cao (Committee Member); Ran Dai (Committee Member) Subjects: Automotive Engineering; Mechanical Engineering

Master of Science in Engineering, University of Akron, 2019, Mechanical Engineering

This thesis illustrates the application of the Extended Kalman filter for online estimation of attitude of a body. The accuracy of controlled attitude largely depends on the performance of the estimation algorithm. In this thesis, the extended Kalman filter (EKF) algorithm consisting of quaternion based state representation is used. The EKF algorithm utilizes gyroscope reading for priori estimation and measurements reading from the accelerometer and the magnetometer to correct the states. In simple terms, the extended Filter is used as the estimation tool by fusing the data from the gyroscope, the accelerometer and the magnetometer. A device that combines the gyroscope, accelerometer and magnetometer is called inertial measurement unit (IMU). The non- accurate scaling, sensor misalignment and non-zero biases of IMU devices are eliminated by proper calibration. The sensors utilized in the estimation have noise and biases which results in propagation of error in time. The noise and biases should be eliminated to get the accurate estimates. In this work, the EKF algorithm with some modification in state equation and in the Kalman filter gain is implemented for both the steady state and the body acceleration conditions. The estimation of the modified EKF is compared with the estimation technique used by VECTORNAV, a well-known commercial IMU. The modified EKF performed well compared to VECTORNAV in steady state condition. However, under body acceleration, the modified EKF did not perform as well as what VECTORNAV did. The attitude estimation technique discussed in this thesis is less expensive and easy compared to those used in missile and aircraft guidance. The algorithm discussed in this thesis can be well implemented in the navigation of robots and drones for home applications.

Committee: Celal Batur (Advisor); Ajay Mahajan (Committee Member); Siamak Farhad (Committee Member) Subjects: Mechanical Engineering

Master of Sciences (Engineering), Case Western Reserve University, 2017, EECS - Electrical Engineering

This thesis explores the use of deep learning for robot localization with applications in re-localizing a mislocalized robot. Seed values for a localization algorithm are assigned based on the interpretation of images. A deep neural network was trained on images acquired in and associated with named regions. In application, the neural net was used to recognize a region based on camera input. By recognizing regions from the camera, the robot can be localized grossly, and subsequently refined with existing techniques. Explorations into different deep neural network topologies and solver types are discussed. A process for gathering training data, training the classifier, and deployment through a robot operating system (ROS) package is provided.

Committee: Wyatt Newman (Advisor); Murat Cavusoglu (Committee Member); Gregory Lee (Committee Member) Subjects: Computer Science; Electrical Engineering; Robotics

Doctor of Philosophy, The Ohio State University, 2014, Mechanical Engineering

This dissertation details the development of a state and position estimator for articulated heavy trucks based entirely on freely available on-board signals. The estimator consists of a quasi-linear vehicle dynamics model, tire cornering stiffness estimator, Kalman filter, and position integrator. Results from testing show that the estimator can provide lane-level (1.5 m) positioning accuracy in urban environments for the duration of typical GPS outages. A hybrid kinematic-dynamic model allows estimation of hitch angle to within half of a degree over the practical range of articulation angles. This presents novel contributions to the state of the art of trailer tire cornering stiffness estimation and hitch angle estimation. Government research has estimated that vehicle-to-vehicle (V2V) collision avoidance systems can address 72% of heavy truck crashes, but this requires localization of the truck and trailer in a variety of environments. Studies have shown that GPS cannot be reliably used for V2V in urban and some suburban environments. This estimator offers a potential supplement to GPS for V2V systems in these environments. Moreover, the current V2V messaging framework does not include an estimate of hitch angle. This can lead to missed warnings and false positives when the implicit assumption of zero hitch angle is grossly violated, such as turning at an intersection. Results from this research indicate that a reliable estimate can be provided without the addition of new sensors.

Committee: Dennis Guenther (Advisor); Gary Heydinger (Committee Member); Junmin Wang (Committee Member); Anthony Luscher (Committee Member) Subjects: Automotive Engineering; Mechanical Engineering

Master of Science in Engineering (MSEgr), Wright State University, 2011, Electrical Engineering

Automobiles depend more and more on electric power. Analysis of warranty data by automotive OEMs shows that faults in the automotive electrical power generation and storage (EPGS) system are often misdiagnosed. Therefore, monitoring of the state of health (SOH) of the automotive EPGS system is vital for early and correct diagnosis of faults in it, ensuring a reliable supply of electric power to the vehicle and reducing maintenance costs. In this research project, a model-based SOH monitoring method for the EPGS system is developed without the requirement of an alternator current sensor. A model representing the dynamic relationship between the battery current and the alternator filed duty voltage cycle is presented. An important model parameter that characterizes the current generation efficiency of the alternator system is adaptively estimated by using a recursive least square algorithm. Based on fault modes and effect analysis, a model-based fault detection and isolation decision scheme is developed for the EPGS system faults under consideration. The SOH monitoring method has been implemented using an EPGS system experimental test bench at GM R and D Center. Real-time evaluation results have shown its effectiveness and robustness.

Committee: Xiaodong Zhang PhD (Committee Chair); Kefu Xue PhD (Committee Co-Chair); Kuldip Rattan PhD (Committee Member); Marian Kazimierczuk PhD (Committee Member); Andrew Hsu PhD (Other) Subjects: Automotive Engineering; Electrical Engineering; Energy; Engineering; Technology

Master of Science in Engineering (MSEgr), Wright State University, 2009, Electrical Engineering

The development of an autopilot system for an RC helicopter presents interesting challengesfrom both a hardware and controls standpoint. The system detailed in this thesis utilizes a 13 state Extended Kalman Filter (EKF) to fuse sensor data and provide a position/velocity/attitude estimate. A novel, state of the art hybrid PID/LQR controller is developed and compared with a full state Linear Quadratic Regulator (LQR). The hybrid controller uses a proportional position, PID velocity outer loop coupled with an inner loop LQR for attitude control. The entire system is developed and implemented in hardware to produce a functional autopilot. The unit was installed on an Align Trex 600 RC helicopter and demonstrated its ability to hover the helicopter at a desired location. Preliminary investigations have also showed the controller is capable of flying waypoints.

Committee: Lang Hong PhD (Committee Chair); Kefu Xue PhD (Committee Member); Xiaodong Zhang PhD (Committee Member) Subjects: Electrical Engineering; Engineering