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

Supremacy in armed conflict comes not merely from superiority in capability or numbers but from how assets are used, down to the maneuvers of individual vehicles and munitions. This document outlines a research plan focused on skirmish-level tactics to militarily relevant scenarios. Skirmish-level refers to both the size of the adversarial engagement -- generally one vs. one, two vs. one, and/or one vs. two -- as well as the fact that the goal or objective of each team is well-established. The problem areas include pursuit-evasion and target guarding, either of which may be considered as sub-problems within military missions such as air-to-air combat, suppression/defense of ground-based assets, etc. In most cases, the tactics considered are comprised of the control policy of the agents (i.e., their spatial maneuvers), but may also include role assignment (e.g, whether to act as a decoy or striker) as well as discrete decisions (e.g., whether to engage or retreat). Skirmish-level tactics are important because they can provide insight into how to approach larger scale conflicts (many vs. many, many objectives, many decisions). Machine learning approaches such as reinforcement learning and neural networks have been demonstrated to be capable of developing controllers for large teams of agents. However, the performance of these controllers compared to the optimal (or equilibrium) policies is generally unknown. Differential Game Theory provides the means to obtain a rigorous solution to relevant scenarios in the form of saddle-point equilibrium control policies and the min/max (or max/min) cost / reward in the case of zero-sum games. When the equilibrium control policies can be obtained analytically, they are suitable for onboard / real-time implementation. Some challenges associated with the classical Differential Game Theory approach are explored herein. These challenges arise mainly due to the presence of singularities, which may appear in even the simplest differenti (open full item for complete abstract)

Committee: Zachariah Fuchs Ph.D. (Committee Member); David Casbeer Ph.D. (Committee Member); Dieter Vanderelst Ph.D. (Committee Member); Meir Pachter Ph.D. (Committee Member); John Gallagher Ph.D. (Committee Member) Subjects: Electrical Engineering

Doctor of Philosophy (PhD), Wright State University, 2017, Engineering PhD

The research on distributed multi-agent systems has received increasing attention due to its broad applications in numerous areas, such as unmanned ground and aerial vehicles, smart grid, sensor networks, etc. Since such distributed multi-agent systems need to operate reliably at all time, despite the possible occurrence of faulty behaviors in some agents, the development of fault-tolerant control schemes is a crucial step in achieving reliable and safe operations. The objective of this research is to develop a distributed adaptive fault-tolerant control (FTC) scheme for nonlinear uncertain multi-agent systems under intercommunication graphs with asymmetric weights. Under suitable assumptions, the closed-loop system's stability and leader-follower cooperative tracking properties are rigorously established. First, a distributed adaptive fault-tolerant control method for nonlinear uncertain first-order multi-agent systems is developed. Second, this distributed FTC method is extended to nonlinear uncertain second-order multi-agent systems. Next, adaptive-approximation-based FTC algorithms are developed for two cases of high-order multi-agent systems, i.e., with full-state measurement and with only limited output measurement, respectively. Finally, the distributed adaptive fault-tolerant formation tracking algorithms for first-order multi-agent systems are implemented and demonstrated using Wright State's real-time indoor autonomous robots test environment. The experimental formation tracking results illustrate the effectiveness of the proposed methods.

Committee: Xiaodong Zhang Ph.D. (Advisor); Kuldip Rattan Ph.D. (Committee Member); Pradeep Misra Ph.D. (Committee Member); Yongcan Cao Ph.D. (Committee Member); Raul Ordonez Ph.D. (Committee Member); Mark Mears Ph.D. (Committee Member) Subjects: Electrical Engineering; Engineering

PhD, University of Cincinnati, 2005, Engineering : Electrical Engineering

Within this dissertation, the problem of the control of the decentralized path planning decision processes of multiple cooperating autonomous aerial vehicles engaged in search of an uncertain environment is considered. The environment is modeled in a probabilistic fashion, such that both a priori and dynamic information about it can be incorporated. The components of the environment include both target information and threat information. Using the information about the environment, a computationally feasible decision process is formulated that can decide, in a near optimal fashion, which path a searching vehicle should take, using a dynamic programming algorithm with a limited look ahead horizon, with the possibility to extend the horizon using Approximate Dynamic Programming. A planning vehicle must take into account the effects of its (local) actions on meeting global goals. This is accomplished using a passive and predictive cooperation scheme among the vehicles. Lastly, a flexible simulator has been developed, using sound simulation analysis methods, to simulate a UAV search team, which can be used to create statistically valid results demonstrating the effectiveness of the model and solution methods.

Committee: Dr. Emmanuel Fernandez (Advisor) Subjects:

MS, University of Cincinnati, 2002, Engineering : Electrical Engineering

The work presented here is part of a research program being conducted in the area of decision and control for autonomous unmanned aerial vehicles (UAV's). Speci.cally, the formulation is presented for the problem of generating near-optimal trajectories to follow in order for several UAV's to cooperatively search for targets in a given area for which some a priori data about target distribution is available. In order to solve this problem, a discrete time decision model is created. The solution based on this model is presented, which utilizes a dynamic programming approach, implemented with a best-.rst search algorithm. This solution predicts the best path for individual vehicles to take under constraints on movement and computational power. A key reduction in computational complexity as compared to the ideal case is made by utilizing a limited look-ahead policy and by modeling other vehicles as stochastic elements. The formulation is .exible enough to respond to additional goals and restrictions, also. A set of simulation studies is provided that shows the utility of this approach. The proposed method is demonstrated against a standard search, and another method that currently exists in the literature.

Committee: Dr. Marios Polycarpou (Advisor) Subjects:

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

The ability to inspect and repair spacecraft or other assets is a critical need for all entities public or private that wish to operate in the space domain. In-space Servicing, Assembly, and Maintenance, or ISAM, encompasses a wide variety of tasks necessary to establish and maintain a permanent human presence in space. Past ISAM tasks such as servicing the Hubble Space Telescope have relied on high-cost and high-risk missions to send astronauts to the spacecraft to perform the task, making ISAM a field that will benefit greatly from increased autonomy both in planning and execution. In this dissertation, the ISAM mission is studied in four parts: (1) Multi-Agent Control uses Reinforcement Learning (RL) to train a multi-agent system to cooperate in order to complete a task, (2) Cubesat Inspection consists of deployment of a team of cubesats about a target object with the goal to provide efficient, low-risk inspection coverage of the object prior to approach by the servicing spacecraft, (3) Artificial Intelligence Manipulator Control uses RL to control a space manipulator system to precisely position the end effector of the manipulator relative to a target object, and (4) Dual Manipulator Stabilization develops techniques to use a second manipulator arm to provide active stabilization of the base spacecraft while the primary manipulator is performing a mission task. The multi-agent control task employs a heterogeneous robotic team to perform an exploration task in a grid-based environment. A method is established for determining the reward criteria (figures of merit) that can be used for training the robot team through reinforcement learning techniques. A hierarchical framework of rewards is used which, at the lowest level, measures the success of an individual robot in performing its task. The success of all robots performing different subtasks is then measured using a cooperation metric and finally, the (open full item for complete abstract)

Committee: Ou Ma Ph.D. (Committee Chair); Donghoon Kim Ph.D. (Committee Member); Rajnikant Sharma Ph.D. (Committee Member); Kelly Cohen Ph.D. (Committee Member) Subjects: Engineering

MS, University of Cincinnati, 2024, Engineering and Applied Science: Aerospace Engineering

With the advancement of communication technology, automobiles are gaining more functionalities at an increasing rate. Connected vehicle (CV) technology enables wireless communication between autos, also known as Vehicle to Vehicle (V2V) technology. Creating interpretable vehicle platooning controllers is crucial for improving both human-vehicle communication and ensuring compliance with responsible artificial intelligence (AI) principles. Unforeseen malfunctions of these devices might result in annoyance and a reduction in driver confidence, underscoring the importance of responsible AI. By ensuring meticulous and accountable development, implementation, and use of artificial intelligence systems, we not only reduce the potential disadvantages in platooning controllers but also guarantee transparency, equity, and security in the broader application of AI. The objective of this thesis is to develop fuzzy logic based controllers for connected vehicle platooning, specifically focusing on longitudinal and latitudinal car-following control. The objective of this study is to enhance a car's ability to sustain a consistent time interval between vehicles and optimize the comfort of highway travel. This research examines various scenarios that closely resemble highway circumstances, all of which are conducted at high speeds on the highway. The driving models that have been created aim to reduce the distance or time gap between the preceding vehicles in each scenario to a following distance or time gap of 1 second. The driving models are evaluated against the Krauss driving model, which emulates a human driver, and the Cooperative Adaptive Cruise Control (CACC) driving model in identical settings. This study will specifically examine the traffic flow and safety precautions, including the distance between the car being studied and the vehicle in front of it, the abrupt changes in acceleration of the vehicle being studied, and the time it takes for a collision to occur (TTC).

Committee: Kelly Cohen Ph.D. (Committee Chair); Anoop Sathyan Ph.D. (Committee Member); Donghoon Kim Ph.D. (Committee Member); Manish Kumar Ph.D. (Committee Member) Subjects: Aerospace Materials

Master of Science in Civil Engineering, Cleveland State University, 2021, Washkewicz College of Engineering

The transportation industry is going through many changes as populations grow and new technologies pave the way to the future of transportation. One of the emerging technologies in transportation is Connected and Autonomous Vehicles (CAVs), promising many improvements by using advanced technologies. There is, however, so much research and development required to reach full deployment of CAVs and fill the transition from human driven vehicles to fully automated vehicles. As a result of positive environmental and economic impacts of automated truck platoons on the transportation industry, it is expected that they will be among the first applications of CAVs deployment. However, it is also important to evaluate their impacts on traffic flow in order to have a comprehensive deployment plan. Previous studies have shown mixed effects of truck platooning on traffic flow; however, many researchers did not consider the impacts of on- and off-ramps as it is the case in many urban freeways. The main objective of this thesis is to examine whether automated truck platoons would have a positive impact on traffic flow in terms of the appropriate performance measures based on different characteristics of platooning. To do so, a comprehensive literature review was conducted. Additionally, using VISSIM microsimulation software, a case study of a 5.5 mile corridor of the I-880 freeway having 8 on-ramps and 5 off-ramps during AM peak hour was conducted to evaluate the impacts of automated truck platoons on traffic flow. According to the results of the case study, automated truck platoons negatively impact traffic flow in terms of average speed, total network delay, and total time spent in the network based on different factors of platooning, including gap, market penetration rate, and platoon size. Statistical tests indicated that only market penetration rates of more than 30% have significant impacts on the performance measures compared to the base scenario. This negative impact (open full item for complete abstract)

Committee: Jacqueline M. Jenkins (Advisor); Hongkai Yu (Committee Member); Mehdi Rahmati (Committee Member) Subjects: Civil Engineering; Transportation; Transportation Planning

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

In Cooperative Localization, agents share their sensor information with each other for collective state estimation. Observability of the system plays a key role in maintaining localization accuracy, irrespective of the estimation technique used. Since the observability of nonlinear systems is state-dependent, controllers can be developed to solve for states that improve observability of the system. In this thesis, we solve such a problem in relative framework. For missions like aerial refueling, landing, formation flying etc., knowledge of relative pose is more important for successful execution of the mission. When flying through GPS-denied hostile environments, the agents must estimate their relative pose with respect to each other using on-board sensors. Depending on the nature of the mission, agents can be constrained to move in certain trajectories that can make the system unobservable, because of which estimation errors accumulate over time. In such cases, additional vehicles can be introduced to provide more measurements, which improves localization accuracy. Furthermore, the path followed by these vehicles can improve the observability of the system. Therefore, we perform a detailed nonlinear observability analysis of the N-vehicle system and derive a cost function from the observability gramian (O'O). We solve for trajectories of these additional supporting vehicles that maximize this cost using Trajectory Optimization coupled with Model Predictive Control (MPC) approach. When multiple supporting vehicles are involved, we can solve for the trajectories of all the vehicles collectively or solve for each vehicle individually. We present simulation results using MATLAB/Simulink that demonstrates the effectiveness and consistency of the controller developed. We show that the distributed approach offers the same accuracy with low computation time compared to the centralized approach. We also study the effect of the number of additional supporting vehicles (open full item for complete abstract)

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

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

Differential games provide a framework for solving dynamic systems of two competing interests. One family of differential games focuses on two competing agents. Another family of differential games has competing teams of agents known as teaming games. Solutions to teaming games have inherit challenges due to the number of agents. By increasing the number of agents on each team, the dimension of the state space increases and additional termination cases are created. Singularities are a challenge in the majority of differential games and increasing the number of agents compounds that difficulty. This study aimed to overcome the challenges presented in solving teaming differential games by solving corresponding optimal control problems. The teaming game in question is a single pursuer, dual evader pursuit-evasion differential game. By fixing the control strategy of the pursuer, the teaming differential game is transformed into an optimal control problem for the team of evaders. Conversely, fixing the control strategies of the evaders results in an optimal control problem for the pursuer. For both problems, the optimal control strategy for the team in question is determined along with any singularities present within the control strategies. The teaming game is then reconsidered. Similarities between the optimal control problems and differential game allow for a simplified development of the solution to teaming game. The work concludes by demonstrating how solving the corresponding optimal control problems helps to overcome the inherit challenges of solving a teaming differential game.

Committee: Zachariah Fuchs Ph.D. (Committee Chair); David Casbeer Ph.D. (Committee Member); Ali Minai Ph.D. (Committee Member); Dieter Vanderelst Ph.D. (Committee Member) Subjects: Electrical Engineering

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

With the increasing demand for real-time control logic, infrastructure-enabled vehicle detector data is being considered for state of art traffic signal control strategies. The conventional detection methods are usually point detection that cannot directly measure vehicle speed and location. This has been the biggest challenge to design a robust traffic control system. Connected Vehicles (CVs) due to the advancements in wireless communication technology are a potential solution to overcome this challenge. The emerging CV technology provides an opportunity to formulate an ambulant data platform that allows the actual data transfer among multiple vehicles as well as with the infrastructure. More significantly, the CV's capability of serving as the mobile trajectory sensors could help us to reduce the dependencies on conventional infrastructure-based vehicle detectors. The connected vehicles can provide increased opportunities and enforce more challenges for the signal control of urban traffic. These include vehicle to infrastructure (V2I), vehicle to vehicle (V2V), and vehicle to something else's(V2X). The core objective of this study is to create a framework in which algorithms, modeling methods, and testing schemes for the optimization of urban traffic signal under mixed traffic conditions are included (coexistence of conventional vehicles and CVs). For isolated intersections or multiple intersections along a corridor, this framework can improve traffic signal timing. Precisely, the major assignments of this research include: 1.Thorough testing in traffic simulation to reinforce the proposed methods. This research evaluated the CCACSTO algorithm at four different penetration rates of CAVs for three different traffic conditions (light traffic, mild traffic, and heavy traffic). The simulation test results show that average vehicle delay and queue length with CCACSTO algorithm reduced by 46.04% and 56.15% respectively under 50% penetration rate of CAVs.

Committee: Heng Wei Ph.D. (Committee Chair); Jiaqi Ma Ph.D. (Committee Member); Nick Yeretzian MS Civil Engineering (Committee Member) Subjects: Transportation

PhD, University of Cincinnati, 2019, Engineering and Applied Science: Mechanical Engineering

Multi-UAV systems are inherently safety-critical systems, which means that safety guarantees must be made to ensure no undesirable configurations, such as collisions, occur. This dissertation focuses on developing optimization algorithms for trajectory planning of single as well as multiple cooperating Unmanned Air Vehicles (UAVs) operating in a cluttered environment that comprise of stationary obstacles and other cooperating as well as non-cooperating moving vehicles. This dissertation presents a Partial Differential Equation (PDE) based generalized method for UAV trajectory planning in a three-dimensional world using a number benchmark multi-UAV cooperative control problems. The PDEs proposed in this dissertation are based on the dynamics governing the multi-phase fluid motion in a porous medium. The method introduces a risk value representing the risk of collision or other hazard associated with every point in the domain. That risk value represents the notion of porosity (permeability) used in fluid flowing through a porous medium. This value is used in the PDE whose solution is obtained via novel numerical methods to calculate the streamlines that constitute the potential paths from a starting point to a target location. In particular, this research proposes a machine learning technique to decrease the computational time for calculations of flow movements in porous medium to 0.7 seconds which leads this technique to be implemented on-board and online. Subsequently, based on the criteria of the optimization problem, we propose post-processing of the streamlines to yield all the flyable paths. The proposed controller, based on multi-phase flows, is executed with a new decentralized manner using a concept of Prediction Sets (PSs). This method has been applied to three different cooperative control problems. IN first problem, large-scale path planning problem of UAVs is considered in shared airspace. The method is qualitatively compared via a simulation study (open full item for complete abstract)

Committee: Manish Kumar Ph.D. (Committee Chair); David Casbeer PhD (Committee Member); Kelly Cohen Ph.D. (Committee Member); Donald French Ph.D. (Committee Member); Tamara Lorenz Ph.D. (Committee Member) Subjects: Mechanical Engineering

Doctor of Philosophy, The Ohio State University, 2017, Electrical and Computer Engineering

Cooperative highway driving systems (CHDSs) consist of collaborating vehicles with automated control units and vehicle-to-vehicle communication capabilities. Such systems are proposed as an important component of intelligent transportation systems (ITS) aiming at improving energy efficiency and driving safety. CHDSs have a broad spectrum of applications, ranging from automated freight systems to highway automation to smart city transit. Modeling and control of cooperative vehicles on highways contributes importantly to CHDS development. This problem is of critical importance in developing safe and reliable controllers and establishing frameworks and criteria verifying CHDS performance. This work focuses on the cooperative control problems in developing CHDSs by investigating distributed model predictive control (DMPC) techniques. In particular, collaboration of connected and automated vehicles is first formulated into a constrained optimization problem. Then, different DMPC strategies are investigated considering features of the cooperative control problem in a CHDS. We focus on non-iterative DMPC schemes with partially parallel information exchange between subsystems. Feasibility and stability properties of the closed-loop system applying non-iterative DMPC are established taking into account the coupling of the control input with state predictions calculated at previous step. Furthermore, a non-iterative DMPC scheme implementing a partitioning procedure is proposed to reduce the conservatism of compatibility constraints while guaranteeing safe inter-vehicle distances. With the DMPC scheme controlling the connected and automated vehicles, we further investigate interactions of cooperative driving groups with surrounding human-operated vehicles in mixed traffic environments. A behavior classification framework is developed to detect driver behaviors of surrounding human-operated vehicles. With the behavior classification framework, a behavior-guided MPC controller (open full item for complete abstract)

Committee: Umit Ozguner (Advisor) Subjects: Electrical Engineering; Robotics; Transportation

MS, University of Cincinnati, 2015, Engineering and Applied Science: Computer Science

Continued evolution of Wireless Body Area Networks (WBANs) has made effective monitoring of vital parameters of a person much faster and efficient, thereby providing better personal healthcare. Sensor nodes of a WBAN acquire critical physiological parameters like heartbeat, neural activity, limb motion, muscle movement and fatigue, temperature, etc. that are monitored by a physician. Important factors in the acceptance of WBAN performance are energy efficiency and the Quality of Service (QoS) supported for such critical data that impact human lives. The sensor nodes of a WBAN are highly constrained in terms of their battery life. Most of the work till date on WBANs uses a star topology which employs single hop communication. This work discusses various factors that affect energy efficiency in a WBAN and establishes the need for a multi-hop tree based topology. It also studies the need for QoS in WBANs and existing support provided by the current Body Area Sensor Network (BASN) Standard. This thesis tackles the all important challenge of providing QoS in autonomous and neighbor-aware multi-hop WBANs in significant detail spread across multiple chapters. In case of independent, autonomous multi-hop WBANs, the aforementioned issue is resolved by implementing a two layer priority-mapping scheme over a reactive Media Access Control (MAC) layer designed to alter durations of the access phases involved as per QoS requirement. In the case of neighbor-aware WBANs, a framework is defined under which a cooperative inter-WBAN routing scheme is implemented through power based weight assignment and fault detection is carried out by employing Kosaraju's two-pass algorithm that discovers the strongly connected components in the network deployment graph.

Committee: Dharma Agrawal D.Sc. (Committee Chair); Raj Bhatnagar Ph.D. (Committee Member); Prabir Bhattacharya Ph.D. (Committee Member) Subjects: Computer Science

PhD, University of Cincinnati, 2013, Engineering and Applied Science: Mechanical Engineering

This work is intended towards the development of a framework for the deployment of a distributed multi-agent system for co-operative monitoring of spatio-temporal processes in applications such as wildland fires and utilizing the information thus obtained for improved situational awareness. The use of such co-operative systems has strong advantages over conventional methods and presents fewer risks than manned aerial missions. Development of such a framework requires addressing several challenges in sensing, control, optimization, estimation and related technologies. Towards such a framework, this dissertation work focuses on two significant aspects of its development: (i) cooperative control in a multi-agent system for distributed data gathering, and (ii) development of a data processing and filtering algorithm for spatio-temporal estimation. To achieve the first objective, this work develops a co-operative control strategy that optimizes the spatial distribution of agents around closed curves that typically represent most dynamic perimeters. A linear cyclic pursuit control model based on double integrator dynamics has been developed, and the convergence of a system of agents governed by this control model to a stable well-distributed pursuit configuration is demonstrated. The theories developed around the co-operative control model and pursuit dynamics are validated over real-time experiments involving a group of ground robots under the influence of the controller. Further, the influence of addition of a non-linear repulsive term to this control model and its influence on the stability of the control model is evaluated numerically. The control model, further, is extended towards tracking general forms of closed towards practical implementation in real-world applications. A two-dimensional controller is developed in this regard, where the aforementioned cyclic pursuit control action acting along a closed curve is augmented with an orthogonal radial perimeter (open full item for complete abstract)

Committee: Manish Kumar Ph.D. (Committee Chair); Randall Allemang Ph.D. (Committee Member); Kelly Cohen Ph.D. (Committee Member); David Thompson Ph.D. (Committee Member) Subjects: Mechanics

MS, University of Cincinnati, 2012, Engineering and Applied Science: Aerospace Engineering

The ever-increasing applications of UAV's have shown the great capabilities of these technologies. However, for many cases where one UAV is a powerful tool, an autonomous swarm all working cooperatively to the same goal presents amazing potential. Environment that are dangerous for humans, are either too small or too large for safe or reasonable exploration, and even those tasks that are simply boring or unpleasant are excellent areas for UAV swarm applications. In order to work cooperatively, the swarm must allocate tasks and have adequate path planning capability. This paper presents a methodology for two-dimensional target allocation and path planning of a UAV swarm using a hybridization of control techniques. Genetic algorithms, fuzzy logic, and to an extent, dynamic programming are utilized in this research to develop a code known as “UNCLE SCROOGE” (UNburdening through CLustering Effectively and Self-CROssover GEnetic algorithm). While initially examining the Traveling Salesman Problem, where an agent must visit each waypoint in a set once and then return home in the most efficient path, the work's end goal was a variant on this problem that more closely resembled the issues a UAV swarm would encounter. As an extension to Dr. Obenmeyer's “Polygon-Visiting Dubins Traveling Salesman Problem”, the Multi-Depot Polygon-Visiting Dubins Multiple Traveling Salesman Problem consists of a set number of visibility areas, or polygons that a number of UAV's, based in different or similar depot must visit. While this case is constant altitude and constant velocity, minimum turning radii are considered through the use of Dubins curves. UNCLE SCROOGE was found to be adaptable to the PVDTSP, where it competed well against the methods proposed by Obenmeyer. Due to limited benchmarking ability, as these are newly formed problems, Obenmeyer's work served as the only basis for comparison for the PVDTSP. UNCLE SCROOGE brought a 9.8% increase in accuracy, and a run-time reduction (open full item for complete abstract)

Committee: Kelly Cohen PhD (Committee Chair); Manish Kumar PhD (Committee Member); Bruce Walker ScD (Committee Member) Subjects: Aerospace Materials

MS, University of Cincinnati, 2010, Engineering : Mechanical Engineering

In recent years, the research activities in the area of multi-robot systems have grown dramatically. Compared to single robot systems, the multi-robot systems are more robust, flexible, and efficient. However, controlling a multi-robot system is not trivial and involves several challenges including uncertainty, stability and scalability problems. This thesis makes two significant contributions to the multi-robot research: ⅰ) development of cooperative control algorithm based on a hydrodynamic model; ⅱ) development of self-localization and tracking scheme in multi-robot system. In the first part, this thesis proposes a fluid model inspired cooperative control algorithm to control multiple robots. Since this approach takes the density of robots as an important control variable, it is given a name called density-based control. Two tasks have been implemented to demonstrate the efficiency of the density-based controller: group motion and shape control, and group segregation. In the second part, this thesis proposes a novel algorithm for self-localization and tracking in a multi-robot system that implements a non-pattern based method. In contrast to the traditional pattern-based localization systems, a localization method without using patterns can be more scalable to the number of robots and more efficient in computation. The efficiency of the non-pattern based self-localization scheme has been demonstrated with the help of extensive simulations and experiments.

Committee: Manish Kumar PhD (Committee Chair); Ali Minai PhD (Committee Member); David Thompson PhD (Committee Member) Subjects: Mechanical Engineering

Doctor of Philosophy, The Ohio State University, 2007, Electrical Engineering

The field of cooperative control involves the study of situations in which groups of mobile agents (e.g., autonomous aircraft, etc.) must collaborate with each other in order to accomplish a collective goal. A large number of these situations may be classified as resource allocation problems, which is to say that the agents must determine how to use their available resources in the most effective manner. Because the agents are mobile, these problems often center around trying to achieve an optimal spatial distribution of the group. This distribution may be defined as either a specific set of coordinates or a specific trajectory for each agent, but more often there are a finite number of discrete sites at which the agents may be located over time (and potentially different tasks within each site to which they may be assigned). This dissertation addresses three different problems of the latter type: maintaining uniform surveillance of an environment, team formation within groups of heterogenous agents, and a cooperative selection process. For each problem we develop algorithms that can be used by a group to either achieve the appropriate distribution of agents or at least make a good effort towards that goal despite very restrictive conditions. This research both extends theoretical results from similar problems and introduces a number of novel problem formulations. Heuristic algorithms with good performance have also been developed for problems that resist analytical treatment.

Committee: Kevin Passino (Advisor) Subjects:

Doctor of Philosophy, The Ohio State University, 2003, Electrical Engineering

Resource allocation involves deciding how to divide a resource of limited availability across multiple demands in a way that optimizes current objectives (e.g., allocating a processor's computing resources to the demand presented by tasks in order to maximize task completion throughput). In “distributed” resource allocation there are multiple resource types each of which can be subdivided, but then each can only be allocated to a subset of the demands (e.g., in a multiprocessor system where each processor can only process certain task types). In this dissertation we focus on three types of resource allocation problems where via an imperfect communication network multiple agents can share the workload presented by multiple task types. First, we define a model for a network of processors processing task types from buffers and show that they lead to the cumulative demand being bounded by a constant. We demonstrate via simulations when they can be superior to one noncooperative strategy. Second, we model a cooperative control problem for a network of uninhabited autonomous vehicles (UAVs) where it is assumed that before the mission starts a set of tasks is given to a set of UAVs, but then after deployment the UAVs must cooperate to decide which UAV should process each task. We introduce cooperative scheduling of tasks for a set of UAVs where the cooperation must occur over a network that has random but bounded delays. We show how to view this as a cooperative scheduling (resource allocation) problem, and how to derive bounds on mission-level performance metrics for cooperative scheduling methods. Simulations will be used to compare the approach to a noncooperative strategy and to provide design guidelines for the cooperative scheduler. Finally, we introduce an inexpensive laboratory testbed for networked cooperative scheduling strategies. We describe the apparatus, highlight the challenges it presents, and we compare the performance of two scheduling strategies that see (open full item for complete abstract)

Committee: Kevin Passino (Advisor) Subjects: