Matching Items (1,166)
Filtering by
- Creators: School of Mathematical and Statistical Sciences
- Creators: Department of Information Systems
- Member of: Theses and Dissertations
Description
Carbon nanomaterials have caught tremendous attention in the last few decades due to their unique physical and chemical properties. Tremendous effort has been made to develop new synthesis techniques for carbon nanomaterials and investigate their properties for different applications. In this work, carbon nanospheres (CNSs), carbon foams (CF), and single-walled carbon nanotubes (SWNTs) were studied for various applications, including water treatment, energy storage, actuators, and sensors.
A facile spray pyrolysis synthesis technique was developed to synthesize individual CNSs with specific surface area (SSA) up to 1106 m2/g. The hollow CNSs showed adsorption of up to 300 mg rhodamine B dye per gram carbon, which is more than 15 times higher than that observed for conventional carbon black. They were also evaluated as adsorbents for removal of arsenate and selenate from water and displayed good binding to both species, outperforming commercial activated carbons for arsenate removal in pH > 8. When evaluated as supercapacitor electrode materials, specific capacitances of up to 112 F/g at a current density of 0.1 A/g were observed. When used as Li-ion battery anode materials, the CNSs achieved a discharge capacity of 270 mAh/g at a current density of 372 mA/g (1C), which is 4-fold higher than that of commercial graphite anode.
Carbon foams were synthesized using direct pyrolysis and had SSA up to 2340 m2/g. When used as supercapacitor electrode materials, a specific capacitance up to 280 F/g was achieved at current density of 0.1 A/g and remained as high as 207 F/g, even at a high current density of 10 A/g.
A printed walking robot was made from common plastic films and coatings of SWNTs. The solid-state thermal bimorph actuators were multifunctional energy transducers powered by heat, light, or electricity. The actuators were also investigated for photo/thermal detection. Electrochemical actuators based on MnO2 were also studied for potential underwater applications.
SWNTs were also used to fabricate printable electrodes for trace Cr(VI) detection, which displayed sensitivity up to 500 nA/ppb for Cr(VI). The limit of detection was shown to be as low as 5 ppb. A flow detection system based on CNT/printed electrodes was also demonstrated.
A facile spray pyrolysis synthesis technique was developed to synthesize individual CNSs with specific surface area (SSA) up to 1106 m2/g. The hollow CNSs showed adsorption of up to 300 mg rhodamine B dye per gram carbon, which is more than 15 times higher than that observed for conventional carbon black. They were also evaluated as adsorbents for removal of arsenate and selenate from water and displayed good binding to both species, outperforming commercial activated carbons for arsenate removal in pH > 8. When evaluated as supercapacitor electrode materials, specific capacitances of up to 112 F/g at a current density of 0.1 A/g were observed. When used as Li-ion battery anode materials, the CNSs achieved a discharge capacity of 270 mAh/g at a current density of 372 mA/g (1C), which is 4-fold higher than that of commercial graphite anode.
Carbon foams were synthesized using direct pyrolysis and had SSA up to 2340 m2/g. When used as supercapacitor electrode materials, a specific capacitance up to 280 F/g was achieved at current density of 0.1 A/g and remained as high as 207 F/g, even at a high current density of 10 A/g.
A printed walking robot was made from common plastic films and coatings of SWNTs. The solid-state thermal bimorph actuators were multifunctional energy transducers powered by heat, light, or electricity. The actuators were also investigated for photo/thermal detection. Electrochemical actuators based on MnO2 were also studied for potential underwater applications.
SWNTs were also used to fabricate printable electrodes for trace Cr(VI) detection, which displayed sensitivity up to 500 nA/ppb for Cr(VI). The limit of detection was shown to be as low as 5 ppb. A flow detection system based on CNT/printed electrodes was also demonstrated.
ContributorsWang, Chengwei, Ph.D (Author) / Chan, Candace K. (Thesis advisor) / Tongay, Sefaattin (Committee member) / Wang, Qing Hua (Committee member) / Seo, Dong (Committee member) / Arizona State University (Publisher)
Created2015
Description
Design problem formulation is believed to influence creativity, yet it has received only modest attention in the research community. Past studies of problem formulation are scarce and often have small sample sizes. The main objective of this research is to understand how problem formulation affects creative outcome. Three research areas are investigated: development of a model which facilitates capturing the differences among designers' problem formulation; representation and implication of those differences; the relation between problem formulation and creativity.
This dissertation proposes the Problem Map (P-maps) ontological framework. P-maps represent designers' problem formulation in terms of six groups of entities (requirement, use scenario, function, artifact, behavior, and issue). Entities have hierarchies within each group and links among groups. Variables extracted from P-maps characterize problem formulation.
Three experiments were conducted. The first experiment was to study the similarities and differences between novice and expert designers. Results show that experts use more abstraction than novices do and novices are more likely to add entities in a specific order. Experts also discover more issues.
The second experiment was to see how problem formulation relates to creativity. Ideation metrics were used to characterize creative outcome. Results include but are not limited to a positive correlation between adding more issues in an unorganized way with quantity and variety, more use scenarios and functions with novelty, more behaviors and conflicts identified with quality, and depth-first exploration with all ideation metrics. Fewer hierarchies in use scenarios lower novelty and fewer links to requirements and issues lower quality of ideas.
The third experiment was to see if problem formulation can predict creative outcome. Models based on one problem were used to predict the creativity of another. Predicted scores were compared to assessments of independent judges. Quality and novelty are predicted more accurately than variety, and quantity. Backward elimination improves model fit, though reduces prediction accuracy.
P-maps provide a theoretical framework for formalizing, tracing, and quantifying conceptual design strategies. Other potential applications are developing a test of problem formulation skill, tracking students' learning of formulation skills in a course, and reproducing other researchers’ observations about designer thinking.
This dissertation proposes the Problem Map (P-maps) ontological framework. P-maps represent designers' problem formulation in terms of six groups of entities (requirement, use scenario, function, artifact, behavior, and issue). Entities have hierarchies within each group and links among groups. Variables extracted from P-maps characterize problem formulation.
Three experiments were conducted. The first experiment was to study the similarities and differences between novice and expert designers. Results show that experts use more abstraction than novices do and novices are more likely to add entities in a specific order. Experts also discover more issues.
The second experiment was to see how problem formulation relates to creativity. Ideation metrics were used to characterize creative outcome. Results include but are not limited to a positive correlation between adding more issues in an unorganized way with quantity and variety, more use scenarios and functions with novelty, more behaviors and conflicts identified with quality, and depth-first exploration with all ideation metrics. Fewer hierarchies in use scenarios lower novelty and fewer links to requirements and issues lower quality of ideas.
The third experiment was to see if problem formulation can predict creative outcome. Models based on one problem were used to predict the creativity of another. Predicted scores were compared to assessments of independent judges. Quality and novelty are predicted more accurately than variety, and quantity. Backward elimination improves model fit, though reduces prediction accuracy.
P-maps provide a theoretical framework for formalizing, tracing, and quantifying conceptual design strategies. Other potential applications are developing a test of problem formulation skill, tracking students' learning of formulation skills in a course, and reproducing other researchers’ observations about designer thinking.
ContributorsDinar, Mahmoud (Author) / Shah, Jami J. (Thesis advisor) / Langley, Pat (Committee member) / Davidson, Joseph K. (Committee member) / Lande, Micah (Committee member) / Ren, Yi (Committee member) / Arizona State University (Publisher)
Created2015
Description
In this dissertation, three complex material systems including a novel class of hyperuniform composite materials, cellularized collagen gel and low melting point alloy (LMPA) composite are investigated, using statistical pattern characterization, stochastic microstructure reconstruction and micromechanical analysis. In Chapter 1, an introduction of this report is provided, in which a brief review is made about these three material systems. In Chapter 2, detailed discussion of the statistical morphological descriptors and a stochastic optimization approach for microstructure reconstruction is presented. In Chapter 3, the lattice particle method for micromechanical analysis of complex heterogeneous materials is introduced. In Chapter 4, a new class of hyperuniform heterogeneous material with superior mechanical properties is investigated. In Chapter 5, a bio-material system, i.e., cellularized collagen gel is modeled using correlation functions and stochastic reconstruction to study the collective dynamic behavior of the embed tumor cells. In chapter 6, LMPA soft robotic system is generated by generalizing the correlation functions and the rigidity tunability of this smart composite is discussed. In Chapter 7, a future work plan is presented.
ContributorsXu, Yaopengxiao (Author) / Jiao, Yang (Thesis advisor) / Liu, Yongming (Committee member) / Wang, Qing Hua (Committee member) / Ren, Yi (Committee member) / Arizona State University (Publisher)
Created2018
Description
Ultrasonication-mediated liquid-phase exfoliation has emerged as an efficient method for producing large quantities of two-dimensional materials such as graphene, boron nitride, and transition metal dichalcogenides. This thesis explores the use of this process to produce a new class of boron-rich, two-dimensional materials, namely metal diborides, and investigate their properties using bulk and nanoscale characterization methods. Metal diborides are a class of structurally related materials that contain hexagonal sheets of boron separated by metal atoms with applications in superconductivity, composites, ultra-high temperature ceramics and catalysis. To demonstrate the utility of these materials, chromium diboride was incorporated in polyvinyl alcohol as a structural reinforcing agent. These composites not only showed mechanical strength greater than the polymer itself, but also demonstrated superior reinforcing capability to previously well-known two-dimensional materials. Understanding their dispersion behavior and identifying a range of efficient dispersing solvents is an important step in identifying the most effective processing methods for the metal diborides. This was accomplished by subjecting metal diborides to ultrasonication in more than thirty different organic solvents and calculating their surface energy and Hansen solubility parameters. This thesis also explores the production and covalent modification of pristine, unlithiated molybdenum disulfide using ultrasonication-mediated exfoliation and subsequent diazonium functionalization. This approach allows a variety of functional groups to be tethered on the surface of molybdenum disulfide while preserving its semiconducting properties. The diazonium chemistry is further exploited to attach fluorescent proteins on its surface making it amenable to future biological applications. Furthermore, a general approach for delivery of anticancer drugs using pristine two-dimensional materials is also detailed here. This can be achieved by using two-dimensional materials dispersed in a non-ionic and biocompatible polymer, as nanocarriers for delivering the anticancer drug doxorubicin. The potency of this supramolecular assembly for certain types of cancer cell lines can be improved by using folic-acid-conjugated polymer as a dispersing agent due to strong binding between folic acid present on the nanocarriers and folate receptors expressed on the cells. These results show that ultrasonication-mediated liquid-phase exfoliation is an effective method for facilitating the production and diverse application of pristine two-dimensional metal diborides and transition metal dichalcogenides.
ContributorsYousaf, Ahmed (Author) / Green, Alexander A (Thesis advisor) / Wang, Qing Hua (Committee member) / Liu, Yan (Committee member) / Arizona State University (Publisher)
Created2018
Description
Two-dimensional transition metal dichalcogenides (TMDCs) such as
molybdenum disulfide (MoS2), tungsten disulfide (WS2), molybdenum diselenide (MoSe2) and tungsten diselenide (WSe2) are attractive for use in biotechnology, optical and electronics devices due to their promising and tunable electrical, optical and chemical properties. To fulfill the variety of requirements for different applications, chemical treatment methods are developed to tune their properties. In this dissertation, plasma treatment, chemical doping and functionalization methods have been applied to tune the properties of TMDCs. First, plasma treatment of TMDCs results in doping and generation of defects, as well as the synthesis of transition metal oxides (TMOs) with rolled layers that have increased surface-to-volume ratio and are promising for electrochemical applications. Second, chemical functionalization is another powerful approach for tuning the properties of TMDCs for use in many applications. To covalently functionalize the basal planes of TMDCs, previous reports begin with harsh treatments like lithium intercalation that disrupt the structure and lead to a phase transformation from semiconducting to metallic. Instead, this work demonstrates the direct covalent functionalization of semiconducting MoS2 using aryl diazonium salts without lithium treatments. It preserves the structure and semiconducting nature of MoS2, results in covalent C-S bonds on basal planes and enables different functional groups to be tethered to the MoS2 surface via the diazonium salts. The attachment of fluorescent proteins has been used as a demonstration and it suggests future applications in biology and biosensing. The effects of the covalent functionalization on the electronic transport properties of MoS2 were then studied using field effect transistor (FET) devices.
molybdenum disulfide (MoS2), tungsten disulfide (WS2), molybdenum diselenide (MoSe2) and tungsten diselenide (WSe2) are attractive for use in biotechnology, optical and electronics devices due to their promising and tunable electrical, optical and chemical properties. To fulfill the variety of requirements for different applications, chemical treatment methods are developed to tune their properties. In this dissertation, plasma treatment, chemical doping and functionalization methods have been applied to tune the properties of TMDCs. First, plasma treatment of TMDCs results in doping and generation of defects, as well as the synthesis of transition metal oxides (TMOs) with rolled layers that have increased surface-to-volume ratio and are promising for electrochemical applications. Second, chemical functionalization is another powerful approach for tuning the properties of TMDCs for use in many applications. To covalently functionalize the basal planes of TMDCs, previous reports begin with harsh treatments like lithium intercalation that disrupt the structure and lead to a phase transformation from semiconducting to metallic. Instead, this work demonstrates the direct covalent functionalization of semiconducting MoS2 using aryl diazonium salts without lithium treatments. It preserves the structure and semiconducting nature of MoS2, results in covalent C-S bonds on basal planes and enables different functional groups to be tethered to the MoS2 surface via the diazonium salts. The attachment of fluorescent proteins has been used as a demonstration and it suggests future applications in biology and biosensing. The effects of the covalent functionalization on the electronic transport properties of MoS2 were then studied using field effect transistor (FET) devices.
ContributorsChu, Ximo (Author) / Wang, Qing Hua (Thesis advisor) / Sieradzki, Karl (Committee member) / Green, Alexander (Committee member) / Chan, Candace (Committee member) / Arizona State University (Publisher)
Created2018
Description
Flame retardants (FRs) are applied to variety of consumer products such as textiles and polymers for fire prevention and fire safety. Substantial research is ongoing to replace traditional FRs with alternative materials that are less toxic, present higher flame retardancy and result in lower overall exposure as there are potential health concerns in case of exposure to popular FRs. Carbonaceous nanomaterials (CNMs) such as carbon nanotubes (CNTs) and graphene oxide (GO) have been studied and applied to polymer composites and electronics extensively due to their remarkable properties. Hence CNMs are considered as potential alternative materials that present high flame retardancy. In this research, different kinds of CNMs coatings on polyester fabric are produced and evaluated for their use as flame retardants. To monitor the mass loading of CNMs coated on the fabric, a two-step analytical method for quantifying CNMs embedded in polymer composites was developed. This method consisted of polymer dissolution process using organic solvents followed by subsequent programmed thermal analysis (PTA). This quantification technique was applicable to CNTs with and without high metal impurities in a broad range of polymers. Various types of CNMs were coated on polyester fabric and the efficacy of coatings as flame retardant was evaluated. The oxygen content of CNMs emerged as a critical parameter impacting flame retardancy with higher oxygen content resulting in less FR efficacy. The most performant nanomaterials, multi-walled carbon nanotubes (MWCNTs) and amine functionalized multi-walled carbon nantoubes (NH2-MWCNT) showed similar FR properties to current flame retardants with low mass loading (0.18 g/m2) and hence are promising alternatives that warrant further investigation. Chemical/physical modification of MWCNTs was conducted to produce well-dispersed MWCNT solutions without involving oxygen for uniform FR coating. The MWCNTs coating was studied to evaluate the durability of the coating and the impact on the efficacy during use phase by conducting mechanical abrasion and washing test. Approximately 50% and 40% of MWCNTs were released from 1 set of mechanical abrasion and washing test respectively. The losses during simulated usage impacted the flame retardancy negatively.
ContributorsNosaka, Takayuki (Author) / Herckes, Pierre (Thesis advisor) / Westerhoff, Paul (Committee member) / Wang, Qing Hua (Committee member) / Arizona State University (Publisher)
Created2018
Description
Pipeline infrastructure forms a vital aspect of the United States economy and standard of living. A majority of the current pipeline systems were installed in the early 1900’s and often lack a reliable database reporting the mechanical properties, and information about manufacturing and installation, thereby raising a concern for their safety and integrity. Testing for the aging pipe strength and toughness estimation without interrupting the transmission and operations thus becomes important. The state-of-the-art techniques tend to focus on the single modality deterministic estimation of pipe strength and do not account for inhomogeneity and uncertainties, many others appear to rely on destructive means. These gaps provide an impetus for novel methods to better characterize the pipe material properties. The focus of this study is the design of a Bayesian Network information fusion model for the prediction of accurate probabilistic pipe strength and consequently the maximum allowable operating pressure. A multimodal diagnosis is performed by assessing the mechanical property variation within the pipe in terms of material property measurements, such as microstructure, composition, hardness and other mechanical properties through experimental analysis, which are then integrated with the Bayesian network model that uses a Markov chain Monte Carlo (MCMC) algorithm. Prototype testing is carried out for model verification, validation and demonstration and data training of the model is employed to obtain a more accurate measure of the probabilistic pipe strength. With a view of providing a holistic measure of material performance in service, the fatigue properties of the pipe steel are investigated. The variation in the fatigue crack growth rate (da/dN) along the direction of the pipe wall thickness is studied in relation to the microstructure and the material constants for the crack growth have been reported. A combination of imaging and composition analysis is incorporated to study the fracture surface of the fatigue specimen. Finally, some well-known statistical inference models are employed for prediction of manufacturing process parameters for steel pipelines. The adaptability of the small datasets for the accuracy of the prediction outcomes is discussed and the models are compared for their performance.
ContributorsDahire, Sonam (Author) / Liu, Yongming (Thesis advisor) / Jiao, Yang (Committee member) / Ren, Yi (Committee member) / Arizona State University (Publisher)
Created2018
Description
Photovoltaics (PV) is one of the promising options for maintaining sustainable energy supply because it is environmentally friendly, a non-polluting and low-maintenance energy source. Despite the many advantages of PV, solar energy currently accounts for only 1% of the global energy portfolio for electricity generation. This is because the cost of electricity from PV remains about a factor of two higher than the fossil fuel (10¢/kWh). Widely-used commercial methods employed to generate PV energy, such as silicon or thin film-based technologies, are still expensive as they are processed through vacuum-based techniques. Therefore, it is desirable to find an alternative method that is open-air and continuous process for the mass production of solar cells.
The objective of the research in this thesis is to develop low-cost spray pyrolysis technique to synthesize oxides thin films for applications in solar cells. Chapter 4 and 5 discuss spray-deposited dielectric oxides for their applications in Si solar cells. In Chapter 4, a successful deposition of Al2O3 is demonstrated using water as the solvent which ensures a lower cost and safer process environment. Optical, electrical, and structural properties of spray-deposited Al2O3 are investigated and compared to the industrial standard Atomic Layer Deposition (ALD) Al2O3/Plasma Enhanced Chemical Vapor Deposition (PECVD) SiNx stack, to reveal the suitability of spray-deposited Al2O3 for rear passivation and optical trapping in p-type Si Passivated Emitter and Rear Cell (PERC) solar cells. In Chapter 5, The possibility of using low-cost spray-deposited ZrO2 as the antireflection coating for Si solar cells is investigated. Optical, electrical and structural properties of spray-deposited ZrO2 films are studied and compared to the industrial standard antireflection coating PECVD SiNx. In Chapter 6, spray-deposited hematite Fe2O3 and sol-gel prepared anatase TiO2 thin films are sulfurized by annealing in H2S to investigate the band gap narrowing by sulfur doping and explore the possibility of using ternary semiconductors for their application as solar absorbers.
The objective of the research in this thesis is to develop low-cost spray pyrolysis technique to synthesize oxides thin films for applications in solar cells. Chapter 4 and 5 discuss spray-deposited dielectric oxides for their applications in Si solar cells. In Chapter 4, a successful deposition of Al2O3 is demonstrated using water as the solvent which ensures a lower cost and safer process environment. Optical, electrical, and structural properties of spray-deposited Al2O3 are investigated and compared to the industrial standard Atomic Layer Deposition (ALD) Al2O3/Plasma Enhanced Chemical Vapor Deposition (PECVD) SiNx stack, to reveal the suitability of spray-deposited Al2O3 for rear passivation and optical trapping in p-type Si Passivated Emitter and Rear Cell (PERC) solar cells. In Chapter 5, The possibility of using low-cost spray-deposited ZrO2 as the antireflection coating for Si solar cells is investigated. Optical, electrical and structural properties of spray-deposited ZrO2 films are studied and compared to the industrial standard antireflection coating PECVD SiNx. In Chapter 6, spray-deposited hematite Fe2O3 and sol-gel prepared anatase TiO2 thin films are sulfurized by annealing in H2S to investigate the band gap narrowing by sulfur doping and explore the possibility of using ternary semiconductors for their application as solar absorbers.
ContributorsShin, Woo Jung (Author) / Tao, Meng (Thesis advisor) / Goryll, Michael (Committee member) / Wang, Qing Hua (Committee member) / Arizona State University (Publisher)
Created2019
Description
Advanced material systems refer to materials that are comprised of multiple traditional constituents but complex microstructure morphologies, which lead to their superior properties over conventional materials. This dissertation is motivated by the grand challenge in accelerating the design of advanced material systems through systematic optimization with respect to material microstructures or processing settings. While optimization techniques have mature applications to a large range of engineering systems, their application to material design meets unique challenges due to the high dimensionality of microstructures and the high costs in computing process-structure-property (PSP) mappings. The key to addressing these challenges is the learning of material representations and predictive PSP mappings while managing a small data acquisition budget. This dissertation thus focuses on developing learning mechanisms that leverage context-specific meta-data and physics-based theories. Two research tasks will be conducted: In the first, we develop a statistical generative model that learns to characterize high-dimensional microstructure samples using low-dimensional features. We improve the data efficiency of a variational autoencoder by introducing a morphology loss to the training. We demonstrate that the resultant microstructure generator is morphology-aware when trained on a small set of material samples, and can effectively constrain the microstructure space during material design. In the second task, we investigate an active learning mechanism where new samples are acquired based on their violation to a theory-driven constraint on the physics-based model. We demonstrate using a topology optimization case that while data acquisition through the physics-based model is often expensive (e.g., obtaining microstructures through simulation or optimization processes), the evaluation of the constraint can be far more affordable (e.g., checking whether a solution is optimal or equilibrium). We show that this theory-driven learning algorithm can lead to much improved learning efficiency and generalization performance when such constraints can be derived. The outcomes of this research is a better understanding of how physics knowledge about material systems can be integrated into machine learning frameworks, in order to achieve more cost-effective and reliable learning of material representations and predictive models, which are essential to accelerate computational material design.
ContributorsCang, Ruijin (Author) / Ren, Yi (Thesis advisor) / Liu, Yongming (Committee member) / Jiao, Yang (Committee member) / Nian, Qiong (Committee member) / Zhuang, Houlong (Committee member) / Arizona State University (Publisher)
Created2018
Description
Coordination and control of Intelligent Agents as a team is considered in this thesis.
Intelligent agents learn from experiences, and in times of uncertainty use the knowl-
edge acquired to make decisions and accomplish their individual or team objectives.
Agent objectives are defined using cost functions designed uniquely for the collective
task being performed. Individual agent costs are coupled in such a way that group ob-
jective is attained while minimizing individual costs. Information Asymmetry refers
to situations where interacting agents have no knowledge or partial knowledge of cost
functions of other agents. By virtue of their intelligence, i.e., by learning from past
experiences agents learn cost functions of other agents, predict their responses and
act adaptively to accomplish the team’s goal.
Algorithms that agents use for learning others’ cost functions are called Learn-
ing Algorithms, and algorithms agents use for computing actuation (control) which
drives them towards their goal and minimize their cost functions are called Control
Algorithms. Typically knowledge acquired using learning algorithms is used in con-
trol algorithms for computing control signals. Learning and control algorithms are
designed in such a way that the multi-agent system as a whole remains stable during
learning and later at an equilibrium. An equilibrium is defined as the event/point
where cost functions of all agents are optimized simultaneously. Cost functions are
designed so that the equilibrium coincides with the goal state multi-agent system as
a whole is trying to reach.
In collective load transport, two or more agents (robots) carry a load from point
A to point B in space. Robots could have different control preferences, for example,
different actuation abilities, however, are still required to coordinate and perform
load transport. Control preferences for each robot are characterized using a scalar
parameter θ i unique to the robot being considered and unknown to other robots.
With the aid of state and control input observations, agents learn control preferences
of other agents, optimize individual costs and drive the multi-agent system to a goal
state.
Two learning and Control algorithms are presented. In the first algorithm(LCA-
1), an existing work, each agent optimizes a cost function similar to 1-step receding
horizon optimal control problem for control. LCA-1 uses recursive least squares as
the learning algorithm and guarantees complete learning in two time steps. LCA-1 is
experimentally verified as part of this thesis.
A novel learning and control algorithm (LCA-2) is proposed and verified in sim-
ulations and on hardware. In LCA-2, each agent solves an infinite horizon linear
quadratic regulator (LQR) problem for computing control. LCA-2 uses a learning al-
gorithm similar to line search methods, and guarantees learning convergence to true
values asymptotically.
Simulations and hardware implementation show that the LCA-2 is stable for a
variety of systems. Load transport is demonstrated using both the algorithms. Ex-
periments running algorithm LCA-2 are able to resist disturbances and balance the
assumed load better compared to LCA-1.
Intelligent agents learn from experiences, and in times of uncertainty use the knowl-
edge acquired to make decisions and accomplish their individual or team objectives.
Agent objectives are defined using cost functions designed uniquely for the collective
task being performed. Individual agent costs are coupled in such a way that group ob-
jective is attained while minimizing individual costs. Information Asymmetry refers
to situations where interacting agents have no knowledge or partial knowledge of cost
functions of other agents. By virtue of their intelligence, i.e., by learning from past
experiences agents learn cost functions of other agents, predict their responses and
act adaptively to accomplish the team’s goal.
Algorithms that agents use for learning others’ cost functions are called Learn-
ing Algorithms, and algorithms agents use for computing actuation (control) which
drives them towards their goal and minimize their cost functions are called Control
Algorithms. Typically knowledge acquired using learning algorithms is used in con-
trol algorithms for computing control signals. Learning and control algorithms are
designed in such a way that the multi-agent system as a whole remains stable during
learning and later at an equilibrium. An equilibrium is defined as the event/point
where cost functions of all agents are optimized simultaneously. Cost functions are
designed so that the equilibrium coincides with the goal state multi-agent system as
a whole is trying to reach.
In collective load transport, two or more agents (robots) carry a load from point
A to point B in space. Robots could have different control preferences, for example,
different actuation abilities, however, are still required to coordinate and perform
load transport. Control preferences for each robot are characterized using a scalar
parameter θ i unique to the robot being considered and unknown to other robots.
With the aid of state and control input observations, agents learn control preferences
of other agents, optimize individual costs and drive the multi-agent system to a goal
state.
Two learning and Control algorithms are presented. In the first algorithm(LCA-
1), an existing work, each agent optimizes a cost function similar to 1-step receding
horizon optimal control problem for control. LCA-1 uses recursive least squares as
the learning algorithm and guarantees complete learning in two time steps. LCA-1 is
experimentally verified as part of this thesis.
A novel learning and control algorithm (LCA-2) is proposed and verified in sim-
ulations and on hardware. In LCA-2, each agent solves an infinite horizon linear
quadratic regulator (LQR) problem for computing control. LCA-2 uses a learning al-
gorithm similar to line search methods, and guarantees learning convergence to true
values asymptotically.
Simulations and hardware implementation show that the LCA-2 is stable for a
variety of systems. Load transport is demonstrated using both the algorithms. Ex-
periments running algorithm LCA-2 are able to resist disturbances and balance the
assumed load better compared to LCA-1.
ContributorsKAMBAM, KARTHIK (Author) / Zhang, Wenlong (Thesis advisor) / Nedich, Angelia (Thesis advisor) / Ren, Yi (Committee member) / Arizona State University (Publisher)
Created2018