본문

서브메뉴

Control of Agentic Systems Using the Newton-Raphson Controller
Control of Agentic Systems Using the Newton-Raphson Controller
Control of Agentic Systems Using the Newton-Raphson Controller

Detailed Information

자료유형  
 학위논문 서양
최종처리일시  
20260202105326
ISBN  
9798263324032
DDC  
500
저자명  
Niu, Kaicheng.
서명/저자  
Control of Agentic Systems Using the Newton-Raphson Controller
발행사항  
[Sl] : Georgia Institute of Technology, 2025
발행사항  
Ann Arbor : ProQuest Dissertations & Theses, 2025
형태사항  
118 p
주기사항  
Source: Dissertations Abstracts International, Volume: 87-05, Section: B.
주기사항  
Advisor: Abdallah, Chaouki T.;Wardi, Yorai.
학위논문주기  
Thesis (Ph.D.)--Georgia Institute of Technology, 2025.
초록/해제  
요약The Newton-Raphson Controller is a tracking controller built on the ideas of output prediction and the Newton-Raphson method. Existing results have already shown that the controller is capable of controlling a variety of nonlinear systems, including the inverted pendulums, autonomous vehicles and quadrotors. Compared to other tracking control approaches, the Newton-Raphson Controller features similar or better tracking accuracies while requiring less computational resources.Despite its early success, several problems still remain in the current version of the Newton-Raphson Controller. To begin with, this controller is model-based, meaning that it requires a mathematical model of the controlled plant in order to calculate a proper control input. Moreover, although there has already been a verifiable stability criterion of the Newton-Raphson Controller for linear systems, such criterion for nonlinear systems is yet to be proposed. Lastly, this controller can only be applied to single-agent systems. To apply the Newton-Raphson Controller in a multi-agent system, we must explicitly design the reference signal for each agent and convert the multi-agent control problem into multiple single-agent control problems.The problems mentioned above have prevented the use of the Newton-Raphson Controller in many control applications. Therefore, in this dissertation, we attempt to solve these problems and improve the existing controller, such that it can be applied to more control scenarios. Specifically, we study the following aspects of the Newton-Raphson Controller: First, we attempt to make the Newton-Raphson Controller model-free using the Deep Neural Network (DNN), such that the controller does not rely on the mathematical models of the controlled plants. Second, we perform stability analysis and propose verifiable stability criteria of the controller for a class of differentially flat systems. Third, we try to extend the current single-agent controller to a multi-agent controller, such that it can achieve leaderless consensus over a class of heterogeneous multi-agent systems.Fourth, we develop the leaderless consensus controller into the leader-follower consensus controller, which allows the controlled agents to follow a leader or a reference signal while maintaining consensus. Finally, to avoid inter-agent collisions during the consensus control, we propose to apply the Integral-Control Barrier Function to maintain a proper distance between agents and keep the agents safe. We will demonstrate the effectiveness of the improved controller using a number of simulations on different nonlinear systems and show that the proposed methods are capable of solving a variety of control problems. The simulation results together with the theoretical developments suggest a potential of the improved controller in future applications.
일반주제명  
Kinematics
일반주제명  
Communication channels
일반주제명  
Control algorithms
일반주제명  
Deep learning
일반주제명  
Bicycles
일반주제명  
Mathematical models
일반주제명  
Neural networks
일반주제명  
Controllers
일반주제명  
Flexibility
일반주제명  
Online instruction
일반주제명  
Design
일반주제명  
Systems stability
일반주제명  
Ordinary differential equations
일반주제명  
Educational technology
일반주제명  
Mathematics
일반주제명  
Transportation
기타저자  
Georgia Institute of Technology.
기본자료저록  
Dissertations Abstracts International. 87-05B.
전자적 위치 및 접속  
로그인 후 원문을 볼 수 있습니다.

MARC

 008260126s2025        us                              c    eng  d
■001000017360239
■00520260202105326
■006m          o    d                
■007cr#unu||||||||
■020    ▼a9798263324032
■035    ▼a(MiAaPQ)AAI32307879
■035    ▼a(MiAaPQ)GeorgiaTech78666
■040    ▼aMiAaPQ▼cMiAaPQ
■0820  ▼a500
■1001  ▼aNiu,  Kaicheng.
■24510▼aControl  of  Agentic  Systems  Using  the  Newton-Raphson  Controller
■260    ▼a[Sl]▼bGeorgia  Institute  of  Technology▼c2025
■260  1▼aAnn  Arbor▼bProQuest  Dissertations  &  Theses▼c2025
■300    ▼a118  p
■500    ▼aSource:  Dissertations  Abstracts  International,  Volume:  87-05,  Section:  B.
■500    ▼aAdvisor:  Abdallah,  Chaouki  T.;Wardi,  Yorai.
■5021  ▼aThesis  (Ph.D.)--Georgia  Institute  of  Technology,  2025.
■520    ▼aThe  Newton-Raphson  Controller  is  a  tracking  controller  built  on  the  ideas  of  output  prediction  and  the  Newton-Raphson  method.  Existing  results  have  already  shown  that  the  controller  is  capable  of  controlling  a  variety  of  nonlinear  systems,  including  the  inverted  pendulums,  autonomous  vehicles  and  quadrotors.  Compared  to  other  tracking  control  approaches,  the  Newton-Raphson  Controller  features  similar  or  better  tracking  accuracies  while  requiring  less  computational  resources.Despite  its  early  success,  several  problems  still  remain  in  the  current  version  of  the  Newton-Raphson  Controller.  To  begin  with,  this  controller  is  model-based,  meaning  that  it  requires  a  mathematical  model  of  the  controlled  plant  in  order  to  calculate  a  proper  control  input.  Moreover,  although  there  has  already  been  a  verifiable  stability  criterion  of  the  Newton-Raphson  Controller  for  linear  systems,  such  criterion  for  nonlinear  systems  is  yet  to  be  proposed.  Lastly,  this  controller  can  only  be  applied  to  single-agent  systems.  To  apply  the  Newton-Raphson  Controller  in  a  multi-agent  system,  we  must  explicitly  design  the  reference  signal  for  each  agent  and  convert  the  multi-agent  control  problem  into  multiple  single-agent  control  problems.The  problems  mentioned  above  have  prevented  the  use  of  the  Newton-Raphson  Controller  in  many  control  applications.  Therefore,  in  this  dissertation,  we  attempt  to  solve  these  problems  and  improve  the  existing  controller,  such  that  it  can  be  applied  to  more  control  scenarios.  Specifically,  we  study  the  following  aspects  of  the  Newton-Raphson  Controller:  First,  we  attempt  to  make  the  Newton-Raphson  Controller  model-free  using  the  Deep  Neural  Network  (DNN),  such  that  the  controller  does  not  rely  on  the  mathematical  models  of  the  controlled  plants.  Second,  we  perform  stability  analysis  and  propose  verifiable  stability  criteria  of  the  controller  for  a  class  of  differentially  flat  systems.  Third,  we  try  to  extend  the  current  single-agent  controller  to  a  multi-agent  controller,  such  that  it  can  achieve  leaderless  consensus  over  a  class  of  heterogeneous  multi-agent  systems.Fourth,  we  develop  the  leaderless  consensus  controller  into  the  leader-follower  consensus  controller,  which  allows  the  controlled  agents  to  follow  a  leader  or  a  reference  signal  while  maintaining  consensus.  Finally,  to  avoid  inter-agent  collisions  during  the  consensus  control,  we  propose  to  apply  the  Integral-Control  Barrier  Function  to  maintain  a  proper  distance  between  agents  and  keep  the  agents  safe.  We  will  demonstrate  the  effectiveness  of  the  improved  controller  using  a  number  of  simulations  on  different  nonlinear  systems  and  show  that  the  proposed  methods  are  capable  of  solving  a  variety  of  control  problems.  The  simulation  results  together  with  the  theoretical  developments  suggest  a  potential  of  the  improved  controller  in  future  applications.
■590    ▼aSchool  code:  0078.
■650  4▼aKinematics
■650  4▼aCommunication  channels
■650  4▼aControl  algorithms
■650  4▼aDeep  learning
■650  4▼aBicycles
■650  4▼aMathematical  models
■650  4▼aNeural  networks
■650  4▼aControllers
■650  4▼aFlexibility
■650  4▼aOnline  instruction
■650  4▼aDesign
■650  4▼aSystems  stability
■650  4▼aOrdinary  differential  equations
■650  4▼aEducational  technology
■650  4▼aMathematics
■650  4▼aTransportation
■690    ▼a0389
■690    ▼a0800
■690    ▼a0710
■690    ▼a0405
■690    ▼a0709
■71020▼aGeorgia  Institute  of  Technology.
■7730  ▼tDissertations  Abstracts  International▼g87-05B.
■790    ▼a0078
■791    ▼aPh.D.
■792    ▼a2025
■793    ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17360239▼nKERIS▼z이  자료의  원문은  한국교육학술정보원에서  제공합니다.

Preview

Export

ChatGPT Discussion

AI Recommended Related Books


    New Books MORE
    Statistics for the past 3 years. Go to brief

    Подробнее информация.

    • Бронирование
    • не существует
    • моя папка
    • Первый запрос зрения
    • Non-Book Loan Application
    • Nighttime Book Loan Application
    материал
    Reg No. Количество платежных Местоположение статус Ленд информации
    TF14476 전자도서 대출가능 My Folder 부재도서신고 비도서대출신청 야간 도서대출신청

    * Бронирование доступны в заимствований книги. Чтобы сделать предварительный заказ, пожалуйста, нажмите кнопку бронирование

    Books borrowed together with this book

    Related Popular Books

    Available after logging in.