Loading...

Control of Nonlinear Dynamic Systems over the Packet Erasure Channel: Applications in Remote Control of Autonomous Vehicles

Parsa, Ali | 2018

682 Viewed
  1. Type of Document: Ph.D. Dissertation
  2. Language: Farsi
  3. Document No: 51430 (05)
  4. University: Sharif University of Technology
  5. Department: Electrical Engineering
  6. Advisor(s): Farhadi, Alireza
  7. Abstract:
  8. Abstract: In this dissertation, the problem of tracking and stabilization of nonlinear dynamic systems over the packet erasure channel is addressed. Using the idea of linearization of nonlinear system around current working point, and by the design of an encoder, decoder, and controller for the linearized system, state tracking and reference tracking of the nonlinear system are addressed. To prove the traking and stability of the system, the strong law of large numbers and the dwell time approach are used. By computer simulations, the theoretical development is applied on the nonlinear dynamic of the unicycle model, which is a suitable model for autonomous vehicles, such as drones, unmanned aerial vehicles, and underwater autonomous vehicles. In this part of the dissertation, for the first time, the problem of nonlinear control of dynamic systems over the packet erasure channel is addressed. Also the classical linearization method for controlling nonlinear systems is extended to the context of networked control systems.
    The next idea developed in this dissertation for tracking and stabilization of nonlinear dynamic systems over the packet erasure channel is based on a new nonlinear encoder and decoder, and uses the vertical axis partition instead of the horizontal axis for coding. The theoretical development based on this idea is also applied to the nonlinear dynamic of the unicycle model, and the satisfactory performance of the proposed encoder, decoder and controller is illustrated via computer simulations.
    Finally, this dissertation addresses the problem of tracking and stabilization of nonlinear dynamic systems over the Additive White Gaussian Noise channel (AWGN). The describing function method is used to find an approximate linear model for nonlinear continuous- time noiseless dynamic systems that have periodic response to sinusoidal inputs. For this problem, the communication channel between the sensor and controller is considered as a continuous-time Gaussian channel, which is subject to power constraint of the antenna. After finding an approximate linear system for the nonlinear dynamic system using the describing function method, the available results for linear control over analog AWGN channel are extended to account for linear continuous- time systems with non - real valued and multiple real valued eigenvalues and for tracking a non-zero reference signal. Then, by applying the extended results on the approximate linear system, a new control technique including an encoder, decoder and a controller is presented for state tracking, reference tracking and stability of nonlinear dynamic system over AWGN channel. The theoretical development is applied on the unicycle dynamic of autonomous vehicles via computer simulations. In this part of dissertation for the first time the frequency domain analysis tools are used for nonlinear control over communication channel and the problem of nonlinear control over AWGN channel is addressed
  9. Keywords:
  10. Tracking ; Stabilization ; Linearization ; Describing Functions ; Nonlinear Dynamical System ; Network Control ; Packet Erasure Channel

 Digital Object List

 Bookmark

No TOC