Prototype and Verification of Advanced Controller for Customized Autonomous Vehicle

  • 劉 保良

Student thesis: Doctoral Thesis

Abstract

For customized autonomous vehicles the control method with the online pose updating to achieve a high-precision path-tracking control has been presented in the literature This thesis extends the presented technology under the concept of Level 6 of Technology Readiness Index (i e system/subsystem model or prototype demonstration in a relevant environment) to create a prototype of an advanced controller for a customized autonomous vehicle Main technologies of this prototype include (i) the multi-model-based modified model predictive control (MPC) with input constraints (ii) a modified supervisory control structure for autonomous vehicles (iii) a new mechanism for selections of multi-system model sequences and (iv) a verification for highly complex paths to have a system/subsystem model or prototype demonstration in a relevant environment Since the vehicle dynamic system is an unknown system multiple models are established according to its characteristics to construct corresponding perfect state estimators The modified MPC is combined with the modified supervisory control structure to track a given path and activate the pose updating online to have a precise path-tracking performance Then based on Hildreth’s quadratic programming a modified observer-based MPC with input constraints is utilized for the proposed method In the end we use the Vehicle Dynamics-MATLAB & Simulink-MathWorks 2018a as a verified system to verify the feasibility of the proposed method We apply to six driving scenarios including three real paths obtained from Google Map after processing to show the effectiveness of the created prototype of an advanced controller for a customized autonomous vehicle
Date of Award2020
Original languageEnglish
SupervisorJason Sheng-Hon Tsai (Supervisor)

Cite this

'