Trajectory tracking control of rear-steered AGVs via LQR and disturbance observer

다국어 초록 (Multilingual Abstract)

Accurate trajectory tracking is important for automated guided vehicles operating in warehouses and manufacturing environments, particularly during turning and transitions between straight and curved path sections. This thesis develops a simulation-based trajec- tory tracking control structure for a rear-driven, rear-steered three-wheel forklift AGV, where tracking performance depends on both vehicle-level error regulation and the re- sponse of the driving and steering motors. A rear-steered kinematic model was used to formulate a four-state tracking-error model consisting of longitudinal, lateral, heading, and velocity errors. Based on the linearized error dynamics, a Linear Quadratic Regulator was designed to generate velocity and steering correction commands. A Genetic Algorithm was applied offline to optimize the LQR weighting parameters and reduce dependence on man- ual trial-and-error tuning. Local motor controllers of driving and steering systems were designed based on feedforward control, feedback control and disturbance observer com- pensation. Identified motor parameters were incorporated into a MATLAB/Simulink model to represent actuator dynamics. The proposed GA-LQR-DOB controller was eval- uated on a racetrack trajectory and compared with PID and manually tuned LQR control- lers. The simulation results indicated improved trajectory tracking, reduced lateral and heading deviations, and quicker recovery from curved and transition sections, with stable velocity and steering command tracking throughout the simulation.

목차 (Table of Contents)

• Abstract 1
• 1. Introduction 2
• 1.1. Background and Motivation 2
• 1.2. Problem Statement 3
• 1.3. Research Objectives 4
• 1.4. Scope and Limitations 5
• 1.5. Thesis Contributions 5
• 1.6. Structure of Thesis 6
• 2. Literature Review 8
• 2.1. Automated Guided Vehicles and Warehouse Automation 8
• 2.2. Trajectory Tracking Control of Mobile Robots 8
• 2.3. Rear-Steered AGV and Forklift-Type Vehicle Control 8
• 2.4. LQR-Based Control Methods 9
• 2.5. Disturbance Observer-Based Control 10
• 2.6. Genetic Algorithm-Based Controller Tuning 10
• 2.7. Research Gap and Motivation 10
• 3. System Description and Mathematical Modeling 12
• 3.1. AGV System Configuration 12
• 3.2. Kinematic Model of the Rear-Steered AGV 14
• 3.3. Trajectory Tacking Error Model 16
• 3.4. Linearized State-Space Model 18
• 3.5. System Identification 19
• 3.6. Trajectory Generation 20
• 4. Controller Design 22
• 4.1. Overall Control Architecture 22
• 4.2. LQR-Based Trajectory Tracking Control 23
• 4.3. Genetic Algorithm-Based LQR Gain Optimization 25
• 4.4. Local Motor Controller Design 27
• 5. Simulation and Results 30
• 5.1. Local Motor Tracking Performance 30
• 5.2. Tracking Error Analysis of the Proposed Controller 31
• 5.3. Trajectory Tracking Comparison 32
• 6. Conclusion and Future Work. 35
• 7. Reference 37
• Appendix 40