1 of 24

Ensuring Tolerances for Trajectory Tracking of Non-Holonomic Robots under Disturbance and Measurement Delay

Controls and Robotics Seminar Series

Frank Lawless

April 1st, 2022

Robotics and Controls Seminar

1

4/01/2022

COAR LAB

2 of 24

Trajectory Planning for Robotic Systems

Applications

    • Ground vehicles
    • Autonomous underwater vehicles
    • Flying robots

Robotics and Controls Seminar

2

4/01/2022

E. Masehian and N. Mohamadnejad, "Path planning of nonholonomic flying robots using a new virtual obstacle method"

3 of 24

Overview

  • Trajectory Tracking
  • Problem Formulation
  • Control I Design
  • Control II Design
  • Introducing Noise and Delay
  • Tolerances
  • Simulation Results
  • Future Works

Robotics and Controls Seminar

3

4/01/2022

4 of 24

Trajectory Tracking for Mobile Robots

  •  

Robotics and Controls Seminar

4

4/01/2022

Reference Trajectory

Robot Trajectory

Robot Trajectory

5 of 24

Previous Works

Robotics and Controls Seminar

5

4/01/2022

K. Shojaei, A. M. Shahri, A. Tarakameh, and B. Tabibian, “Adaptive trajectory tracking control of a differential drive wheeled mobile robot,” 2011

Plamen Petrov and Ivan Kralov, “A Look-ahead approach to mobile robot path tracking based on distance-only measurements,” 2019

R. L. S. Sousa, M. D. do Nascimento Forte, F. G. Nogueira and B. C. Torrico, "Trajectory tracking control of a nonholonomic mobile robot with differential drive," 2016

Adaptive Feedback Linearization

Look-ahead Feedback Linearization

 

Feedforward Control

 

Compensate for parametric uncertainty

6 of 24

The Problem

    • Problem Description

-Develop a controller such that the tracking error E(t) converges to zero. �-Develop a controller to track any human provided reference trajectory.

    • Guarantee Tolerances

-Maximum error

-Maximum velocities

    • Decrease High Error Velocities

-Develop a controller such that E(t) converges non-exponentially

    • State-transition model

-E(t) converges in a linear manner at high error and exponentially after a user defined threshold

Robotics and Controls Seminar

5

4/01/2022

Control

Control II

Control I

 

 

Exponential

Non-

Exponential

7 of 24

Differential Drive Robots

Robotics and Controls Seminar

7

4/01/2022

 

Centroid Dynamics

Base-point Dynamics

8 of 24

Controller I Design

Robotics and Controls Seminar

8

4/01/2022

 

Proposed state-feedback controller

 

9 of 24

Reference Vs. Robot Trajectory

Robotics and Controls Seminar

9

4/01/2022

 

D = 1

D = 10

 

 

10 of 24

Control I Velocity Bounds

Robotics and Controls Seminar

10

4/01/2022

  • Saturation Velocities

-Maximum Linear and Angular Velocities

11 of 24

Control I Velocity Bounds

Robotics and Controls Seminar

11

4/01/2022

  • Develop maximum bounds for velocities

Bounded velocities

12 of 24

Simulation Results of Control I

Robotics and Controls Seminar

12

4/01/2022

Reference Trajectory

13 of 24

Control II Design

Robotics and Controls Seminar

13

4/01/2022

 

Proposed High Error controller

Control I

Control II

Through Lyapunov analysis

14 of 24

Control II Velocity Bounds

Robotics and Controls Seminar

14

4/01/2022

Bounded velocities

  • Develop maximum bounds for velocities

15 of 24

Transition Model

Robotics and Controls Seminar

15

4/01/2022

 

 

 

 

Control II

Control I

Control

m = 1

m = .1

  • Maintain linear convergence at high error while still converging to zero with time.

16 of 24

Simulation Results of Control I&II

Robotics and Controls Seminar

16

4/01/2022

Control I

Control I&II

17 of 24

Introducing noise

Robotics and Controls Seminar

17

4/01/2022

Two main ways for noise to be introduced into the system:

  • Hardware: Robot components - wheel actuators.
  • Positional: Robot location - GPS data.

 

-Error tolerance must be decided by user

-High noise might result in the robot to veer off the reference trajectory causing operational and safety concerns

18 of 24

Delay

Robotics and Controls Seminar

18

4/01/2022

Estimated heading angle

When current position cannot be determined the heading angle must be estimated.

Bounded heading angle error

Estimated heading angle error

Initial heading angle

Inaccurate guess

Better guess

Finding the error in heading angle

19 of 24

Lyapunov Stability for control I

Lyapunov function

Robotics and Controls Seminar

19

4/01/2022

Prove E(t) converges to zero with noise and delay

20 of 24

Lyapunov Stability for control I

Lyapunov function

Through Young’s inequality we can bound our relationship

Robotics and Controls Seminar

20

4/01/2022

 

To guarantee E(t) to converge to zero

Prove E(t) converges to zero with noise and delay

21 of 24

Error Bounds

Robotics and Controls Seminar

21

4/01/2022

Error with high noise

Error with low noise

Bounded error with noise

22 of 24

Noise and Delay Simulation Results

Robotics and Controls Seminar

22

4/01/2022

23 of 24

Conclusion and Future Works

Robotics and Controls Seminar

23

4/01/2022

  • Control II Tolerances with noise and delay
  • Event-based Control
  • TurtleBot3 Implementation
  • Generalize and implement control for alternative dynamics

24 of 24

Thank you�Q&A

ROBOTICS AND CONTROLS SEMINAR

24

4/01/2022