Embedded Files
This is the project for Linear Control Systems Class
Project Introduction
Project Introduction
In this class project, we delve into the dynamic realm of autonomous systems and unmanned aerial vehicles (UAVs) using Webots, a versatile open-source robotics simulation software. The overarching objective of this undertaking is to gain profound insights into control methods, a critical aspect in the development of efficient and intelligent autonomous vehicles.
As we navigate through the project, we aim to unravel the complexities of control mechanisms through a systematic approach. The outlined phases involve linearizing the state space model, designing PID controllers for both lateral and longitudinal control, ensuring the controllability and stabilizability of the system, and exploring advanced techniques such as designing an optimal lateral controller and implementing the A* path planning algorithm.
Furthermore, we venture into the intricacies of Simultaneous Localization and Mapping (SLAM) through the implementation of the Extended Kalman Filter (EKF). This enables us to control the vehicle even in the absence of default sensor input. The ultimate challenge awaits as we engage in friendly competition, racing against other Buggy competitors within the class.
By utilizing Webots as our simulation platform, we not only acquire technical expertise but also contribute to the broader field of robotics and autonomous systems. The hands-on experience gained from this project lays a foundation for understanding real-world challenges in autonomous vehicle control, making meaningful strides toward the future of intelligent transportation and robotics.
Methods
Methods
The project has been divided into 4 parts to complete:
Phase 1: Linearization and PID Controller Design
Task 1a: Linearize the state space model
Task 1b: Design PID controllers for lateral and longitudinal control
Phase 2: Controllability Analysis and PID Controller Refinement
Task 2a: Assess the controllability and stabilizability of the linearized system
Task 2b: Refine PID controllers for lateral and longitudinal control
Phase 3: Optimal Control and Path Planning
Task 3a: Design an optimal lateral controller
Task 3b: Implement the A* path planning algorithm
Phase 4: Sensor Fusion and Competitive Racing
Task 4a: Implement Extended Kalman Filter (EKF) SLAM for vehicle control without default sensor input
Task 4b: Engage in a competitive race with other Buggy competitors in the class
We uses a bicycle model for the vehicle, a popular choice in the study of vehicle dynamics. As depicted in Figure 1, the car is modeled as a two-wheel vehicle with two degrees of freedom, described separately in longitudinal and lateral dynamics.
Results - Demo Video
Results - Demo Video
Proportional–Integral–Derivative(PID) Controller
Proportional–Integral–Derivative(PID) Controller
For both vehicle longitudinal and lateral motion control.
State Feedback Controller Using Pole Placement
State Feedback Controller Using Pole Placement
For only the vehicle lateral motion control, the longitudinal motion is controlled using PID.
Linear Quadratic Regulator Controller
Linear Quadratic Regulator Controller
For only the vehicle lateral motion control, the longitudinal motion is controlled using PID.
A* Path Planning
A* Path Planning
The vehicle lateral motion is controlled with LQR, the longitudinal motion is controlled using PID.
Conclusion
Conclusion
Through Webots, our simulation platform of choice, we not only cultivate technical expertise but also contribute to the broader landscape of robotics and autonomous systems. The hands-on experience garnered in this project lays a solid foundation for tackling real-world challenges in autonomous vehicle control, propelling us toward a future of intelligent transportation and robotics.
Page updated
Google Sites
Report abuse