Purpose

The purpose of this repository/training course is to build a foundational understanding and skills for GNC topics including data acquisition and filtering, fusion algorithms, basic system controllers, and then touching on more realistic and advanced system controllers.

Control Systems Training Outline

Phase 1: Foundations

Lesson 1: Reading Sensor Data

• Purpose: Learn to acquire raw accelerometer data from the Pixhawk and visualize it in MATLAB.
• Focus: Connection setup, basic data sampling, and visualization.

Lesson 2: Simple Data Filtering

• Purpose: Introduce filtering methods (e.g., moving average) to reduce noise and understand the impact of noise on data.
• Focus: Compare raw vs. filtered data and understand why filtering is crucial.

Lesson 3: Gyroscope Integration

• Purpose: Extend to gyroscope data to understand rotational velocity and its relationship with accelerometer data.
• Focus: Data acquisition and logging for multi-sensor systems.

Lesson 4: Complementary Filter

• Purpose: Combine accelerometer and gyroscope data to estimate orientation with a simple sensor fusion technique.
• Focus: Practical introduction to basic sensor fusion concepts.

Lesson 5: Introduction to Kalman Filtering

• Purpose: Implement a basic Kalman filter for 1D data to understand state estimation and uncertainty reduction.
• Focus: Step-by-step breakdown of Kalman filter theory and implementation.

Phase 2: Control Systems

Lesson 6: Simulink Basics

• Purpose: Familiarize with Simulink and model simple physical systems like a mass-spring-damper.
• Focus: Basic modeling, simulation setup, and visualization in Simulink.

Lesson 7: Implementing a PID Controller

• Purpose: Build a PID controller in MATLAB and Simulink to control a simulated system.
• Focus: Understanding PID tuning parameters and their effect on system behavior.

Lesson 8: Pixhawk PID Loop

• Purpose: Implement a PID controller on the Pixhawk for controlling a simple simulated or real system.
• Focus: Transitioning from simulation to hardware-in-the-loop (HITL).

Phase 3: Advanced Topics

Lesson 9: Extended Kalman Filter (EKF)

• Purpose: Implement an EKF for attitude estimation using accelerometer and gyroscope data.
• Focus: Multi-dimensional Kalman filtering and advanced sensor fusion.

Lesson 10: System Identification

• Purpose: Model a physical system by analyzing input-output data using system identification techniques.
• Focus: Building accurate models for control design.

Lesson 11: Sensor Fusion Algorithms

• Purpose: Explore and compare advanced sensor fusion algorithms (e.g., Madgwick, Mahony).
• Focus: Implement and evaluate algorithms for orientation estimation.

Lesson 12: Simulating a Control System with HITL

• Purpose: Integrate sensor fusion and control in a hardware-in-the-loop setup with Pixhawk and Simulink.
• Focus: Bridging simulation and hardware for real-world applications.

Phase 4: Advanced Applications

Lesson 13: 2D Control Systems

• Purpose: Simulate a 2D control system (e.g., a planar pendulum) using sensor fusion and PID.
• Focus: Multi-axis control and enhanced visualization.

Lesson 14: Multiple Input Single Output (MISO) Systems

• Purpose: Model and control systems with multiple inputs and a single output, such as balancing forces from different actuators.
• Focus: Practical handling of MISO systems in MATLAB and Simulink.

Lesson 15: Coupled Systems

• Purpose: Analyze and control systems with coupled dynamics (e.g., two masses connected by a spring).
• Focus: Advanced modeling and simulation techniques.

Lesson 16: Nonlinear Systems

• Purpose: Explore the challenges of controlling nonlinear systems, such as systems with friction or saturation.
• Focus: Understanding limitations of linear control methods and introducing nonlinear approaches.

Lesson 17: Model Predictive Control (MPC)

• Purpose: Introduce model predictive control and its application to advanced control problems.
• Focus: Learn to anticipate future states and optimize control inputs.