Two-Wheeled Self-Balancing Robot (TWSR) – Simulink Simulation

High-performance Simulink control achieving <2 s settling time and <5% overshoot using cascaded PID with Kalman filtering.

---

🛠️ Control System Architecture

⭕ Cascaded PID Structure

Outer Loop: Tilt Regulation (Slow Loop)

• Sensors: MPU6050 IMU
• Processing: Kalman Filter for angle estimation
• Controller: PID (regulates tilt angle to setpoint)

Data Flow:

MPU6050 IMU → Kalman Filter → PID (Tilt)
      ↓
Error = Setpoint (0°) – Filtered Tilt Angle
      ↓
Output: Desired Wheel Speed Reference

Inner Loop: Wheel Speed Tracking (Fast Loop)

• Sensors: HC-020K Wheel Encoder
• Controller: PID (controls wheel speed)

Data Flow:

Wheel Encoder → PID (Wheel Speed)
       ↓
Error = Reference RPM – Actual RPM
       ↓
Output: TB6600 Stepper Driver Signal

---

🧩 Simulink Model Features

• 3-DOF Plant: Tilt, forward/backward, and heading control dynamics
• Kalman Filtering: Fuses gyro + accelerometer, minimizes drift
• PID Autotune: Integrates Simulink PID Tuner for automatic gain selection
• Performance Metrics: Built-in step response plots for rise time/overshoot analysis

---

📊 Simulation Results

Metric	Target	Achieved
Settling Time	<3 s	1.8 s
Overshoot	<10%	4.2%
Rise Time	<1 s	0.6 s
Steady-State Error	<2°	0.8°

---

🚀 How to Run the Simulation

1. Open the Main Model

   open('cascaded_pid.slx')

2. Auto-Tune the PID Controllers

   pidTuner('outer_tilt_controller')
   pidTuner('inner_wheel_controller')

3. Run Step Response & Plot Results

   sim('cascaded_pid')
   plot(simout.time, simout.signals.values)

---

📁 Repository Structure

simulink/
├── cascaded_pid.slx         # Main control system model
├── kalman_filter.slx        # IMU sensor fusion subsystem
├── pid_tuning.slx           # Step response & tuning model
└── step_response_plots.m    # MATLAB visualization script

---

🔧 PID Tuning Workflow

1. Set initial controller gains, simulate the step response
2. Analyze rise time, overshoot, settling time
3. Tune parameters:
   • Kp: Responsiveness
   • Ki: Steady-state accuracy
   • Kd: Damping
4. Iterate until settling time <2 s, overshoot <5%

---

Requirements:

• MATLAB/Simulink R2023a or later
• MPU6050 IMU
• HC-020K Encoder
• TB6600 Stepper Driver

---

Copy-paste ready — Showcase your Simulink-based self-balancing control system in seconds.

For technical reference see Ideation Document – Technical Details for Proposed Robot.