ROS2 PID Controller for Velocity

Problem Statement

Implement a PID velocity controller as a ROS2 node that drives a robot's actual velocity to a configurable target.

Requirements

Your node must:

• Declare a parameter target_vel (default 0.5) — the desired linear velocity in m/s
• Subscribe to /current_velocity (std_msgs/Float32) for the measured velocity feedback
• Implement a PIDController class with method compute(setpoint, measurement, dt) → float
  • P term: kp * error
  • I term: accumulated ki * error * dt — clamp to [-1.0, 1.0] (anti-windup)
  • D term: kd * (error - prev_error) / dt
• Use a create_timer(0.05, ...) (20 Hz) control loop that:
  • Reads target_vel parameter each cycle (allows runtime updates via ros2 param set)
  • Calls pid.compute(target, measured, 0.05)
  • Clamps output to [-1.0, 1.0]
  • Publishes geometry_msgs/Twist to /cmd_vel with linear.x = clamped_output
• Log "PID Controller Ready" on startup

Input / Output

Input: std_msgs/Float32 on /current_velocity (measured speed) Output: geometry_msgs/Twist on /cmd_vel

🤖 Why This Matters in Real Robots

The ros2_control stack internally runs PID controllers for every hardware joint. Understanding how to implement one from scratch is essential before using the framework. Tune kp=1.0, ki=0.1, kd=0.05 and observe the step response — underdamped (oscillates), overdamped (slow), or critically damped (fast + smooth).

Input/Output Format

Input: /target_vel, /current_vel | Output: /cmd_vel

⚠️ Common Pitfalls

• Integral windup
• Derivative noise spikes

📚 Helpful Resources

• ROS2 Concept: PID Controller for Velocity