A Comparative Study of PID, LQG, and MPC Controllers for Temperature Control in Solder Reflow Ovens

This paper presents an experimental comparison of three control strategies: Proportional-Integral-Derivative (PID), Linear Quadratic Gaussian (LQG), and Model Predictive Control (MPC) for temperature control in a solder reflow oven application. An accurate model of the oven's thermal dynamics was developed through system identification. Each controller was designed and implemented in MATLAB/Simulink and on hardware, with PID gains optimized using Particle Swarm Optimization. The controllers were evaluated based on their step response characteristics and their ability to track a typical reflow temperature profile for 183° C solder paste. Results indicate that while the PID controller provided balanced performance and the LQG controller achieved the fastest rise time, the MPC controller demonstrated superior overall performance with minimal overshoot (0.02 %), quick settling time (<tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{1. 0 6 ~ s}$</tex>), and excellent steady-state accuracy (steady-state error of -0.0002). These findings suggest that MPC is the most suitable approach for precise temperature control in reflow soldering applications, offering the best combination of accuracy, stability, and constraint-handling capabilities.