A thermoelectric module was modeled and analyzed in COMSOL Multiphysics using a finite element method (FEM). An optimum coefficient of performance of 0.4 is achieved at an operational current of 2 A, and a maximum heat load of 0.52 W is obtained at zero temperature gradient. The system’s dynamic coolingbehavior was simulated and the input and output data were exported to the system identification toolbox in MATLAB for the prediction of the mathematical model. The predicted model was validated and a suitable proportional integral (PI) controller was designed for the system. However, the controller needs to beoptimized. Optimization techniques namely: Grey Wolf Optimizer (GWO), Particle Swarm Optimization (PSO), Artificial Bee Colony (ABC), and Harris Hawk Optimization (HHO) were tested on the controller. GWO having a settling time of 199 seconds and an overshoot of 5.96% is found to be the best for this model.