Gaussian adaptive PID control optimized via genetic algorithm applied to a step-down DC-DC converter

In this paper, the proposal of an adaptive PID based on Gaussian functions is presented. Gaussian functions are of interest because they are smooth, with smooth derivatives, upper and lower bounded and with an adjustable concavity. The resulting Gaussian Adaptive PID (GAPID) have six parameters to be defined, that are optimized for the lowest settling-time with low overshoot by employing Genetic Algorithm (GA). Here, the choice of keeping the GAPID parameters linked to the PID parameters has the advantage of sharing the same design requirements, making it easy to replace existing PID controllers in most systems. Comments about GA and the way it was used (parameters of GA, individuals, evolution over each generation) are presented too. This technique is applied to a step-down Buck converter. The performance enhancement is notable (52.8%). Simulation and experimental results are also shown.