A new robust controller for non-linear periodic single-input/single-output systems using genetic algorithms

Abstract In this paper, a robust control approach designed for periodic systems and based on the well-known genetic algorithms is presented. This control scheme introduces the innovative concept of indirect encoding through finite Fourier series, which greatly contributes to the efficiency of the algorithm. Moreover, emerging concepts such as multi-parent crossover and local mutation are employed. These features bring undeniable exploitation and exploration capabilities to the algorithm, which are essential for controlling ever-changing environments such as our case of application: a cardiac bioreactor. For about a decade, our research team has been working on the development of this system for the growth of tissue-engineered heart valves. The work presented in this article address the major challenge of optimal and robust control of the flowrate in the bioreactor. Several in silico and physical simulations based on a simplified model of this complex system allowed for a quick development process and are presented in this paper. Finally, experimental validation of the simulations were conducted on the cardiac bioreactor. These confirmed that the proposed approach can effectively be used for the optimal and adaptive control of this non-linear periodic single-input/single-output system.