I. Introduction
Research on Neural Networks (NN) as function approximators has been present in the control theory for an extensive time [1]. The fact that the control problem can be formulated as an optimization problem contributed to the initial motivation for the use of Artificial Intelligence mechanisms, especially in controlling intricate systems, such as the unwanted nonlinear cases [2]. From the several Intelligent control papers that have been published, we can observe a notorious gap between research and real-world applications. One of the root causes for this gap is the well-known disadvantage of Deep-learning (DL) models: Data-dependence and its black-box nature [3].