Data-Driven Cooperative Intelligent Controller Based on the Endocrine Regulation Mechanism

Abstract

A data-driven mechanism can achieve effective control by utilizing the online/offline data of the target system, although its performance has not been tuned to a better level. The endocrine regulating mechanism in the human body establishes a rapid responding system to maintain the balance of the body, which can be mathematically derived and therefore provide an inspiration for optimizing the industrial controller. In this paper, a novel data-driven cooperative intelligent controller inspired by the regulating principle of the endocrine system in the human body is proposed. The data-driven component of the proposed controller optimizes the controller parameters by collecting and processing runtime data of the target system. The endocrine regulation-inspired enhancing component tunes the intensity of control signals adaptively. Both the components are further organized by an adaptive distributor so that their behaviors can be regulated dynamically. A dynamic tension control system for acrylic fiber production is taken to verify the performance of the proposed controller. Simulation results show that the proposed controller can realize effective control on systems with unknown or varying models, meanwhile featuring rapid response and effective regulation against external disturbance.