Energy efficiency improvement through MPC-based peripherals management for an industrial process test-bench

摘要

High energy costs evince the growing need for energy efficiency in industrial companies. This paper presents a solution at the industrial machine level to obtain efficient energy consumption. Therefore, a controller inspired by the well-known model predictive control (MPC) strategy was developed for the management of peripheral devices. The validation of the control requires a test-bench to emulate the energy consumption of a manufacturing machine. The test-bench has four devices, two used to emulate the periodic and fixed energy consumption of the manufacturing process and two as peripherals, subject to rules associated with the process. Consequently, a subspace identification (SI) was employed to identify energy models to simulate the behavior of the device. As a final step, a performance comparison between a rule-based control (RBC) and the proposed predictive-like controller revealed the remarkable energy savings. The MPC results show an energy saving of around 3% with respect to RBC as well as an instant maximum energy consumption reduction of 8%, approximately.