Experimental Analysis of Real-time Energy Control Potentials for Sustainable Manufacturing Systems

Energy efficiency improvement as well as carbon footprint reduction in modern manufacturing systems has been of high interests to both academia and industry in recent years. However, most existing research efforts in energy efficiency improvement only focus on either single machine or process level due to the complexity of modern manufacturing systems, and few literatures concentrating on the energy efficiency improvement for the typical manufacturing systems with multiple machines and buffers can be found. Therefore, there lacks a concrete understanding related to the potential of the energy efficiency improvement by real-time energy control strategy and the practicality of the integration of energy control module into the existing control systems. Generally, to improve the energy efficiency of manufacturing systems with multiple machines and buffers, the following steps are designed: a) Identify the opportunities of energy control for each machine; b) Decide the optimal state of each machine in the system; c) Make decisions and execute actions; and d) Repeat steps a), b) and c) for continuous improvement. Due to the infeasibility of validating energy control models in real manufacturing lines practically, lab based experiment becomes a useful tool in analyzing the performance of those models. In this thesis, an experimental based method is proposed to study various strategies for energy efficiency improvement of complex manufacturing systems. A typical production line with multiple machines and buffers is established in both software testbed and hardware testbed, and a framework of real time energy control is also proposed and implemented in both testbeds. The objective of this thesis is to 1) identify the potential of energy savings in modern manufacturing systems with multiple machines and buffers by energy control policy, 2) examine the feasibility of the application new energy control module under the existing control system and 3) provide a generalized testbed framework with different functionality modules by using object-oriented programming which can be easily adjusted and fit into different systems for different research purposes.