Energy saving policies for a machine tool with warm-up, stochastic arrivals and buffer information

Abstract

One of the measures for saving energy in manufacturing is the implementation of control strategies that reduces energy consumption during the machine idle periods. Specifically, the paper proposes a framework that integrates different control policies for switching the machine off when the production is not critical, and on either when the part flow has to be resumed or the queue has accumulated to a certain level. A general policy is formalized by modeling explicitly the power consumed in each machine state. A threshold policy is analyzed and the optimal parameter is provided for an M/M/1/K system. Numerical results are based on data acquired with dedicated experimental measurements on a real machining centre, and a comparison with common practices in manufacturing is also reported

Cross-level model of a transfer machine energy demand using a two-machine generalized threshold representation

“…Similar assumption for power demand of a MT was considered Frigerio et al (2013) who using a Markov-chain "automata"based approach to define an optimum switch-off strategy, knowing only part-arrival distribution, but not knowing the state of remaining machines. Other implemented policies, relying on time of last parts departure/arrival (Frigerio and Matta, 2014), used the same assumption: constant power demand per state. Similar assumption was used by Zavanella et al (2015), who applied queueing theory to realize a switch-on model of a machine in an industrial manufacturing system.…”

“…In recent years, research community developed methods for finding optimum work-point of a single processing unit, by adjusting process parameters, rearranging feature processing order or optimizing toolpath (Diaz C. et al, 2020), in an a priori optimization/simulation (Mawson and Hughes, 2019) or in real-time at the shop floor level (Hu et al, 2020) . At manufacturing system level, the scope is primarily optimal resource and buffer allocation, including energy-aware job dispatching strategies (Chou et al, 2020), or energy saving policies for machines shutdown (Frigerio and Matta, 2014). However, in a typical scenario, these optimization measures are applied sequentially, without any interaction between machine and system levels.…”