Energy- and labor-aware flexible job shop scheduling under dynamic electricity pricing: A many-objective optimization investigation

Abstract

Energy-aware production scheduling is a promising way to adapt the factories’ energy consumption behavior to the volatile electricity prices in the demand response initiative of smart grids. However, it may not be economical by simply scheduling production loads to the periods with lower electricity prices, as these periods often have higher labor wage, e.g., nights and weekends. Based on this gap, this paper proposes a many-objective integrated energy- and labor-aware flexible job shop scheduling model. Many objectives refer to the number of optimization objectives surpasses three (i.e., five objectives: makespan, total energy cost, total labor cost, maximal workload, and total workload), whereas the existing energy-aware production scheduling research is limited within three objectives. To enable energy awareness in the conventional production scheduling algorithms, a state-based shop floor wide energy model is proposed. To enable labor awareness, the number and type of human workers are matched to the scheduled production loads, with varying labor wage over shifts. As one of the most complex shop floor configurations, the partial flexible job shop further considers job recirculation and operation sequence-dependent machine setup times. The recently-proposed nondominated sorting genetic algorithm-III (NSGA-III) is tailored for this many-objective optimization problem (MaOP), including scheduling solution encoding and decoding, crossover, mutation, and solution evaluation using the energy- and labor-aware discrete-event simulation framework. Through numerical experiments under real-time pricing (RTP) and time-of-use pricing (ToUP), insights are statistically obtained on the relation among these five production objectives; the effectiveness and efficiency of NSGA-III in solving a MaOP are also demonstrated. This proposed scheduling method can be used to automated and enhance the decision making of factory managers in jointly allocating machine, human worker, and energy resources on the shop floor, such that the production cost is minimized even under time-varying electricity and labor prices.