Predictive maintenance integrated production scheduling by applying deep generative prognostics models: approach, formulation and solution

Abstract To harness the full potential of predictive maintenance (PdM), PdM information has to be used optimally plan production and actions. Hence, operation-specific modelling degradation, i.e. predictions health condition under time-varying operational conditions, realized. By utilizing degradation information, can planned with regard each other thus, integrated scheduling (PdM-IPS) is enabled. This publication proposes a novel PdM-IPS approach consisting two interacting modules: an Prognostics Health Management (PHM) module planning (IPSMP) module. Specifically, mathematical problem IPSMP based on extended version flexible job shop formulated. A two-stage genetic algorithm efficiently solve this designed subsequently applied simulated monitoring, as well real industrial data. Results indicate that able find feasible high quality schedules.