Study On Flexible Flow Shop Scheduling For The Carbon Efficiency Optimization

Carbon emissions are the root of the deteriorating environment and foggy,and carbon emissions from the manufacturing process account for half of the global carbon emissions.As an advanced manufacturing process,the flexible flow shop(FFS)can meet the requirements of multi-variety and small batch production,mass customization and single-piece production.Moreover,the study on the carbon efficiency optimization of FFS is an important basis to implement the flexible low carbon manufacturing,and it helps to achieve the strategic layout for China to become powerful in manufacturing.This paper focuses on the carbon efficiency optimization of FFS,and attempts to optimize the carbon efficiency,maximum completion time and maximum machine load simultaneously.Although phased and hierarchical decision making approach can integrate the optimization of processing parameters into the scheduling program to obtain better results than those of unilateral decision making approach,it fails to achieve the simultaneous optimization of processing parameters and scheduling scheme,which can cause information isolated islands and lower optimization performance.Hence,this paper proposes the carbon-efficient scheduling where the processing parameters and the scheduling scheme are optimized simultaneously,and constructs the mathematical model of the synchronization decision of the processing parameters and the scheduling scheme based on the unit-specific event-based continuous-time formulation.Subsequently,this paper designs the decoding method for synchronization decision based on the Monte Carlo simulation and heuristic rules,and develops an integrated weighted target evolution algorithm and a multi-objective evolutionary algorithm.The simultaneous optimization method of three objectives under the synchronization decision of the processing parameters and the scheduling scheme can provide multiple alternative Pareto non-dominated scheduling schemes,which can be selected by the manager.First,the calculation formulas of carbon emission and turning time are provided after analyzing the turning process.Based on the above research,these objective functions of the carbon emission,maximum completion time and maximum machine load are establishedwith the feed rate,spindle speed and scheduling decision as independent variables.Andcombined with the material and machine constraints a 0-1 mixed integer programming model of FFS carbon-efficient scheduling is construct based on the unit-specific event-based continuous-time approach afterwards.Then the double-deck decoding method is designed for the carbon-efficient scheduling of FFS based on Monte Carlo simulation and heuristic rules.In the first deck,the processing order of the jobs is determined according to their available time,and in the second deck,the machine allocation and processing parameters are synchronously determined based on the Monte Carlo simulation.After that,the improved genetic algorithm(GA),estimation of distribution algorithm(EDA)and gravitational search algorithm(GSA)are designed according to the flow of algorithms.Specifically,the improved GA is acquired combined with three improved operations-initialization,neighborhood search and restart,and the improved EDA is also obtained by combining the dynamic probability and the three operations.Furthermore,the experimental results verify the effectiveness of the three proposed algorithms.Next,the nature characteristics of non-dominated solution are studied,and later a method of quickly constructing non-dominated solution is designed,which is faster than the electoral method and exclusion method.Based on the above method,Monte Carlo simulation and heuristic rules,the decoding method of multi-objective EDA is developed for the carbon efficiency optimization of FFS,and the decoding method is decomposed into two layers to achieve lower complexity.Then based on the reference point setting method and decoding method,the optimal population selection of EDA is designed and the multi-objective EDA for carbon-efficient scheduling of FFS is developed.Subsequently,an improved multi-objective EDA is developed based on the neighborhood search of NEH and the restart operation of the external archive set.After the evaluation indicators of multi-objective algorithm are systematically studied to reveal the differences among convergence indicator,distribution indicator and ductility indicator,the calculation formulas of these indicators are reconstructed according to the three goals and they are utilized to evaluate the performance of the original multi-objective EDA and improved multi-objective EDA algorithm.Finally,on the basis of the above research results and the detailed analysis of turning process and energy consumption in a roll enterprise machining workshop,this paper develops the carbon effeciency optimization system of roll turning and this system is tested by using the actual processing data of the roll processing workshop.And the test results verify the optimizing effect of the synchronous decision of the processing parameters and the scheduling decision is much better than the separate decision,futhermore,they also demonstrate the effectiveness of thesedesigned algorithms.