Research On Energy-saving Flexible Job Shop Scheduling Base On Swarm Intelligence Algorithms

Due to exacerbated environmental pollution in China,the central and local governments are paying more and more attention to the environmental protection,in order to effectively meet the situation.Meanwhile,enterprises should adopt more flexible and reasonable production plans and workshop scheduling schemes.At the same time,in the process of scheduling management,the environmental protection and energy consumption index requirements are needed to be seriously considered.In the job shop system,it can make the product machining path become flexible by means of equipment adjustment,and effectively improve the efficiency and flexibility of shop scheduling.However,the addition of flexible machining path increases the complexity of the workshop scheduling problem.Previous studies have proved that the flexible job-shop scheduling problem is a combinatorial optimization problem with the NP-hard characteristics.However,in today's severe situation of environmental protection,adding optimization indexes such as energy consumption and carbon emission in the actual workshop scheduling process can effectively reduce environmental pollution while improving production efficiency and reducing manufacturing cost,resulting in the benefits of providing the enterprises with a green and sustainable production mode.Therefore,it is of great theoretical and practical significance to construct a reasonable energy-saving scheduling methodology based on theoretic breakthroughs and practical applications of flexible job shop and state-of-the-art scheduling optimization algorithms.In this paper,according to the production environment characteristics of different workshops,different types of energy-saving scheduling models are studied and constructed,and efficient solution approaches are designed.The following research tasks have been completed:Firstly,the paper studies the single-objective energy-saving flexible job shop scheduling problem.Then,a problem model is developed to minimize the sum of processing cost and energy consumption cost as the objective function.In the established problem model,the energy consumption cost of equipment includes two parts: processing energy consumption and no-load energy consumption costs.In order to solve the problem effectively,an improved whale optimization algorithm is designed with the four innovative aspects.A Rank-Order-Value(ROV)transformation mechanism based on random keys is introduced to realize the transformation between the FJSP problem solution and the individual position vector of whales.A hybrid population initialization method is developed on the basis of a certain proportion of global,local and random search,so as to improve the quality of the initial population.The nonlinear convergence factor and the adaptive inertia weight coefficient are designed to enhance the ability of the algorithm to coordinate global search and local optimization.The adaptive search strategy is introduced to prevent the algorithm to fall into the local optimum.The correctness of the proposed model and the effectiveness of the proposed algorithm are verified by simulation experiments using random and standard instances.Secondly,we then propose a discrete whale optimization algorithm to solve the single-objective energy-saving flexible job shop scheduling problem.In the solving process of the improved whale optimization algorithm,due to the introduction of conversion mechanism,its solving speed is slow.Thus,a new metaheuristic algorithm called the discrete whale optimization algorithm is designed by referring to the idea of whale optimization algorithm.In the proposed algorithm,three different crossover operations have been designed to replace the three original iterative methods in the traditional whale optimization algorithm as "Shrinking Encircling Mechanism","Spiral Updating Position",and "Search for Prey",which make the algorithm procedure efficient to solve the single-objective energy-saving flexible job shop scheduling problem.In order to effectively balance the global search ability and local optimization ability of the algorithm,six different dynamic adjustment curves are used to improve the convergence factor a,and the corresponding six improved algorithms are developed.Through the simulation experiments of 19 instances consisting of 14 standard instances and 5 random instances,it is verified that the proposed discrete whale optimization algorithm has a satisfactory performance in solving the single-objective energy-saving flexible job shop scheduling problem.Thirdly,we investigate the low-dimensional multi-objective energy-saving flexible job shop scheduling problem.Considering the fact that the single-objective energy-saving flexible job shop scheduling problem contains only one criterion of minimizing the cost,we extend to construct a low-dimensional multi-objective energy-saving flexible job shop scheduling problem model with the criteria of minimizing the makespan,the total tardiness and the total energy consumption of equipment.In the model,the speed of all machines is constant Then we propose an improved NSGA-? based on sparseness theory.The improved NSGA-?algorithm can obtain the children population of local search by performing different neighborhood search operation on the sparse solution,which forms a new population with original populations.In this case,the diversity of the population can be enhanced and the accuracy of the obtained solutions can be improved.The weighted method is used to select the optimal compromise solution from the obtained Pareto solution set.The advantages of the proposed algorithm are verified by a comparative simulation of an enterprise case and several standard instances.Finally,we examine the many-objective energy-saving flexible job shop scheduling problem.It is observed that low-dimensional multi-objective energy-saving flexible job shop scheduling problem fails to consider the influence of the machine speed on scheduling results.To overcome this limitation,we establish a many-objective energy-saving flexible job shop scheduling problem model with the objective of minimizing the maximum completion time,the total tardiness,the total load of equipment and the total energy consumption of equipment.In this new model,every machine is supposed to have multiple processing speeds to select.As the improved NSGA-? is not suitable for solving many-objective optimization problem,we develop an improved NSGA-? that has more stronger search ability to solve the model.By analyzing the characteristics of the many-objective energy-saving flexible job shop scheduling problem,four different combinations of crossover and mutation strategies are applied to improve the traditional NSGA-?,and four improved NSGA-? algorithms are proposed.The performance of four improved strategies and the superiority of the improved NSGA-III algorithm are verified in comparison to classical algorithms based on extensive computational experiments.