Research On Welding Shop Scheduling Models And Methods For Large Components

As a very important assembly process the welding process is widespread in the modern manufacturing industry, such as aerospace, shipbuilding, automotive, engineering machinery, etc., and the energy consumption is very high. Therefore, the welding shop scheduling technology has great influences on the performance of whole production system. However, welding shop is quite different from the machinery shop, such as multi-machines can process one job at the same time in welding shop and the energy consumption in welding shop is quite different from machinery process. Therefore, the previous scheduling models, methods and energy consumption model based on machine work cannot be applied in WSSP, and there were very few researches on the welding shop scheduling problem(WSSP).In this paper, a single objective model and an energy-aware multi-objective model for the large component welding shop is formulated, then a discrete artificial bee colony algorithm(DABC) and a Multi-objective ABC(MOABC) are proposed respectively to solve the problems. Finally, a case study considering a real-world welding shop scheduling problem is conducted to validate the proposed methods.Firstly, a special hybrid flow shop scheduling problem where multi-machines can process one job at the same time which is quite different from the classical scheduling problem is proposed and defined. And a mixed integer programming model for the WSSP is formulated based on the permutation flow shop scheduling problem with setup time.Secondly, based on the single objective model of WSSP, an effective DABC has been proposed to solve the WSSP, considering job permutation and machine allocation at the same time. To improve the performance of proposed DABC algorithm, effective operators have been designed. To evaluate the effectiveness of proposed method, 3 instances with different scales have been used and the experimental results show that the proposed DABC is superior to GA.Then, the energy consumption in WSSP is investigated and an energy aware multi-objective model is developed. A MOABC is designed to solve the problem and numerical experiments show that MOABC is superior to SPEA2.Finally, a case study from a practical girder welding shop is conducted. The results show that the single objective and multi-objective methods reduce 55% and 30% production time respectively comparing with the traditional method and the scheduled machine allocation provides more reasonable arrangements for workers and machine loads.