Study On Flexible Job Shop Scheduling Problem Based On Hybrid Quantum Algorithm

With the rapid development of world markets and production, the global competition of manufacturing, which is an important basic industry of national, is intensifying. Since the 1990s, the domestic manufacturing industry has gradually become the main pillar of national economy development. It developed rapidly but also faced severe tests. Under the global market competition, how to improve the competitiveness of the enterprise through the low cost, high quality, fast speed and satisfied services? How to improve the product management mechanism to satisfy modern production's requirement of small product demand, complex consumption structure and short-time order delivery? How to find the strategy to shorten the distance of theoretical research and practical production? Manufacturing enterprises need to focus on these core problems with improving the management. They are the key of enterprise competition.The efficiency of job-shop scheduling may influnce the effectiveness of the manufacturing. The effective control of job-shop scheduling cost is directly related to the development of manufacturing industry. So, the research of flexible job-shop scheduling problem, which is the key link of job-shop scheduling, is given more and more attention. Scholars made lots of researches on the flexible job-shop scheduling problem in modeling, choosing optimal solution and application of algorithm. But the theoretical research focus on certain and static problem. It would not consistent with the actual production scheduling. When the problem size is bigger, current method can't adapt to the change of state space of problem. It leads to decline of solving efficiency. At present, the evaluation standard of algorithm for solving the flexible job-shop scheduling problem is not unified. That may make the evaluation of algorithm not objective. In view of this, this paper researches on the flexible job-shop scheduling problem, this is consistent with practical producing scheduling, and its performance evaluation. It has theoretical and practical value. This paper mainly researched on general multi-objective FJSP, fuzzy FJSP and dynamic FJSP. And then the paper used the classic test set and actual production job-shop scheduling data to verify the validation of proposed method.Firstly, in the actual job-shop scheduling, multiple objective index may have conflict. The thesis sets up the model of minimizing delivery penalty value, makespan, processing machine load, producing cost and maximizing customers'satisfaction.; sets different target weight coefficient; and introduces the crowding distance selection strategy to optimize the non-dominated solutions.Secondly, in multiobjective FJSP, completion time and date of delivery may be fuzzy. The thesis sets up the fuzzy mathematical model, proposes chaotic local optimization strategy and improved quantum rotation angle CQPSO algorithm, solve the convergence problem caused by loss of population diversity.Thirdly, in flexible job-shop production process, some dynamic interference factors may occur, such as, emergency artifacts insert and machine breakdown. The paper defines and classifies the dynamic aspects of problem, designs rolling window strategy based on rescheduling cycle driven and rescheduling event driven. At the same time, the thesis uses double-stranded coding method with two genes chains represent machine distribution and process sequence. In order to balance scheduling efficiency and stability, the thesis introduces the mixed multiphase quantum particle swarm to solve problems.At last, the thesis uses the standard Kacem test set, Brandimarte test and the actual data of workshop, analyses the existing problems in the production process, designs FJSP scheduling prototype system and study how to apply this method to the production practice.