Research On Optimal Scheduling Technology For Flexible Production

With the advancement of German Industry 4.0 and Made in China 2025 plan,individualization,customization and flexible production mode have become the main trend of development of manufacturing enterprises.The traditional workshop can not adapt to more flexible manufacturing mode of manufacturing industry today because the machine does not have the flexibility.In flexible production mode,the machine are allowed to process different processes in a more flexible way,which results in the flexible job shop scheduling problem(FJSP).Because the machine in FJSP with flexible features make the complexity of the problem be greatly improved which lead that it is difficult to be solved.So the solution of FJSP in intensive study has important significance both in terms of actual production application or in theoretical research.At present,FJSP is mainly solved by particle swarm optimization algorithm and ant colony algorithm,but both of them have the disadvantage of falling into the local extreme.In this thesis,in order to advance the performance of the algorithm,improvements are carried out according to the shortcomings of ant colony algorithm and particle swarm optimization algorithm.The main work and innovation are as follows:As the size of pheromone value will affect global convergence ability and convergence speed of algorithm,an update method of pheromone adaptive change is proposed,which is used to solve the FJSP and verify the performance of improved algorithm.The results of experiments show that the improved ant colony algorithm has great advancements both in global search capability and convergence speed,and the performance of two standard test cases of 6×6 FJSP with 18 processes and 10×10 FJSP with 30 processes are improved by 16%and 15.7%respectively.For the imbalance between exploration ability and development ability in basic ant colony algorithm,which leads to premature convergence of algorithm,a modified particle swarm optimization algorithm based on time-varying linearly decreasing inertia weight is proposed and used to solve FJSP.The results of experiments show that the improved algorithm can avoid falling into premature convergence,and the performance is increased by 16.5%and 16%respectively in two standard test cases of 6×6 FJSP with 18 processes and 10×10 FJSP with 30 processes.The intelligent optimization method of individual group is often poor in solving complex problems such as FJSP.In this thesis,a two-stage improved ant colony-particle swarm blend optimization algorithm is proposed to solve FJSP,in which an ant colony algorithm is applied to determine the best process route,then the particle swarm optimization algorithm is used to determine the optimal scheduling plan.The results of experiments show that the performance of improved hybrid algorithm is increased by 32%and 32.8%respectively in two standard test cases of 6×6 FJSP with 18 processes and 10×10 FJSP with 30 processes.Although the improved algorithm is validated only in two standard test cases of 6×6 FJSP with 18 processes and 10×10 FJSP with 30 processes,it also shows good performance for other FJSP problems.The above illustrates that the improved algorithm can meet intelligentized,personalized and flexible production mode of manufacturing industry today and has important application value.