Discrete Teaching And Learning Optimization Algorithm To Solve Complex Shop Scheduling Problems

Production scheduling is an important part in modern manufacturing systems.Effective scheduling optimization algorithms are an important way to improve enterprise efficiency and competitiveness,and belong to an important research field of intelligent manufacturing.Parallel machine scheduling problem and permutation flow shop scheduling problem are two kinds of combinatorial optimization problems widely existed in manufacturing.In terms of computational complexity,they are belong to NP-Hard problem.Therefore,it is of great engineering significance and theoretical value to study effective algorithms for solving uncorrelated parallel machine scheduling problems and permutation flow shop scheduling problemsBased on the standard teaching-learning-based optimization algorithm,this paper proposes three improved discrete teaching and learning algorithms,and applies it to solve complex parallel machine scheduling problems and permutation pipeline scheduling problems.Firstly,an improved discrete teaching and learning optimization algorithm is proposed for a class of multi-process parallel machine scheduling problems with optimization constraints to minimize the maximum completion time.According to the characteristics of the two-stage individual updating formula in the standard teaching and learning algorithm,under the premise of retaining the individual updating formula framework in each stage,the designed permutation is used for each core operation of the real individual or vector in the formula.The operation is replaced so that population updates based on standard teaching and learning algorithm can be performed directly in the discrete solution space.Secondly,considering the sequence dependent setup time between workpieces,a class of complex parallel machine scheduling problems with multiple constraints is obtained.A hybrid discrete teaching and learning optimization algorithm is proposed to solve the problem.In the discrete teaching stage,the self-learning process is introduced.The discrete learning stage introduces the learning probability to avoid the blind trust of the students.Combined with the local search based on Interchange and Insert neighborhood,the search depth and overall performance of the algorithm are enhanced.Finally,in order to solve the problem of permutation pipeline scheduling commonly used in manufacturing production process,considering the optimization goal to minimize the maximum completion time,a hybrid discrete teaching and learning optimization algorithm based on probability model is proposed In order to improve the quality of the initial population,we introduce the NEH rule and FBR1 rule in the initialization.In the discrete stage,the probabilistic model updating mechanism of the two-dimensional probability matrix is introduced,which enables class students to obtain better knowledge from high-quality students,and designs local search based on Insert and Inverse neighborhoods to enhance the breadth and depth of algorithm search.By comparing the simulation experiments and algorithms of different test problem sets,it is verified that the proposed three algorithms can effectively solve the corresponding problems,and compared with the algorithms in other literatures,the proposed algorithms show high efficiency and robustness.