The Research Of BMO Hybrid Algorithm For Production Scheduling Problem

Due to its simplicity and flexibility, evolutionary algorithm in solving complicated problems has been much used.Such as solving the problem of science and engineering,and in artificial computational systems to solve complex computing problems has been the biggest success.A kind of evolutionary algorithms is a meta heuristic optimization method based on population, this method tries to mimic some mechanism of biological evolution. Although evolutionary algorithm will have differences, but all of these evolutionary algorithm to solve the problem of technology of the basic idea is the same. This paper proposes a novel evolutionary algorithm, BMO(Bird Mating Optimizer).BMO is a kind of inspired by birds during the mating season mating strategy optimization algorithm, namely the BMO and simulate the behavior of the birds breed offspring with good genes, to design the optimal search technology.For BMO is proposed a new algorithm in recent years, it is not yet mature.So this article in adjustable parameter Settings of the BMO algorithm are studied, and 23 set of benchmark functions to test it. BMO is verified by comparison with other algorithms, the algorithm has good performance, and the setting of parameters is a scientific and effective.For flow shop multi-objective scheduling optimization problems, combining the theory of genetic evolution and mutation factor analysis method, a hybrid algorithm of BMO was proposed.Genetic evolution and mutation factor was used to calculate fitness value, improving the search performance of the algorithm.The method was a collection of multiple scheduling process as a flock, by simulating the birds breeding progeny with excellent gene optimization to solve three target flow shop scheduling problem.Finally the shop scheduling test case on the MATLAB platform experiment, was able to get uniform distribution of Pareto front, get the solution of the proposed algorithm is verified better than other algorithms.For multiple flexible-shops, a new algorithm for modeling of multiple flexible-shops is proposed, SOA-BMO. Combining the SOA( Surrogate Optimization Algorithm)principles to solve the problem of discrete optimization with the BMO mating principle generated by the four strategies, SOA and BMO ensure the diversity and avoid prematureconvergence. Considered the three multi-shop scheduling problems as an example for Gantt chart, the feasibility and effectiveness of the method is verified.Results show that:this method has a great success to assign work-piece processing workshop and the reasonable planning work-piece path.