Research On Multi-objective Algorithm And Its Application In Flow Shop Scheduling

Genetic algorithm is a kind of bionic optimization algorithm,which comes from the process of reproduction and evolution of creatures in nature.In the genetic algorithm,the individuals in the population are copied,crossed,mutated and selected to produce the offspring individuals,and then the population is updated to select the excellent individuals.The concept of genetic algorithm is characterized by simple concept,less parameter adjustment and easy operation.Based on the research of genetic algorithm,this paper studies the multi-objective optimization algorithm.This paper mainly studies the multi-objective optimization algorithm based on decomposition,and proposes an improved multi-objective genetic algorithm to solve the problem of high computational cost.In addition,the proposed multi-objective genetic algorithm is used to solve the production scheduling problem.Firstly,the genetic algorithm is studied deeply.The bionics principle,main idea and algorithm description of genetic algorithm are described,and the algorithm flow chart is shown.Secondly,the multi-objective optimization algorithm is studied.Aiming at the high computational cost,a multi-objective genetic algorithm based on Agent Assistant is proposed.In the algorithm,the fitness estimation strategy is used to predict the objective function value of new individuals by using the objective function values of neighboring individuals and parent individuals.Meanwhile,radial basis function network is adopted to predict the objective function value of the new species group through trained population network.When an individual enters a renewal population,the real objective function value must be calculated to ensure that all individuals in the population have the real objective function value.The above two agent assistant strategies are used to reduce the calculation times of the real objective function value of the individual,so as to reduce the calculation cost of the algorithm.In order to prove the effectiveness of the proposed algorithm,ZDT and DTLZ test function series are used to compare the proposed multi-objective genetic algorithm with other algorithms.The simulation results of test function show that the algorithm can reduce the computational cost and obtain a Pareto front with good convergence and distribution.Finally,the multi-objective genetic algorithm is applied to the production scheduling problem.This paper describes the flow shop scheduling problem,establishes a two-objective model and determines the objective function.The simulation results show that the proposed multi-objective genetic algorithm can effectively solve the flow shop scheduling problem.