The Research Of Multiobjective Immune Algorithm And Its Application On The Flexible Shop Scheduling

In the fields of engineering application and scientific research,there exist a lot of complex optimization problems.Because of their complexity,dynamic and difficulty to modeling,conventional operational research methods cannot solve them well.However,heuristic-based intelligence algorithm has advantages of solving these optimization problems.Among them,the Artificial Immune System is a new intelligent system developed by simulating the information processing principle and mechanism of biological immune system,which provides the evolutional learning mechanism such as noise tolerance,learning without supervisors,self-organization and memory.It can provide a novel method and mind for solving complex problems.Thus,artificial immune algorithm has been widely attention by the scholars in many fields.This paper studies immune algorithm mainly from the view of optimization problems.Firstly,introducing the present situation of the immune algorithm,the biological basis theory and its algorithm principle,the present situation of flexible job shop scheduling and its problem description.Then,analyzing multi-objective optimization problems and put forward the improved immune algorithm.Finally,verifying its performance on 21 test problems.After that,improving the proposed algorithm according to the characteristic of flexible job shop scheduling problems to further verify that the proposed algorithm also has better performance in practical application.The main work of this paper is described as follows:(1)When solving multi-objective optimization problems,this paper proposes a dynamic population strategy immune algorithm(MOIA-DPS).The main point is proposed a dynamic population strategy(DPS),which can control the population size according to the status of external archive.Therefore,it can reasonable use of computing resources during the evolutionary progress,avoid premature convergence and increase the population diversity.In addition,an effective DE operator is designed to enhance the robustness of MOIA-DPS,which combines the advantages of rand/2/bin and rand/1/bin.Then,the simulation experiments are proceeded on 21 test issue.MOIA-DPS compares to five classic algorithms and five recently proposed immune algorithms,meanwhile verifies the effectivenss of DPS and TDE operators.The experimental results show that the proposed algorithm MOIA-DPS has obvious advantages in the multi-objective optimization problem.(2)When solving flexible job shop problems,this paper proposes a dynamic clone population strategy immune algorithm(DCPS-MOIA).DCPS-MOIA proposes a dynamic clone population strategy.When the improvement rate of the whole population is less than the set value,the clone population size is increased to enhance the genetic diversity.Using this method to balance the population diversity and convergence.Meanwhile,using TDE which proposed in Section 3 to enhance the local search ability and population diversity.Finally verifing the effectiveness on three problem instances.