On The Improvement And Application Of Co-evolutionary Multi-objective Optimization Algorithm

In the process of scientific research and engineering practice, there are a number of Multi-Objective Optimization problems(MOP),Evolutionary Algorithm has been proved to be an effective method for solving these problems.However, Multi-Objective Evolutionary Algorithm has its own issues,such as premature,slow convergence, poor distribution and uniformity.In fact,Co-evolutionary Multi-Objective Evolutionary Algorithm is proposed in recent years, which is based on one or more population that simultaneously evolved.Compared with the traditional Multi-Objective Evolutionary Algorithm, to some extent, it can avoid premature and improve local convergence. But the basic Co-evolutionary Multi-Objective Evolutionary Algorithm doesn’t significantly promote global convergence of Multi-objective Evolutionary Algorithm, the distribution and uniformity of its solution set may not be satisfactory, either.In this paper, the basic Co-evolutionary Multi-objective Evolutionary Algorithm will be researched, and then apply the improved algorithm to Job-shop scheduling problems.In this way,the specific tasks are discussed as follows:1.Make a brief introduction of the research process and status of Multi-objective Evolutionary Algorithm,Co-evolutionary Multi-objective Evolutionary Algorithm. Besides, their basic principles,algorithms and processes will also be stated in the following chapters.2.Then,present Co-evolutionary Multi-objective Algorithm based on aggregation density:Firstly, calculate the aggregation density of each individual in population and define a poset. After that, select individual updates P from the poset based on a certain ratio.According to numerical experiments and quantitative measurements, this improved algorithm could make the distribution and uniformity of its solution better.3.Put forward an Adaptive Co-evolutionary Multi-objective Evolutionary Algorithm, which is:the co-evolution operators are automatically called according to the change rate of objective function;When the population evolution is normal,cooperation operator and merging operator are called, and otherwise division operator is called.Finally,it is turned out that the global convergence has been significantly improved by means of new algorithm.4.Apply the improved Co-evolutionary Multi-objective Evolutionary Algorithm to solving Job-shop problems, then the performance and practicality of improved algorithm will be tested in simulation experiments.