Research And Application Of Multi-objective Lion Swarm Optimization Algorithm

For a long time,optimization technology,which is very important in engineering,production and society,has continuously attracted researchers to improve and explore.As an important tool in the field of optimization,swarm intelligence optimization was first inspired by the biological world and then developed.As a new algorithm proposed in recent years,lion swarm optimization has excellent mechanism design and great development potential,which is worthy of in-depth study.The lion swarm optimization simulates the evolution and hunting process of the lion group.It divides the swarm into the lion king,female lions,and young lions.They have their own tasks and cooperate with each other to adapt to different scenarios.The multi-model and diversity of the algorithm's behavior mechanism enables it to search and optimize more comprehensive questions,and reduce the probability of falling into a local extreme value.So it can adapt better to different situations.There are information exchange mechanisms between different kinds of lions,which is conducive to the preservation and spread of outstanding characteristics within the population.At the same time,the algorithm can guarantee a certain convergence speed.Aiming at the problem that the lion swarm optimization has very little research work on the multi-objective problem and the performance is not ideal,the thesis proposed a new multi-objective lion swarm optimization.It is based on the theory of multi-objective optimization.It refers to several classic multi-objective optimization algorithms,and introduces their reasonable mechanism,including fast Pareto sorting,crowded distance and external files.Then,in order to improve the performance of the new algorithm,the thesis further studied and proposed two improved algorithms.Aiming at the problem that the multi-objective lion swarm optimization tends to be concentrated locally and the search ability is not good when the dimension is high,the thesis refers to the resampling mechanism in sequential Monte Carlo sampling,and introduces the resampling method in the algorithm.It can adjust the spatial density of the population to avoid the problem of premature convergence and unable to search the whole Pareto front.And then the improved algorithm is applied to the multi-objective optimization model of multi-UAV task assignment which is constructed in this thesis.Through the simulation experiment,it can be determined that the multi-objective lion swarm optimization can successfully solve the problem,provide us with a set of Pareto solutions.After that,the thesis refers to cloud model theory and cloud generator algorithm,and proposes a multi-objective lion swarm optimization based on cloud model mutation.It introduces a cloud model mutation mechanism for individuals,improves the diversity of the population,helps balance the global search ability and local search ability.It can help the algorithm jump out of the local extreme value,and avoid the population from falling into the local optimum and local over-concentration.It increases the range of the population search and enhances the optimization ability of the algorithm.After comparative experiments,the improved algorithm has more obvious advantages in the high-dimensional multi-objective functions.Similarly,the thesis establishes a multi-objective optimization model for the job shop scheduling problem,and optimizes it with the improved lions swarm optimization.And we can obtain satisfactory results,which proves the algorithm's ability to solve multi-objective problems.Finally,the thesis summarized the progress of the current work,and pointed out the shortcomings and future research directions.