The Application Of Improved Genetic Algorithm In Clonal Arrangement Design For Seed Orchard

Forest genetic breeding is the basis and prerequisite for the development of the forestry.The construction of seed orchard,which aims for producing tree species,is the key issue of forest breeding.Today,the research on seed orchard construction is mainly focused on the high-generation seed orchard.However,due to the sudden drop in the number of parental species,which belong to the high-generation seed orchard,compared to the parental species' number in the initial seed orchard,the probability of inbreeding between parents is greatly increased,and the genetic gain of the target traits cannot be improved effectively.Based on the genetic relationship between parents,this paper uses optimization algorithm to optimize the design of clonal arrangement for seed orchard,aiming to reduce the chance of inbreeding and enlarging the genetic basis of the offspring.Genetic algorithm has been widely applied in various fields.And,there are many methods to improve genetic algorithm for fitting in different areas.The problem of optimizing arrangement design for seed orchard based on genetic relationship,is an NP-hard problem.This paper uses genetic algorithm to optimize the seed orchard clonal arrangement design and combines with domain issues to improve the genetic algorithm.Firstly,the IAPGA(Improved Adaptive Parallel Genetic Algorithm)is given.The points of improvement are as follows:adding the greedy algorithm to species initialization,making the probability of crossover and mutation be adjusted automatically according to species fitness,and dividing the initial species into multiple populations.Secondly,by revising the migration strategy between subpopulations,the New-IAPGA is proposed.Finally,applying both New-IAPGA and I[APGA to the seed orchard clonal arrangement design to carry out the comparative analysis.Besides,the New-IAPGA is applied to the design of seed orchard with flowering period and genetic distance.The seed orchard design created by IAPGA and New-IAPGA is better than the traditional sequential dislocation method and other methods.Moreover,the New-IAPGA is more stable than the IAPGA and has searched the global optimal fitness for many times.In summary,using IAPGA or New-IAPGA to design seed orchard can maximize physical distance between the clones which have closer genetic distance,reduce the possibility of inbreeding and maintain the offspring's genetic diversity.In the end,it provides advices for the clonal distribution of the high-generation seed orchard.