| At present,many achievements have been made in forestry work of China,and the seed orchard construction is also improving and developing towards higher generations.Recognizing the species and quantity of trees in the seed orchard will inevitably decrease,and will eventually face inbreeding decline.This paper designed the clonal seed orchard based on the clonal genetic distance of Pinussylvestris var.Mongolica and the simulated flowering and pollen quantity of the parents.The fruit fly optimization algorithm,formulated originally as global optimization algorithm,is expanded into realistic problems of different fields.This paper uses fruit fly optimization algorithm to optimize clonal seed orchard design.First of all,the SIFOA(Step-modified improved fruit fly optimization algorithm)is given.The results show that the maximum value of SIFOA is lower than the minimum value of other algorithms after executing 200 times respectively.In the final seed orchard design scheme,the completely random design scheme has multiple clones of the same clone as neighbors,but SIFOA does not have this phenomenon.The proportion of the two clones(Clone 3 and Clone 7)with the largest genetic distance in the experimental data is higher than that of the genetic algorithm,which shows that SIFOA is more conducive to the hybridization of clones.Then,this paper designs a two-population improved adaptive step-length fruit fly optimization algorithm(TIASFOA).Then SIFOA and TIASFOA are applied to the design of clonal seed orchard that integrated flowering and pollen amount.The consequences show that TIASFOA can keep the same flowering time and sufficient pollen amount among the clones.The fitness values of mean,maximum,minimum and variance of the two algorithms are smaller than that of SIFOA.When the population size is small,TISFOA has ob vious stability advantages. |
