Font Size: a A A

Research And Application Of Evolutionary Algorithm Based On The Tissue-like P System Of Trigger Ablation Rule

Posted on:2022-03-12Degree:MasterType:Thesis
Country:ChinaCandidate:X SangFull Text:PDF
GTID:2480306332487864Subject:Management Science and Engineering
Abstract/Summary:PDF Full Text Request
Membrane computing,also known as P system,is a distributed computing model with extremely parallel characteristics,and it is obtained through the evolution of genetic material in cells and the communication between cells.The P system is mainly divided into three types:cell-like P system,tissue-like P system,and neural-like P system.Among them,the tissue-like P system is a model based on the cell-like P system,which abstracted from cooperation that realizes the exchange,communication and evolution of genetic material.The Biogeography-based Optimization(BBO)is an evolutionary algorithm,inspired by biogeography,to complete the evolution of species through information exchange in the form of migration,mutation and extinction of species between different habitats,so that the optimal solution to the practical application problem corresponding to it can be found.As a widely used algorithm,there are still problems such as poor diversity and easy to fall into local optima.Differential Evolution algorithm(DE)is an evolutionary algorithm based on population,which realizes its own evolution through cooperation and competition between groups.It has the characteristics of fast convergence speed and better search ability,but for complex optimization problems,it is easy to fall into the local optimum,and the parameter settings are relatively fixed,different optimization performances are often shown for different problems.In order to make full use of the characteristic and computing power of the tissue-like P system,and better solve some of the shortcomings of the two evolutionary algorithms,this paper starts from three aspects to study include the tissue-like P system of trigger ablation rule,improved BBO algorithm and improved DE algorithm,and the main work is as follows:First of all,this paper proposes a new type of tissue-like P system of trigger ablation rule,which sets two forms of rules for triggering ablation when other objects in the cell other than themselves meet certain number or state conditions.The characteristics of the ablation of genetic material and the parallelism of the execution of the rules in the cell are proved and calculation ability of the new tissue-like P system is proved.Secondly,this paper proposes an improved BBO algorithm based on the tissue-like P system of trigger ablation rule(DCGBBO-TATP).The migration operator is dynamically adjusted,and the characteristics of the migration into and out of the habitat are obtained and dynamically integrated,the Gaussian distribution function is used to divide the mutation process into two stages for dynamic control,and the reverse learning mechanism is used to make the algorithm jump out of the current local optimum,and test it on 8 classic test functions which obtain good optimization results.The improved algorithm with the proposed new system are combined,the trigger ablation rule is used to realize the operation of the algorithm.Then,this paper proposes an improved DE algorithm based on the tissue-like P system of trigger ablation rule(ACMDE-TATP).The difference vector of the mutation operator is adjusted,and a dynamic adaptive adjustment to the crossover operator is made,dynamically change the crossover probability with the operation of the algorithm,and then jump out of the local optimum,and test on 6 classic benchmark functions which achieved good optimization performance.The improved algorithm with the proposed new system are combined,and the operation of the algorithm by trigger ablation rule is realized.Finally,this article applies the evolutionary algorithm based on the tissue-like P system of trigger ablation rule in image segmentation.The DCGBBO-TATP algorithm is applied to color image segmentation,and the ACMDE-TATP algorithm is used to solve the gray-scale image segmentation problem.Different evaluation indicators prove that two kinds of algorithms have achieved better segmentation results.In summary,this paper proposes a new type of tissue P system for triggering ablation rules.Under its framework,combined with the concept of triggering ablation rules,the improved BBO algorithm and the improved DE algorithm are respectively proposed,and the optimized evolutionary algorithm is proposed.The model is applied to image segmentation to improve the effect of segmentation.
Keywords/Search Tags:Tissue-like P system, Biogeography-based Optimization, Differential Evolution Algorithm, Image Segmentation
PDF Full Text Request
Related items