Researches On The Simulation Of Membrane Computing And Membrane Evolutionary Algorithm

Membrane computing(also known as P system)is a distributed parallel computing model inspired by the structure and functionning of biological cells,proposed by Professor P?un,an academician of the European Academy of Sciences and an academician of the Romanian Academy of Sciences.It has gained more and more attention of researchers because of its strong computing ability.In view of the limitations of the biological implementation of the P system,P system simulators have become an important tool for the verification and research of P systems.The P system simulation tool like P-Lingua cannot adapt to the rapid development of membrane computing due to their poor scalability and unmanned maintenance and update.New features and hybrid computing models are emerging in the field of membrane computing.This thesis has carried out related research from the expansibility and maintainability of simulator.In the research of the model's application ability,researchers have proposed application solutions including numerical computing,solving computationally hard problems(SAT problem,0-1 knapsack problem,etc.),and membrane algorithm.Among them,membrane algorithm is a heuristic algorithm inspired by the idea of membrane computing.However,in previous studies,many other heuristic algorithms such as genetic algorithms and local search algorithms were integrated with membrane algorithm in the application of membrane algorithm,which means the evolving ability of membrane algorithm relays on other heuristic algorithms.From the point of evolution ability,this thesis studied membrane evolutionary algorithm and applied it to solve minimum vertex cover problem.Based on previous researches of exsiting P system models,this thesis designed and implemented a new P system simulation tool,proposed and implemented a new kind of membrane algorithm called membrane evolutionary algorithm.The main contributions are listed below:(1)Designed and implemented a new P system simulation tool called UPSimulator.By simulating mechanisms of P systems,UPSimulator can simulate three basic types of P system and some newly proposed P systems.Moreover,the supported concepts can be combined together freely that means with the suppport of more and more new concepts,the number of UPSimulator suported P systems will increase exponentially,which can somehow solve the diffculties in the simulation of hybrid P systems.(2)Proposed membrane computing idea inspired membrane evolutionary algorithm framework(MEAF).MEAF has its own operators: Division,Fusion,Selection,and Cytolysis.Based on MEAF,this thesis implemented the four operators to solve the minimum vertex cover problem.The experiment results are better and more stable than local search algorithm Fast VC.Besides,this thesis also improved the local search algorithm for minimum vertex cover problem,and got better results and performance.The designed and proposed P system simulator UPSimulator will make the simulation process of P systems easier for P system researchers,and also will decrease the diffculties in simulating new P system models.The proposed membrane evolutionary algorithm has four unique operators,which broke the situation that membrane algorithms must integrate some other heuristic algorithms.