Evolutionary design of a simple membrane system

Abstract

The programmability of membrane systems is an ongoing and challenging issue. This paper focuses on the automatic design of a simple membrane system for fulfilling a specific task by using a quantum-inspired evolutionary algorithm and the P-lingua simulator. The design consists of the pre-defined membrane structure and initial objects, a set of possible evolution rules, the coding technique of membrane systems, evolutionary operators and a fitness function for evaluating different membrane systems. Experiments conducted on P-lingua simulator show that the presented design approach is feasible and effective to automatically evolve a membrane system for solving some specific tasks. The results also show that a quantum-inspired evolutionary algorithm is more appropriate than a genetic algorithm, recently reported in the literature, for designing a membrane system. © 2012 Springer-Verlag.