A Study Of Endoplasmic Reticulum Inspired Membrane Computing Optimization Algorithms

Membrane computing (also known as P systems) is presented originally as a distribute and parallel computational model, which is abstracted from living cell structure, functioning and interrelationships in cell membrane. Inspired by the structure and function of endoplasmic reticulum, the membrane computing optimization algorithm and its applications are studied in this thesis. The main contents of this thesis are as follows:(1) Based on the basic framework of standard membrane computing, and inspired by the structure and function of endoplasmic reticulum, Endoplasmic Reticulum inspired Membrane Computing (ER-MC) was presented. Two new evolution rules (Statistics rule and combination rule) are introduced to accelerate the convergence speed and improve the convergence precision. Tested by 9 typical testing functions and compared with other algorithms show that ER-MC algorithm got higher performance, especially in the ability of against premature, convergence speed and precision.(2) An improved ER-MC algorithm with new communication mode and penalty function method was presented(ncER-MC) by improving the exchange mechanism between membranes.Compared with other algorithms proves the advantage of the ncER-MC algorithm.(3) The ER-MC algorithm is applied for parameter estimation of chemical engineering processes. Comparison was made with nmDNA-GA, the advantages of ER-MC algorithm were proved by the comparison results. The ncER-MC algorithm is applied for an three day gasoline blending model which contains price uncertainty. Compared with other algorithms proves the effectiveness of the ncER-MC algorithm.