| Inspired by the membrane computing theory and evaluation communication topology, the membrane computing optimization algorithms and applications in parameter estimation and controller design are studied in this thesis. The tests and comparisons among the algorithms are made with typical testing functions. The main contents of this thesis are as follows:(1) Based the basic framework of standard membrane computing, and inspired by evaluation communication topology like the structure of computer network, a star topology-membrane computing (ST-MC) algorithm is presented. The algorithm adopts three sub-systems in parallel, using evolution rules of rewriting, crossover, inversion and communication. In order to compare with other methods, seven testing functions with larger searching space, more local minimum points and better deceptive characters are adopted to build the testing environment. The experimental results demonstrate the effectiveness of the proposed method, especially in the advantages of searching ability and precise, stability.(2) Considering the uncertainty in the evolution rule's operation probability in ST-MC and combining the fuzzy logic with the ST-MC algorithm, the fuzzy star topology-membrane computing algorithm (FST-MC) is proposed. This algorithm builds a double-in-single-out fuzzy controller to evaluate the inversion probability fussily. With increasing evolution generations, the inversion probability is updating in self-adaptive. The advantages of algorithm are proved by the test result from typical test functions.(3) The ST-MC algorithm is applied for parameters estimation of chemical engineering processes, such as the Enzyme effusion and the heavy oil thermal cracking. The superiority of ST-MC is shown in comparison with other methods and SGA. The FST-MC algorithm is adopted to solve the optimization problem of the controller design of a marine diesel engine. The desired parameters are obtained in order to keep the controller stability. |
PDF Full Text Request