DNA Genetic Algorithm And Its Application In Process Industrial

DNA computing is a novel research field in recent years. Its basic principle is to utilize the double helix structure of DNA molecular and the principle of base pairing to code the problem that mapped to a specific DNA fragment, then generate the data pool by the controllable of biological enzyme, and finally get the possible solution of the problem. Genetic algorithm is a stochastic global optimization based on Darwin’s evolution theory and Mendel’s natural genetic variation theory. It does not require the continuity, conductivity and others stringent assumptions of the actual problems; it is not based on specific areas, and can provide a universal framework for solving complex systems. However, some disadvantages of GAs such as the low search efficiency, the tendency to premature convergence, and the fact that the parameter range of initial solution are relying on the set of experience point and so on.DNA genetic algorithm is based on GA and DNA computing. DNA-GA outperforms tradition GA with the help of DNA coding and operators, whose technique can effectively overcome binary hamming cliffs problem, and premature convergence, as well as enhance the diversity of the population, update and improve the algorithm. Based on the previous DNA research, the study is combined with the advantage of others algorithms to improve the DNA genetic algorithm, and the proposed algorithm is applied the parameter estimation. The main contents of the papers are as follows:1. Deep research to the DNA genetic algorithm is conducted based on the previous research. It adopts DNA coding. Besides, three crossover operators and adaptive mutation operator are designed. The diversity of population can improved, which avoids the premature convergence effectively. A new fitness function is designed, which can compare individuals of high similarity. Meanwhile, in order to release the dependence of the range of initial solution on experience set, and to strengthen the global and local search ability, the multi-step evolution strategy with interrupting genetic, simulated annealing algorithm and parameters interval evolution strategy are developed. Numerical experiment on four typical test functions and heavy oil thermal cracking parameter model are carried out to show the efficiency and effectiveness of the proposed algorithms.2. In order to increase the diversity of genetic population and improve the accuracy of reaching results of GA, a novel DNA genetic algorithm based is present on blending population after every given generations. Therefore, the results prove the feasibility and availability of this algorithm in the optimizing of PID controllers’parameters.3. Variance and information entropy to control the evolution process is introduced, and a novel DNA genetic algorithm with population information is designed. A new computing method is designed for crossover probability so as to get a better control of the evolving process of genetic algorithms. The feasibility and effectiveness of the proposed algorithm is experimentally confirmed by the oil blending and scheduling problem.