Research On DNA Encoding Based On Quantum Computing

In DNA computing, DNA encoding is a module that the practical problems are mapped to the DNA sequence. DNA encoding is influenced by many and complex factors, so DNA encoding is a typical intractable combinatorial optimization problems. Focused on the design of DNA sequence, the paper proposed several new DNA encoding schemes, combined with that Quantum Intelligent Algorithm converges fast and has stronger search capability. The main work of this paper is as follows:Firstly, the relevant theoretical knowledge of DNA encoding was introduced in detail and the mathematical models of the DNA encoding constraints were established. In this paper, five typical constraints of DNA encoding were chose as the objective functions for DNA sequence optimization. In addition, two mathematical methods were used for DNA sequence design: single-objective optimization method and multi-objective optimization method.Secondly, a single-objective optimization method for design of DNA sequences was investigated, and the paper proposed a single-objective method for DNA sequence design based on Quantum Intelligent Algorithm. In the paper, Quantum Genetic Algorithm, Quantum Particle Swarm Algorithm and Quantum Ant Colony Algorithm were respectively used to optimize DNA sequences. Multiple objective functions of DNA encoding problem were converted into single objective function as fitness function by the weighted sum, and then an optimal set of DNA sequences were got by three Quantum Intelligent Optimization Algorithms. With comparison to the traditional method for DNA sequence design, the experiment results show that the proposed new method is better than the traditional methods overall.Thirdly, a multi-objective optimization method for designing DNA sequences was investigated, and a Multi-Objective Quantum Genetic Algorithm based on Pareto domination was introduced to optimize DNA sequences, and another method for DNA encoding was proposed based on MOQGA. In the experiment, the new proposed method was compared to the DNA encoding method based on NSGA-II, and results show that the Pareto optimal solutions set obtained by new method based on MOQGA has a better convergence and the solutions in Pareto optimal set distribute more evenly than NSGA-II.