Research Of Modeling And Optimization Of Biological Networks Based On The Improved Particle Swarm Optimization

As an interdisciplinary subject of biology and engineering,synthetic biology combines various disciplines from control theories and system identification,such as feedback and oscillatory behaviors.Due to the complexity of biological systems,existing experience and prior knowledge is not enough to accurately capture their major dynamics and carry out in-depth analysis and research,so it is necessary to construct effective mathematical models by which to obtain the mechanism and characteristics of systems.In recent years,a variety of intelligent optimization algorithms,represented by the particle swarm optimization algorithm,have made significant progress in the field of system modeling.But there is still a problem that algorithms are easy to fall into a local optimal or suboptimal point.To solve the problem of local optimal,this thesis modifies the topological structure of the particle swarm optimization algorithm to improve its global search ability,and carries out the applied research on parameter estimation and structure design of biological networks.The main work of this paper includes:1.In order to solve the problem of falling into a local optimum and high sensitivity to parameters,considering local topological structure has the characteristics of enhancing the global search ability of the algorithm,this thesis proposes random drift particle swarm optimization algorithm with Von Neumann topological structure.The performance of the improved algorithm is verified through simulations on several classics functions.2.There are nonlinear puzzles when we estimate the parameters of the biological networks.The proposed algorithm which has strong global search ability is used to solve this problem,and the results are compared with other four algorithms.Results show that the proposed algorithm can effectively improve the performance of parameter estimation.3.In order to design the synthetic genetic oscillation network with strong robust performance,the discrete form of the improved algorithm is proposed.Combined with the idea of two-step optimization,we conduct the optimization design of robust performance on the basis of the optimization and determination of the network structure.Results show that the robust performance of networks on the structure and noise is improved greatly after the robust performance optimization.4.When designing the coupled oscillator network with the strong synchronous behavior,if we consider the influence of structures and parameters on the synchronous behavior at the same time,the problem is a mixed integer optimization problem.The proposed optimization algorithm is used to the simulation research of a typical instance.The result shows that this method can get the coupled oscillator network with the strong synchronizing characteristics.