混沌蚁群优化算法与H-R神经元网络动力学研究

发布时间：2018-03-18 15:42

本文选题：混沌蚁群优化算法　切入点：混沌保密通信　出处：《安徽师范大学》2015年硕士论文　论文类型：学位论文

【摘要】：同步现象是神经元网络动力学的一个重要问题。在神经元网络动力学研究中,科研工作者们一直被同步问题所吸引,并开展了大量的研究工作,比如网络拓扑结构对同步的影响,耦合强度和时间延迟与同步模式转变的关系等等。本文利用混沌蚁群优化算法计算H-R神经元网络的最优耦合关系,处于最优耦合的H-R神经元网络将达到最优同步状态。本文主要包含混沌蚁群优化算法研究和H-R神经元网络最优耦合关系计算两部分。混沌蚁群优化算法是群体智能优化算法的一种,但它和其他群体智能优化算法一样也存在结束条件不易精确设置的缺点。在第二章中,本文根据混沌蚁群优化算法的特点,设计一个非常有效的结束条件,通过多个测试函数验证结束条件的有效性。数值试验表明本文提出的结束条件可以实现多次搜寻最终逼近到最优解。其研究结果已发表在物理学报62卷17期上。在第三章中,利用混沌蚁群优化算法可以对已知部分序列的混沌系统进行参数辨识,参数辨识的目的是获得混沌系统的全部信息。利用混沌系统部分序列参数辨识提出一种简单可行的混沌保密通信方法。文中利用Lorenz系统进行数值试验,数值试验结果表明利用混沌蚁群优化算法可以实现混沌系统部分序列参数辨识,并验证了混沌保密通信方法的可行性。数值试验中发现参数辨识得到的混沌系统并不能长时间与原系统维持同步。在此混沌保密通信方法中,可准确获得的混沌系统序列不仅用于参数辨识,还用于保密通信过程的校正。当两系统偏差较大时需要再次参数辨识以维持与原系统同步,这样在混沌保密通信过程中经过多次参数辨识,可以实现长时间保密通信。其研究结果已发表在物理学报63卷1期上。在第四章中,将H-R神经元网络最优耦合关系问题转化成最优化问题,即寻找一个最优耦合矩阵使神经元网络处于最佳同步状态。然后,利用混沌蚁群优化算法求解此最优化问题,从而得到所要寻找的最优耦合矩阵。通过数值模拟验证搜寻到的最优耦合矩阵可以使H-R神经元网络处于较好的同步状态。最后,本文对将来的研究工作进行了展望。
[Abstract]:Synchronization is an important problem in neuronal network dynamics. In the research of neuronal network dynamics, researchers have been attracted by the synchronization problem and have carried out a lot of research work. For example, the influence of network topology on synchronization, the relationship between coupling intensity and time delay and synchronization mode transformation, etc. In this paper, the optimal coupling relationship of H-R neural network is calculated by using chaotic ant colony optimization algorithm. The H-R neural network in the optimal coupling will achieve the optimal synchronization state. This paper mainly includes two parts: chaos ant colony optimization algorithm and H-R neural network optimal coupling relationship calculation. Chaotic ant colony optimization algorithm is colony intelligence. One of the algorithms that can be optimized, However, like other swarm intelligence optimization algorithms, it also has the disadvantage that the end condition is not easy to set accurately. In the second chapter, according to the characteristics of chaotic ant colony optimization algorithm, a very effective ending condition is designed. The validity of the end condition is verified by several test functions. Numerical experiments show that the end condition proposed in this paper can realize multiple searches and finally approach the optimal solution. The results of the study have been published in the Journal of Physics 62 vol. 17. In Chapter 3, Chaotic ant colony optimization algorithm can be used to identify the parameters of chaotic systems with known partial sequences. The purpose of parameter identification is to obtain all the information of chaotic system. A simple and feasible chaotic secure communication method is proposed by using partial sequence parameter identification of chaotic system. Numerical experiments are carried out using Lorenz system in this paper. Numerical results show that chaotic ant colony optimization algorithm can be used to identify some sequence parameters of chaotic system. The feasibility of the chaotic secure communication method is verified. It is found in the numerical experiments that the chaotic system can not be synchronized with the original system for a long time. In this chaotic secure communication method, The accurately obtained chaotic system sequences are used not only for parameter identification, but also for the correction of secure communication processes. When the deviation between the two systems is large, it is necessary to identify the parameters again in order to maintain synchronization with the original system. In this way, a long time secure communication can be realized through multiple parameter identification in the process of chaotic secure communication. The results of this study have been published in the first issue of 63 volumes of the Journal of Physics. In Chapter 4th, The optimal coupling relation problem of H-R neural network is transformed into an optimization problem, that is, to find an optimal coupling matrix to make the neural network in the optimal synchronization state, and then the chaos ant colony optimization algorithm is used to solve the optimization problem. Through numerical simulation, the optimal coupling matrix can make H-R neural network in good synchronization state. Finally, the future research work is prospected in this paper.
【学位授予单位】：安徽师范大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP18;O415.5;TN918

【参考文献】