發(fā)布時(shí)間：2018-03-18 15:42

  本文選題：混沌蟻群優(yōu)化算法　切入點(diǎn)：混沌保密通信　出處：《安徽師范大學(xué)》2015年碩士論文　論文類型：學(xué)位論文

【摘要】：同步現(xiàn)象是神經(jīng)元網(wǎng)絡(luò)動(dòng)力學(xué)的一個(gè)重要問題。在神經(jīng)元網(wǎng)絡(luò)動(dòng)力學(xué)研究中,科研工作者們一直被同步問題所吸引,并開展了大量的研究工作,比如網(wǎng)絡(luò)拓?fù)浣Y(jié)構(gòu)對(duì)同步的影響,耦合強(qiáng)度和時(shí)間延遲與同步模式轉(zhuǎn)變的關(guān)系等等。本文利用混沌蟻群優(yōu)化算法計(jì)算H-R神經(jīng)元網(wǎng)絡(luò)的最優(yōu)耦合關(guān)系,處于最優(yōu)耦合的H-R神經(jīng)元網(wǎng)絡(luò)將達(dá)到最優(yōu)同步狀態(tài)。本文主要包含混沌蟻群優(yōu)化算法研究和H-R神經(jīng)元網(wǎng)絡(luò)最優(yōu)耦合關(guān)系計(jì)算兩部分�；煦缦伻簝�(yōu)化算法是群體智能優(yōu)化算法的一種,但它和其他群體智能優(yōu)化算法一樣也存在結(jié)束條件不易精確設(shè)置的缺點(diǎn)。在第二章中,本文根據(jù)混沌蟻群優(yōu)化算法的特點(diǎn),設(shè)計(jì)一個(gè)非常有效的結(jié)束條件,通過多個(gè)測試函數(shù)驗(yàn)證結(jié)束條件的有效性。數(shù)值試驗(yàn)表明本文提出的結(jié)束條件可以實(shí)現(xiàn)多次搜尋最終逼近到最優(yōu)解。其研究結(jié)果已發(fā)表在物理學(xué)報(bào)62卷17期上。在第三章中,利用混沌蟻群優(yōu)化算法可以對(duì)已知部分序列的混沌系統(tǒng)進(jìn)行參數(shù)辨識(shí),參數(shù)辨識(shí)的目的是獲得混沌系統(tǒng)的全部信息。利用混沌系統(tǒng)部分序列參數(shù)辨識(shí)提出一種簡單可行的混沌保密通信方法。文中利用Lorenz系統(tǒng)進(jìn)行數(shù)值試驗(yàn),數(shù)值試驗(yàn)結(jié)果表明利用混沌蟻群優(yōu)化算法可以實(shí)現(xiàn)混沌系統(tǒng)部分序列參數(shù)辨識(shí),并驗(yàn)證了混沌保密通信方法的可行性。數(shù)值試驗(yàn)中發(fā)現(xiàn)參數(shù)辨識(shí)得到的混沌系統(tǒng)并不能長時(shí)間與原系統(tǒng)維持同步。在此混沌保密通信方法中,可準(zhǔn)確獲得的混沌系統(tǒng)序列不僅用于參數(shù)辨識(shí),還用于保密通信過程的校正。當(dāng)兩系統(tǒng)偏差較大時(shí)需要再次參數(shù)辨識(shí)以維持與原系統(tǒng)同步,這樣在混沌保密通信過程中經(jīng)過多次參數(shù)辨識(shí),可以實(shí)現(xiàn)長時(shí)間保密通信。其研究結(jié)果已發(fā)表在物理學(xué)報(bào)63卷1期上。在第四章中,將H-R神經(jīng)元網(wǎng)絡(luò)最優(yōu)耦合關(guān)系問題轉(zhuǎn)化成最優(yōu)化問題,即尋找一個(gè)最優(yōu)耦合矩陣使神經(jīng)元網(wǎng)絡(luò)處于最佳同步狀態(tài)。然后,利用混沌蟻群優(yōu)化算法求解此最優(yōu)化問題,從而得到所要尋找的最優(yōu)耦合矩陣。通過數(shù)值模擬驗(yàn)證搜尋到的最優(yōu)耦合矩陣可以使H-R神經(jīng)元網(wǎng)絡(luò)處于較好的同步狀態(tài)。最后,本文對(duì)將來的研究工作進(jìn)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.
【學(xué)位授予單位】：安徽師范大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2015
【分類號(hào)】：TP18;O415.5;TN918

【參考文獻(xiàn)】

相關(guān)期刊論文 前1條

1 祝大偉;涂俐蘭;;隨機(jī)擾動(dòng)下Lorenz混沌系統(tǒng)的自適應(yīng)同步與參數(shù)識(shí)別[J];物理學(xué)報(bào);2013年05期

，

本文編號(hào)：1630223