神经网络的分叉理论设计方法

本文用分叉理论的规范形方程设计和综合期望贮存静、动态记忆模式的神经网络。对于期望贮存静态记忆模式的网络，该规范形方程为叉形分叉的；若期望贮存的记忆模式是周期振荡形式，该规范形方程为高余维数Ｈｏｐｆ分叉的，由满足设计约束的规范形系数得到的突触连接系数可以保证期望贮存的记忆模式都能成功地存贮于所设计的网络，且是网络仅有的吸引子，没有伪吸引子，吸引域的范围足够大。     

规范形网络中的混沌吸引子

讨论多余维Hopf分叉三阶规范形的普适开折形成的网络更进一步的复杂动力学行为．通过对余维二Hopf分叉的规范形网络多级分叉的分析，发现在参数空间的某个区域会出现二环面，将S形非线性加入规范形网络，在出现二环面的区域内可以出现混沌．本文给出了该混沌吸引子的相图及其二阶Poincare映射的图景．由这些图可以看到该混沌吸引子具有非常奇妙的形态：某些二阶Poincare映射像一只逼真的蝴蝶．     

规范形网络中的混沌吸引子

讨论多余维Hopf分叉三阶规范形的普适开折形成的网络更进一步的复杂动力学行为．通过对余维二Hopf分叉的规范形网络多级分叉的分析，发现在参数空间的某个区域会出现二环面，将S形非线性加入规范形网络，在出现二环面的区域内可以出现混沌．本文给出了该混沌吸引子的相图及其二阶Poincare映射的图景．由这些图可以看到该混沌吸引子具有非常奇妙的形态：某些二阶Poincare映射像一只逼真的蝴蝶．     

驱动长方形腔内流动非稳定性的数值模拟

本文对长宽比为２的驱动腔内流动进行了数值模拟．采用非均匀交错网格上的修正隐式Ｔｅｍａｍ格式，以及压力修正投影法，分别计算了Ｒｅ数为１００、４００、１０００、２０００、３０００、３５００、５０００、１００００的驱动长方形腔内流场。当Ｒｅ≤３０００时，流场收敛到定常状态；而Ｒｅ≥３５００时，只能得到渐近周期结果；其中应用了谱分析等方法说明数值是周期性变化，可见，Ｈｏｐｆ分叉点出现在Ｒｅ数３０００与３５００之间．     

多孔介质中热对流的分叉机理研究

本文利用解析分析方法研究了数值模拟发现的多孔介质层中出现的对流分叉机理，指出控制方程中的Ｒａｙｌｅｉｇｈ数，是决定流动的特征参数。当Ｒａｙｌｅｉｇｈ数小于临界数值时，多孔介质内流动处于静止传热状态，并且这种状态是稳定的。如果Ｒａｙｌｅｉｇｈ数大于临界数值，非线性方程出现分叉解，文中指出，存在多个使平凡解失稳而分叉的临界Ｒａｙｌｅｉｇｈ数，当Ｒａｙｌｅｉｇｈ数由小到大经历这些临界数值时，其由平凡解发展起来的分叉解的流态，依次由单回流区转变为双回流区及三回流区。理论分析给出了分叉解和分叉解的振幅方程，阐明了分叉的机理，其结论和数值结果定性一致．     

温盐双扩散系统的多平衡稳定性

采用数值试验的方法研究了温盐双扩散系统中的多平衡稳定性在系统中引入非线性状态方程、温通量边界条件和变强迫因子在数值模拟中引入边界拟会坐标系,并按照温盐浮力比λ大于1,等于1和小于1将流场分为温度控制场、温盐均势场和盐度控制场得到了四种稳定平衡状态,特别是,大强迫因子和强非线性影响下的四胞环流对称状态．在对称的定常强迫边条件作用下的流场、温度场和盐度场均为对称分布引入温通量边条件后,对称流场在小干扰下失去对称性,发展为非对称分布,并因为干扰位置的不同而造成不同的非对称性,出现了从对称状态到非对称状态的音叉分叉,得到了a-Ras的稳定曲线及分叉图,并解释了分叉的本质     

非线性问题和分叉问题及其数值方法

本文给出了一个一般性的分叉定义，说明了伪弧长算法在分叉计算中的应用，概述了静分叉点定位、用单纯形算法准确确定静分叉后各分叉解枝初始方向的算法，以及Ｈｏｐｆ分叉点定位和大范围连续追踪周期解轨道的数值方法。     

并列双圆柱流致振动的不对称振动和对称性迟滞研究

对雷诺数Re = 100 间距比s/D = 2.5 和5.0 的并列双圆柱流致振动进行了数值模拟研究, 其中圆柱质量比m = 2.0, 折合流速U_r 在2.0~10.0 之间, 两圆柱仅能做横流向振动. 研究发现, 当间距比s/D = 2.5 时, 在折合流速4.4 < U_r< 4.8区间内, 两圆柱流致振动响应出现不对称振动现象, 在折合流速4.4 < U_r< 4.8 区间内, 两圆柱流致振动响应出现对称性迟滞现象; 而当间距比s/D = 2.5时, 圆柱流致振动响应与单圆柱涡激振动响应相似, 没有出现不对称振动和对称性迟滞现象. 在不对称振动区间内, 两圆柱的升、阻力参数也出现了不相等的情况. 此外, 当两圆柱不对称振动时, 圆柱间隙流稳定地偏斜向其中的一个圆柱; 相应地, 尾涡也出现了宽窄不等的模式. 窄尾流圆柱的振幅和升、阻力均较宽尾流圆柱的大. 通过对比不对称振动现象发生前后的尾涡模式, 对新现象的产生机制进行了阐述.      

密频近线性系统2阶超谐共振Hopf分叉

研究了一个两自由度密频近线性系统２阶超谐共振Ｈｏｐｆ分叉，揭示出由于密频内共振的作用，系统将存在较复杂的动力学行为。     

时滞Lienard非线性系统的Hopf分岔

本文研究了Ｌｉｅｎａｒｄ非线性时滞系统的线性稳定性和Ｈｏｐｆ分岔，考究了特征方程随两参数变化时根的分布，应用中心流形和范式分析失去线性稳定性出现的Ｈｏｐｆ分岔及其稳定性。     

Observing Symmetry-Breaking and Chaos in the Normal Form Network

The complex dynamical behaviors of neural networks may deducenew information processing methodology. In this paper, the dynamics of anormal form network with Z ₂ symmetry is studied. Thesecondary Hopf bifurcation of the network is discussed and a two-torusis observed. Examining the phase-locking motions of the two-torus, wepresent the regularity of symmetry-breaking occurring in the system. Ifthe ratio of the two frequencies of the codimension-two Hopf bifurcationis represented by an irreducible fraction, symmetry-breaking occurs wheneither the numerator or the denominator of the fraction is even. Chaoticattractors may be created with sigmoid nonlinearities added to theright-hand side of the normal form equations. The trajectory andsecond-order Poincaré maps of the chaotic attractor are given.The chaotic attractor looks like a butterfly on some of the second-orderPoincaré maps. This is a marvelous example for chaos mimickingnature.     

冲击消振器的概周期碰振运动分析

建立了冲击消振器对称周期运动的Poincar啨映射方程 ,讨论了对称周期运动的稳定性与局部分岔。通过数值仿真研究了冲击消振器在非共振、弱共振和强共振条件下的概周期碰振运动及其向混沌的转迁过程。     

神经网络时滞系统非共振双Hopf分岔及其广义同步

本文建立了具有自连接和抑制-兴奋型他连接的两个同性神经元模型。其中自连接是由于兴奋型的突触产生,而他连接则分别对应于两神经元兴奋、抑制型的突触。发现如果有兴奋型自连接就会有双Hopf分岔,而没有时滞自连接时双Hopf分岔就会消失,因此自连接引起了双Hopf分岔。作为一个例子,通过变动连接中的时滞和他连接中的比重,1／√2双Hopf分岔得到了详细研究。通过中心流形约化,分岔点邻域内各种不同的动力学行为得到了分类,并以解析形式表出。神经元活动的分岔路径得以表明。从得到的解析近似解可以发现,本文所研究的具有兴奋一抑制型他连接的两相同神经元的节律不能完全同步而只能广义同步。时滞也可以使其节律消失,两神经元变为非活动的。这些结果在控制神经网络关联记忆和设计人工神经网络方面有着潜在的应用。     

Hopf bifurcation analysis of asymmetrical rotating shafts

The Hopf and double Hopf bifurcations analysis of asymmetrical rotating shafts with stretching nonlinearity are investigated. The shaft is simply supported and is composed of viscoelastic material. The rotary inertia and gyroscopic effect are considered, but, shear deformation is neglected. To consider the viscoelastic behavior of the shaft, the Kelvin–Voigt model is used. Hopf bifurcations occur due to instability caused by internal damping. To analyze the dynamics of the system in the vicinity of Hopf bifurcations, the center manifold theory is utilized. The standard normal forms of Hopf bifurcations for symmetrical and asymmetrical shafts are obtained. It is shown that the symmetrical shafts have double zero eigenvalues in the absence of external damping, but asymmetrical shafts do not have. The asymmetrical shaft in the absence of external damping has a saddle point, therefore the system is unstable. Also, for symmetrical and asymmetrical shafts, in the presence of external damping at the critical speeds, supercritical Hopf bifurcations occur. The amplitude of periodic solution due to supercritical Hopf bifurcations for symmetrical and asymmetrical shafts for the higher modes would be different, due to shaft asymmetry. Consequently, the effect of shaft asymmetry in the higher modes is considerable. Also, the amplitude of periodic solutions for symmetrical shafts with rotary inertia effect is higher than those of without one. In addition, the dynamic behavior of the system in the vicinity of double Hopf bifurcation is investigated. It is seen that in this case depending on the damping and rotational speed, the sink, source, or saddle equilibrium points occur in the system.     

Symmetry breaking and preliminary results about a Hopf bifurcation for incompressible viscous flow in an expansion channel

This computational study shows, for the first time, a clear transition to two-dimensional Hopf bifurcation for laminar incompressible flows in symmetric plane expansion channels. Due to the well-known extreme sensitivity of this study on computational mesh, the critical Reynolds numbers for both the known symmetry-breaking (pitchfork) bifurcation and Hopf bifurcation were investigated for several layers of mesh refinement. It is found that under-refined meshes lead to an overestimation of the critical Reynolds number for the symmetry breaking and an underestimation of the critical Reynolds number for the Hopf bifurcation.     

Double Hopf bifurcation of time-delayed feedback control for maglev system

This paper undertakes an analysis of a double Hopf bifurcation of a maglev system with time-delayed feedback. At the intersection point of the Hopf bifurcation curves in velocity feedback control gain and time delay space, the maglev system has a codimension 2 double Hopf bifurcation. To gain insight into the periodic solution which arises from the double Hopf bifurcation and the unfolding, we calculate the normal form of double Hopf bifurcation using the method of multiple scales. Numerical simulations are carried out with two pairs of feedback control parameters, which show different unfoldings of the maglev system and we verify the theoretical analysis.     

Stability and bifurcation analysis in tri-neuron model with time delay

A simple delayed neural network model with three neurons is considered. By constructing suitable Lyapunov functions, we obtain sufficient delay-dependent criteria to ensure global asymptotical stability of the equilibrium of a tri-neuron network with single time delay. Local stability of the model is investigated by analyzing the associated characteristic equation. It is found that Hopf bifurcation occurs when the time delay varies and passes a sequence of critical values. The stability and direction of bifurcating periodic solution are determined by applying the normal form theory and the center manifold theorem. If the associated characteristic equation of linearized system evaluated at a critical point involves a repeated pair of pure imaginary eigenvalues, then the double Hopf bifurcation is also found to occur in this model. Our main attention will be paid to the double Hopf bifurcation associated with resonance. Some Numerical examples are finally given for justifying the theoretical results.     

随机干扰与随机参数激励联合作用下的Hopf分叉

本文研究van der Pol-Duffing型的非线性振子在随机干扰和随机参数联合作用下的Hopf分叉现象。本文所得结果证实了当系统处在于Hopf分叉点附近时,对系统的参数的变化具有敏感性。在研究过程中,我们利用Markov扩散过程逼近系统的随机响应,得到了沿稳定矩的概率1稳定和矩稳定的条件。对于非线性振子,我们得到了振幅过程的稳态概论密度函数。研究发现,确定性系统的Hopf分叉点在随机参数作用下具有漂移现象,这种漂移是由系统的性质所决定的,当分叉点为超临界的,分叉点向前漂移;而当分叉点为亚临界时,这种漂移是向后的。当系统处在外部随机干扰作用下时,系统出现非零响应。另外我们发现,稳态矩的分叉与其阶数无关。