A deep learning method for solving third-order nonlinear evolution equations

It has still been difficult to solve nonlinear evolution equations analytically. In this paper, we present a deep learning method for recovering the intrinsic nonlinear dynamics from spatiotemporal data directly. Specifically, the model uses a deep neural network constrained with given governing equations to try to learn all optimal parameters. In particular, numerical experiments on several third-order nonlinear evolution equations, including the Korteweg–de Vries (KdV) equation, modified KdV equation, KdV–Burgers equation and Sharma–Tasso–Olver equation, demonstrate that the presented method is able to uncover the solitons and their interaction behaviors fairly well.     

A deep learning method for solving high-order nonlinear soliton equations

We propose an effective scheme of the deep learning method for high-order nonlinear soliton equations and explore the influence of activation functions on the calculation results for higher-order nonlinear soliton equations. The physics-informed neural networks approximate the solution of the equation under the conditions of differential operator, initial condition and boundary condition. We apply this method to high-order nonlinear soliton equations, and verify its efficiency by solving the fourth-order Boussinesq equation and the fifth-order Korteweg–de Vries equation. The results show that the deep learning method can be used to solve high-order nonlinear soliton equations and reveal the interaction between solitons.     

复摆方程的一种求解方法

利用复摆测量物体的转动惯量是一种重要方法,本文对复摆的转动惯量公式进行拉普拉斯变化,用能量和时间来表达转动惯量,丰富了不规则物体转动惯量的测量方法.     

A method for solving the functional equation with variable delay

Mathematical simulation of physical conditions at a moving boundary in problems of wave reflection leads to a finite-difference equation with variable delay. If the reflector velocity is smaller than the wave-propagation velocity, the existence and uniqueness of the equation solution is proved using the principle of contraction maps and the method of successive approximations. It is shown that the solution can be expressed as a function of the composite argument that depends on one variable. The required accuracy of numerical calculations is ensured by partial sums when the argument is represented by an infinite converging series. N. I. Lobachevsky State University, Nizhny Novgorod, Russia. Translated from Izvestiya Vysshikh Uchebnykh Zavedenii, Radiofizika, Vol. 42, No. 9, pp. 900–905, September, 1999     

A physics-constrained deep residual network for solving the sine-Gordon equation

Despite some empirical successes for solving nonlinear evolution equations using deep learning,there are several unresolved issues.First,it could not uncover the dynamical behaviors of some equations where highly nonlinear source terms are included very well.Second,the gradient exploding and vanishing problems often occur for the traditional feedforward neural networks.In this paper,we propose a new architecture that combines the deep residual neural network with some underlying physical laws.Using the sine-Gordon equation as an example,we show that the numerical result is in good agreement with the exact soliton solution.In addition,a lot of numerical experiments show that the model is robust under small perturbations to a certain extent.     

一种深度学习的立体阵波达方向估计方法

针对常规波束形成主瓣宽且目标分辨能力低的问题,提出一种基于深度卷积神经网络的波达方向估计方法。算法使用常规波束形成计算二维空间功率谱,将预处理后的空间功率谱图输入深度卷积神经网络。该文利用神经网络学习解卷积映射关系,输出主瓣宽度更窄的空间功率谱图,从而实现高分辨率二维波达方向估计。该算法对阵列结构没有限制,适用于立体阵。仿真结果表明该文方法在不同目标个数、快拍数及信噪比参数下均能准确估计目标方向。该文方法目标分辨能力优于常规波束形成方法。在低快拍情况下,目标方向估计误差低于自适应波束形成方法。     

A method of solving the Dirac equation

Electromagnetic fields are found for which the Dirac equation reduces to a system of independent second-order differential equations.Translated from Izvestiya Vysshikh Uchebnykh Zavedenii, Fizika, No. 12, pp.41–43, December, 1980.     

A polynomial expansion method for solving the laser Fokker-Planck equation

Distribution functions of the laser amplitude and intensity can be determined by solving the laser Fokker-Planck equation. By a suitable expansion of the distribution functions in Laguerre polynomials, a system of ordinary differential equations for the coefficients of the expansion is derived and is shown to have the form of a recurrence relation with length four. Applying it to the transient solution, the averaged amplitude and the first four cumulants of the intensity distribution are obtained even for those pump parameters where the hitherto known numerical solution is not applicable.     

水下目标多模态深度学习分类识别研究

水下目标的分类识别对于水声探测具有重要意义。提出一种水下目标多模态深度学习分类识别方法。针对水声信号的一维时域模态和二维频域模态特征建立一种多模态特征融合的深度学习结构,结合长短时记忆网络和卷积神经网络的优点,对一维时域信号和二维频谱信号分别进行并行处理,对输出进行典型相关分析,形成特征融合表示,并利用相邻帧的相关性进行参数优化。利用实测水声信号对算法进行了验证。结果表明：提出的算法对于水下目标识别的精度有显著的提高。     

Sinc-collocation method for solving the Blasius equation

Sinc-collocation method is applied for solving Blasius equation which comes from boundary layer equations. It is well known that sinc procedure converges to the solution at an exponential rate. Comparison with Howarth and Asaithambi's numerical solutions reveals that the proposed method is of high accuracy and reduces the solution of Blasius' equation to the solution of a system of algebraic equations.     

Soliton,breather, and rogue wave solutions for solving the nonlinear Schrödinger equation using a deep learning method with physical constraints

The nonlinear Schro¨dinger equation is a classical integrable equation which contains plenty of significant properties and occurs in many physical areas.However,due to the difficulty of solving this equation,in particular in high dimensions,lots of methods are proposed to effectively obtain different kinds of solutions,such as neural networks among others.Recently,a method where some underlying physical laws are embeded into a conventional neural network is proposed to uncover the equation’s dynamical behaviors from spatiotemporal data directly.Compared with traditional neural networks,this method can obtain remarkably accurate solution with extraordinarily less data.Meanwhile,this method also provides a better physical explanation and generalization.In this paper,based on the above method,we present an improved deep learning method to recover the soliton solutions,breather solution,and rogue wave solutions of the nonlinear Schro¨dinger equation.In particular,the dynamical behaviors and error analysis about the one-order and two-order rogue waves of nonlinear integrable equations are revealed by the deep neural network with physical constraints for the first time.Moreover,the effects of different numbers of initial points sampled,collocation points sampled,network layers,neurons per hidden layer on the one-order rogue wave dynamics of this equation have been considered with the help of the control variable way under the same initial and boundary conditions.Numerical experiments show that the dynamical behaviors of soliton solutions,breather solution,and rogue wave solutions of the integrable nonlinear Schro¨dinger equation can be well reconstructed by utilizing this physically-constrained deep learning method.     

焊缝超声相控阵检测数据深度学习降噪方法

超声相控阵检测技术在焊缝检测中具有广泛的应用.超声相控阵检测技术检测信号中常混入噪声导致检测成像时难以分辨真实的缺陷特征.这些噪声主要为无关的反射信号和局部相关的结构噪声,传统的超声图像降噪方法难以有效滤除这些噪声,且存在计算效率低、参数优化复杂等问题.该文提出了一种基于深度学习的焊缝超声相控阵检测技术检测S扫图像的降...     

Soliton,breather,and rogue wave solutions for solving the nonlinear Schr(o)dinger equation using a deep learning method with physical constraints

The nonlinear Schr(o)dinger equation is a classical integrable equation which contains plenty of significant properties and occurs in many physical areas.Howeve...     

基于深度学习的材料超声回波衰减预测方法

材料超声回波衰减是评价材料均匀一致性的常用方法, 针对具有复杂结构的航空发动机盘件难以进行材料底面超声回波衰减评价的问题, 本文提出了利用超声背散射波信号直接预测底面回波衰减的方法。采用10MHz聚焦探头进行超声背散射波数据的采集, 利用深度学习技术构建和训练模型,建立了基于深度学习的材料底面回波衰减预测方法, 同时讨论了采用不同信号形式的超声波信号分类识别模型的准确率差异。研究发现：基于深度学习技术可实现通过超声背散射波预测材料的底面回波衰减, 预测结果和实际底面回波衰减试验结果具有良好的一致性。     

A novel approach for solving the Fisher equation using Exp-function method

In this Letter, Exp-function method is employed to obtain traveling wave solutions of the Fisher equation. It is shown that, on this example, the Exp-function method is easy to implement and concise method for nonlinear evolution equations in mathematical physics.     

求解Kuramoto-Sivashinsky方程的平移基无单元Galerkin方法

Kuramoto-Sivashinsky方程是一种可以描述复杂混沌现象的高阶非线性演化方程.方程中高阶导数项的存在,使得传统无单元Galerkin方法采用高次多项式基函数构造形函数时,形函数违背了一致性条件.因此,本文提出了一种采用平移多项式基函数的无单元Galerkin方法.与传统无单元Galerkin方法相比,该方法在方程离散时依然采用Galerkin进行离散,但形函数的构造采用了基于平移多项式基函数的移动最小二乘近似.通过对具有行波解和混沌现象的Kuramoto-Sivashinsky方程的数值模拟,验证了本文方法的有效性.     

Numerical method for solving matrix coefficient elliptic equation with sharp-edged interfaces

Solving elliptic equations with sharp-edged interfaces is a challenging problem for most existing methods, especially when the solution is highly oscillatory. Nonetheless, it has wide applications in engineering and science. An accurate and efficient method is desired. We propose an efficient non-traditional finite element method with non-body-fitting grids to solve the matrix coefficient elliptic equations with sharp-edged interfaces. Extensive numerical experiments show that this method is second order accurate in the L^∞ norm and that it can handle both sharp-edged interface and oscillatory solutions.     

求解非线性薛定谔方程的简便方法

通过引入变换和选准试探函数,把难于求解的非线性偏微分方程化为易于求解的代数方程,然后用待定系数法确定相应的常数,从而简洁地求得了非线性薛定谔方程的解析解.