code2

采用动量梯度下降算法训练BP网络，采用两种训练方法，即 L-M 优化算法（trainlm）和贝叶斯正则化算法（trainbr），用以训练 BP 网络(Gradient descent algorithm using momentum BP network training, using two training methods, namely, LM optimization algorithm (trainlm) and Bayesian regularization algorithm (trainbr), to facilitate the training of BP network)

FFT源代码 FFT源代码 FFT源代码(FFT source code FFT source code FFT source code)

分别采用Roe格式和AUSM格式求解一维黎曼问题，并将计算结果与解析解进行比较，发现这两种计算格式都能获得高的计算精度。(Roe scheme were used in this procedure and AUSM format to solve a 1-D RIEMANN problem, and numerical results were compared with the analytical solution, the results can be seen from these two formats have a solution with high accuracy.)

说明：  二维FDTD程序计算近场模型，包括建模文件以及电磁场递推文件，入射角度为0°。模型时一个介质方柱。(The two-dimensional FDTD program calculates near-field models, including modeling files and electromagnetic field recursion files, with an incident angle of 0. The model is a medium square column.)

计算基尔霍夫近似下镜面与非镜面的反射系数，同时得到距离散射点2米的接收功率(Calculation Kirchhoff approximation of specular and non-specular reflection coefficient, while the reception power obtained from the scattering points of 2 meters)

该程序用于瞬变电磁勘探近区场计算，只要输入不同参数便可以得到近区场的视电阻率(The procedure transient electromagnetic exploration for near field calculations, as long as the input parameters can be different near field apparent resistivity)

复杂网络的结构检测室近几年的研究热点，目前最快的算法是标签传播！(Community detection is hot topic in recent years. The most fast algorithm for community detection is Label Propagation Algorithm(LPA),and this is the program!)

说明：  这是一个ACM试题中的算法，叫分而治之算法，上面有PPT还有实例代码，希望喜欢ACM的同学可以喜欢(This is a Question in the ACM algorithm, called divide and conquer algorithms, there are examples of the code above, there are PPT hope that students like to be like the ACM)

计算风力机的转换功率的算例 根据流体力学激盘素理论推出 最大转换功率不可能超过0.47(cofficience of power change from wind energy to eletricity energy.)

利用Mie散射原理计算多层介质球的散射特性(Principles calculation using Mei scattering scattering properties of multilayer dielectric sphere)