说明：  插值与拟合，解方程，数据分析，包含特征值与特征向量的提取、训练样本以及最后的识别，基于分段非线性权重值的Pso算法，鲁棒性好，性能优越，计算加权加速度,BP神经网络用于函数拟合与模式识别。
( Interpolation and fitting, solution of equations, data analysis, Contains the eigenvalue and eigenvector extraction, the training sample, and the final recognition, Based on piecewise nonlinear weight value Pso algorithm, Robustness, superior performance, Weighted acceleration, BP neural network function fitting and pattern recognition.)