project_soft

written by budi santosa to implement modified PSO on multivariables function budi_s@ie.its.ac.id based on Engineering Optimization Theory and Practice book by Singiresu S. Rao

计算房间冲击响应的matlab函数，亲测可用，界面操作(Matlab code for calculating the pulse response of a room, and the interface operation is available.)

subtightplot is a merger of Pekka Kumpulainen s tight_subplot and Nikolay S. s subplot_tight. It extends the former, which supports asymmetric subplots, to allow for variable margins in addition to gaps, as in the latter.

用递推式法求解；用二分法，牛顿迭代法，迭代法求解方程的近似根(Using recursive method for solving with the dichotomy, Newton iteration method, iteration method for solving the approximate root)

位置指令为幅值为1.0的阶跃信号，r（k）=1.0。网络结构为1-4-1，高斯函数的参数值取Cj=[-2 -1 1 2]T ，B=[0.5 0.5 0.5 0.5]T 。 网络权值学习参数为η=0.30，α=0.05 。PID控制各参数的初RBF网络控制，被控对象为G（s）= 取采样时间为1ms，采用Z变换进行离散化，离散化后的被控对象为 y(k)=-den(2)*y（k-1）-den(3)*y(k-2)+num(2)*u(k-1)+num(3)*u(k-2) 始值为，kp=20， kd=0.3， ki=0.1。 (Position command for the step signal amplitude of 1.0, r (k) = 1.0. The network structure is 1-4-1, the Gaussian function parameters taken Cj = [-2-1 1 2] T, B = [0.5 0.5 0.5 0.5] T. Weight learning parameter η = 0.30, α = 0.05. PID control parameters at the beginning of each RBF network control, the controlled object is G (s) = take the sampling time is 1ms, using the Z transform discrete, discretized controlled object is y (k) =-den (2)* y (k-1)-den (3)* y (k-2)+num (2)* u (k-1)+num (3)* u (k-2) initial value, kp = 20, kd = 0.3, ki = 0.1.)

LMS算法，可以画出自适应滤波器系数收敛曲线(LMS algorithm, the adaptive filter coefficients can draw convergence curve)

感知准则梯降法求线性判别函数权向量的算法(using perceptual criterion to find weightvector of a linear decision function)

edge detection using matlab

部分选主元的高斯消元法的MATLAB代码，并用该方法结合2次回代接方程组(Part of the pivoting Gaussian elimination of the MATLAB code, and combined with this method 2 times back to the pickup equations)

一个解释主成分分析PCA算法的文档,并附有简单的MATLAB程序说明(a document explain principle component analyse)