turbo

做过模糊PID的朋友很多遇到过这种问题，就是仿真启动不起来，停留在e-7等地方或者仿真很慢，自次上次有个朋友说改变步长能解决这个问题，最后的曲线能达到满足要求的时候，我让他找个例子他就没有后文，以前我做了很多这样的模糊PID，但靠改变步长并没有解决问题，仿真曲线直接发散出去了，我自己这两天也总在思考这个问题，通过不断的调试总结，明白了不少，给遇到这个问题的朋友讲解一下解决办法吧，以一个例子详解：(Done a lot of friends of the fuzzy PID encountered this problem, simulation start up, stay in the e-7 and other local or simulation is very slow, have a friend who said to change the step size to solve this problem since the times the last, the finalcurve to achieve to meet the requirements when I asked him to find examples of his later, I have done a lot of fuzzy PID, but by changing the step does not solve the problem, the simulation curve directly divergence out, Itwo days always thinking about this problem, summarized by the debug and understand a lot of friends to experience this problem to explain the solution, an example Detailed:)

RVM的工具箱,包含拟合，退化预测，分类的源程序和例子。(RVM toolbox, including fitting, degradation prediction, classification of source code and examples.)

模型预测控制，针对广义预测控制s函数的编写，(The writing of s function for generalized predictive control)

matlab 基本知识，基本概念和语法规则，编程技巧。武汉大学教学用的课件(matlab basic knowledge, basic concepts and syntax rules, programming skills. Wuhan University teaching courseware)

unlock v79 by sound track

AUV的建模及其参数整定的Matlab程序，可以复制粘贴到Matlab中(PARAMETER IDENTIFICATION OF ARIES)

multisurface_plasticity

上述程序是美国RICE大学的一个学生编写的用于压缩感知的恢复算法，可以实现很多模型的稀疏恢复，包括BP，l1,l2,l1-l2混合型等比较多的类型(The above procedure is a student of the University of the United States RICE prepared for compressed sensing recovery algorithm can achieve a lot of model sparse recovery, including BP, l1, l2, l1-l2 hybrid more types)

Schroeder s reverberator implemented by matlab

利用polynomial order kernel function来秀出SVM执行时可能产生出来的错误并汇图表示(Use of polynomial order kernel function to SVM showed off that may arise out of implementation errors and exchange graph)