decode_meijite

Q 学习方面的MATLAB程序，对于研究强化学习和自适应动态规划的朋友很有参考价值(Q learning MATLAB. For the study of reinforcement learning and adaptive dynamic programming )

LQR最优控制，可以应用于基于磁流变阻尼器汽车悬架等半主动控制(LQR optimal control can be applied to semi-active control of vehicle suspension based on MR damper.)

MATLAB遗传算法工具箱,有遗传算法工具箱具体的应用说明(MATLAB genetic algorithm toolbox, a specific application description of the genetic algorithm toolbox)

提供adams matlab联合仿真，供大家参考(adams matlab)

Matlab fdtd （时域有限差分）一维，二维，三维集合(Matlab fdtd (FDTD) one-dimensional, two-dimensional, three-dimensional collection)

说明：  对比了dv-hop基本算法，和加入pso后得算法。由于结果是随机产生的，不能保证每次得结构都能提高，但大部分时候，比基本算法得定位准确度好。(DV hop algorithm is compared with PSO algorithm. Because the results are generated randomly, it can not be guaranteed that every time the structure can be improved, but most of the time, the positioning accuracy is better than the basic algorithm.)

这个代码是用于独立主分量分析的核心部分，里边有现成的例子，说明的也比较详细了，希望大家西哈(this code is for the independent component analysis to the core, inside a ready example, Note also the more detailed, I hope you subjects)

用matlab实现广义最小二乘的数值运算，欢迎大家参考。(Achieve numerical computation of generalized least squares)

M=63 J=(M-1)/2 Wp=0.5*pi Ws=0.6*pi Np=100 Ns=100 weightp=1 weights=100 通带和阻带的加权系数 WpN=linspace(0,Wp,Np) WsN=linspace(Ws,pi-0.05,Ns) Wm=[WpN WsN] p=[ones(1,50),ones(1,50)/2] d=[p,zeros(1,Ns)] W=diag([weightp*ones(1,Np),weights*ones(1,Ns)]) C=cos(Wm*[0:J]) Q=diag(cos(Wm*0.5)) g=(W*Q*C)(W*d) g=g h=(g(J:-1:2)+g(J+1:-1:3))/4 h=[g(J+1)/4 h g(1)*0.5+g(2)/4] h=[h fliplr(h)] W=linspace(0,pi-0.1,512) H=freqz(h,1,W) plot(W/pi,20*log10(abs(H))) (filter design)

详细介绍了matlab中的p文件，asv文件以及mat文件，对matlab初学者者有用(Described in detail in matlab p files, asv files and mat files, useful to beginners on matlab)