mirror of
https://github.com/ohmyjesus/RBF_NeuralNetwork.git
synced 2026-08-17 17:11:39 +08:00
145 lines
3.5 KiB
Matlab
145 lines
3.5 KiB
Matlab
function [sys,x0,str,ts] = Book6141_2_Controller(t,x,u,flag)
|
|
% 以下程序是 基于RBF神经网络的直接鲁棒自适应控制
|
|
switch flag
|
|
case 0 %初始化
|
|
[sys,x0,str,ts]=mdlInitializeSizes;
|
|
case 1 %连续状态计算
|
|
sys=mdlDerivatives(t,x,u);
|
|
case {2,4,9} %离散状态计算，下一步仿真时刻，终止仿真设定
|
|
sys=[];
|
|
case 3 %输出信号计算
|
|
sys=mdlOutputs(t,x,u);
|
|
otherwise
|
|
DAStudio.error('Simulink:blocks:unhandledFlag', num2str(flag));
|
|
end
|
|
|
|
function [sys,x0,str,ts]=mdlInitializeSizes %系统的初始化
|
|
global c b node
|
|
% 神经网络采用2-19-1结构
|
|
node = 19;
|
|
c = 0.5 * [-4.5 -4 -3.5 -3 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5;
|
|
-4.5 -4 -3.5 -3 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5]; % 高斯函数的中心点矢量 维度 IN * MID 2*19
|
|
b = 2 * ones(19,1); % 高斯函数的基宽 维度node * 1 19*1 b的选择很重要 b越大 网路对输入的映射能力越大
|
|
sizes = simsizes;
|
|
sizes.NumContStates = node; %设置系统连续状态的变量 W
|
|
sizes.NumDiscStates = 0; %设置系统离散状态的变量
|
|
sizes.NumOutputs = 3; %设置系统输出的变量
|
|
sizes.NumInputs = 3; %设置系统输入的变量
|
|
sizes.DirFeedthrough = 1; %如果在输出方程中显含输入变量u，则应该将本参数设置为1
|
|
sizes.NumSampleTimes = 0; % 模块采样周期的个数
|
|
% 需要的样本时间，一般为1.
|
|
% 猜测为如果为n，则下一时刻的状态需要知道前n个状态的系统状态
|
|
sys = simsizes(sizes);
|
|
x0 = 0.1 * ones(node,1); % 系统初始状态变量 代表W和V向量
|
|
str = []; % 保留变量，保持为空
|
|
ts = []; % 采样时间[t1 t2] t1为采样周期，如果取t1=-1则将继承输入信号的采样周期；参数t2为偏移量，一般取为0
|
|
|
|
|
|
function sys = mdlDerivatives(t,x,u) %该函数仅在连续系统中被调用，用于产生控制系统状态的导数
|
|
global c b node
|
|
% 仿真中应根据网络输入值的有效映射范围来设计 c和b 从而保证有效的高斯映射 不合适的b或c均会导致结果不正确
|
|
IN = 2;
|
|
Out = 1;
|
|
qd = sin(t);
|
|
dqd = cos(t);
|
|
ddqd = -sin(t);
|
|
|
|
qd = u(1);
|
|
q = u(2); % e = q - qd
|
|
dq = u(3);
|
|
e = q - qd;
|
|
de = dq - dqd;
|
|
|
|
% 参数的定义
|
|
M = 10;
|
|
gama = 1200;
|
|
alph = 3;
|
|
kp = alph^2;
|
|
kv = 2 * alph;
|
|
Q = [50 0; 0 50];
|
|
A = [0 1; -kp -kv];
|
|
P = lyap(A' , Q);
|
|
B = [0; 1/M];
|
|
k1 = 0.001;
|
|
|
|
Input = [e; de];
|
|
h = zeros(node , 1); %19*1矩阵
|
|
for i =1:node
|
|
h(i) = exp(-(norm(Input - c(:,i))^2) / (2*b(i)^2));
|
|
end
|
|
W = [x(1:19)]'; %node*1 19*1
|
|
|
|
method = 1;
|
|
if method == 1 % 自适应方法一
|
|
dw = gama * h * Input' * P * B;
|
|
for i = 1:node
|
|
sys(i) = dw(i);
|
|
end
|
|
else % 自适应方法二
|
|
dw = gama * h * Input' * P * B + k1 * gama * norm(Input) * W;
|
|
for i = 1:node
|
|
sys(i) = dw(i);
|
|
end
|
|
end
|
|
|
|
|
|
function sys = mdlOutputs(t,x,u) %产生（传递）系统输出
|
|
global c b node
|
|
IN = 2;
|
|
Out = 1;
|
|
qd = sin(t);
|
|
dqd = cos(t);
|
|
ddqd = -sin(t);
|
|
|
|
qd = u(1);
|
|
q = u(2); % e = q - qd
|
|
dq = u(3);
|
|
e = q - qd;
|
|
de = dq - dqd;
|
|
|
|
% 参数的定义
|
|
M = 10;
|
|
gama = 1200;
|
|
alph = 3;
|
|
kp = alph^2;
|
|
kv = 2 * alph;
|
|
Q = [50 0; 0 50];
|
|
A = [0 1; -kp -kv];
|
|
P = lyap(A' , Q);
|
|
B = [0; 1/M];
|
|
k1 = 0.001;
|
|
|
|
Input = [e; de];
|
|
h = zeros(node , 1); %5*1矩阵
|
|
for i =1:node
|
|
h(i) = exp(-(norm(Input - c(:,i))^2) / (2*b(i)^2));
|
|
end
|
|
W = [x(1:19)]; %node*1 19*1
|
|
% 神经网络的输出
|
|
fx = W' * h;
|
|
d = -15 * dq - 30 *sign(dq);
|
|
|
|
some = 1;
|
|
if some == 1
|
|
ut = M * (ddqd - kv*de - kp*e) - fx;
|
|
elseif some == 2
|
|
ut = M * (ddqd - kv*de - kp*e) - d;
|
|
else
|
|
ut = M * (ddqd - kv*de - kp*e);
|
|
end
|
|
sys(1) = ut;
|
|
sys(2) = fx;
|
|
sys(3) = d;
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|