Simulink_RBF神经网络PID控制

function [sys, x0, str, ts] = nnrbf_pid( t,x,u,flag,T,nn,K_pid,... eta_pid, xite, alfa, beta0, w0) switch flag, case 0, [sys, x0, str, ts] = mdlInitializeSizes( T,nn) ; case 2, sys = mdlUpdates( u) ; case 3, sys = mdlOutputs( t, x, u, T, nn, K_pid,eta_pid,... xite, alfa, beta0, w0) ; case { 1, 4, 9}, sys = [] ; otherwise, error ( ['Unhandled flag = ' , num2str(flag)]); end function [sys,x0,str,ts] = mdlInitializeSizes(T, nn) sizes = simsizes; sizes. NumContStates = 0; sizes.NumDiscStates = 3; sizes. NumOutputs = 4+ 5* nn; sizes.NumInputs = 9+15* nn; sizes. DirFeedthrough = 1; sizes. NumSampleTimes =1; sys= simsizes( sizes) ; x0= zeros( 3, 1) ; str= [ ] ; ts=[T 0] ; function sys = mdlUpdates( u) sys= [ u(1) - u(2) ; u(1) ; u(1) + u(3) - 2* u(2) ]; function sys = mdlOutputs( t, x, u,T, nn, K_pid, eta_pid,... xite, alfa, beta0, w0) ci3= reshape( u(7: 6+ 3* nn) , 3, nn) ; ci2= reshape(u( 7+ 5* nn: 6+ 8* nn) , 3, nn) ; ci1= reshape( u( 7+ 10* nn: 6+ 13* nn) , 3, nn) ; bi3= u( 7+ 3* nn: 6+ 4* nn) ; bi2= u( 7+ 8*nn: 6+ 9* nn) ; bi1= u( 7+ 13* nn: 6+ 14* nn) ; w3= u( 7+ 4* nn: 6+ 5* nn) ; w2= u( 7+ 9* nn: 6+ 10* nn) ; w1= u( 7+ 14* nn: 6+ 15* nn) ; xx= u( [ 6; 4; 5] ) ; if t== 0 ci1= w0( 1) * ones( 3, nn) ; bi1= w0( 2) *ones( nn, 1) ; w1= w0( 3) * ones( nn, 1) ; K_pid0= K_pid; else, K_pid0= u( end-2: end) ; end for j= 1: nn h(j, 1) = exp( - norm( xx- ci1( : , j) )^2/(2* bi1(j) * bi1(j))) ; end dym= u(4) - w1'* h; w= w1+ xite* dym* h+alfa* (w1- w2) + beta0*( w2- w3) ; for j= 1: nn dbi(j,1) = xite* dym* w1(j) * h(j) * (bi1(j) ^(- 3)) * norm( xx- ci1(:,j))^2; dci( : ,j) = xite*dym* w1(j)* h(j) * (xx- ci1(:,j)) * (bi1(j)^(- 2) ); end bi= bi1+ dbi+ alfa* (bi1- bi2) + beta0*(bi2- bi3) ; ci= ci1+ dci+ alfa* (ci1- ci2) + beta0*(ci2- ci3) ; dJac= sum( w.*h.*( - xx (1) + ci (1,:)' ) ./bi.^2) ; % Jacobian KK(1)= K_pid0(1)+ u(1) * dJac* eta_pid(1)* x(1); KK(2)= K_pid0(2)+ u(1) * dJac* eta_pid(2)* x(2); KK(3)= K_pid0(3)+ u(1) * dJac* eta_pid(3)* x(3); sys= [ u( 6) + KK* x; KK'; ci( : ) ; bi( : ) ; w( : ) ] ;