【拟合】基于RBF-NN 进行非线性系统识别附matlab代码.zip

%% START clc; close all; clear all; fp = 0.9; alpha_f=0.5; len = 500; % Length of the signal epochs = 1000; % Number of times signal passes through ADF for weight adaptation runs = 10; tempIc=0; tempIf=0; tempIst=0; for run=1:runs meu_c = 2e-5;% Step size meu_f = 2e-5;% Step size meu_st = 1e-2;% Step size x=[ones(1,round(len/4)) -ones(1,round(len/4)) ones(1,round(len/4)) -ones(1,round(len/4))]; x=awgn(x,10); %% Defining Unknown System h = [2 -0.5 -0.1 -0.7 3]; c = [-5:2:5]; n1=length(c); W_c = randn(1,n1); % Weights W_f = randn(1,n1); % Weights W_st = randn(3,n1); % Weights beeta=1; b_c=randn(1); b_f = b_c; b_st = b_c; noise_level = sqrt(0.01); for k=1:epochs Ic(k)=0; If(k)=0; Ist(k)=0; U = zeros(3,1); U(2:end)=[-1 -1]; for i1=1:len U(1:end-1)=U(2:end); U(end)=x(i1); for i2=1:n1 ED_cf(:,i2)=exp((-(norm(U-c(i2))^2))/beeta^2); end for i2=1:n1 ED_st(:,i2)=exp((-(abs(U-c(i2))))/beeta^2); end %% RBF yc(i1)=sum(diag(W_c*ED_cf'))+b_c; d(i1)= h(1)*U(end) +h(2)*U(end-1)+h(3)*U(end-2)+h(4)*(cos(h(5)*U(end)) +exp(-abs(U(end))))+noise_level*randn(); e_c=d(i1)-yc(i1); Ic(k)=Ic(k)+e_c*e_c'./len; %%% Objective Function W_c=W_c+meu_c*e_c*ED_cf; b_c=b_c+meu_c*e_c; %% Fractional RBF yf(i1)=sum(diag(W_f*ED_cf'))+b_f; % d(i1)= h(1)*U(end) +h(2)*U(end-1)+h(3)*U(end-2)+h(4)*(cos(h(5)*U(end)) +exp(-abs(U(end))))+0.1*randn(); e_f=d(i1)-yf(i1); If(k)=If(k)+e_f*e_f'./len; %%% Objective Function W_f=W_f+meu_f*e_f*ED_cf.*((1-alpha_f)+alpha_f*abs(W_f.^fp)); b_f=b_f+meu_f*e_f.*((1-alpha_f)+alpha_f*abs(b_f.^fp)); %% ST-RBF yst(i1)=sum(diag(W_st*ED_st'))+b_st; % d(i1)= h(1)*U(end) +h(2)*U(end-1)+h(3)*U(end-2)+h(4)*(cos(h(5)*U(end)) +exp(-abs(U(end))))+0.1*randn(); e_st=d(i1)-yst(i1); Ist(k)=Ist(k)+e_st*e_st'./len; %%% Objective Function W_st=W_st+meu_st*e_st*ED_st;%*G; b_st=b_st+meu_st*e_st; % q = alpha_q*q + gamma_q*eq^2; % if (q>q_max) % q=q_max; % end end % q_track(k) = q; end % time=toc % tempq_track= tempq_track + q_track; tempIc=tempIc+Ic; tempIf=tempIf+If; tempIst=tempIst+Ist; end % q_track = tempq_track./runs; Ic=tempIc./runs; If=tempIf./runs; Ist=tempIst./runs; save comparison_train.mat % plot(q_track) figure semilogy(Ic,'r') hold on semilogy(If,'k') semilogy(Ist,'b') figure len = 200; % Length of the signal y=0; Ic=0; yf=0; If=0; d=0; x=0; yst=0; Ist=0; x=[-1 -1 ones(1,round(len/4)) -ones(1,round(len/4)) ones(1,round(len/4)) -ones(1,round(len/4))]; x=awgn(x,10); U(2:end)=[-1 -1]; n1=length(c); Ic=0; for i1=1:len-1 U(1:end-1)=U(2:end); U(end)=x(i1); for i2=1:n1 ED_cf(:,i2)=exp((-(norm(U-c(i2))^2))/beeta^2); end for i2=1:n1 ED_st(:,i2)=exp((-(abs(U-c(i2))))/beeta^2); end yc(i1)=sum(diag(W_c*ED_cf'))+b_c; d(i1)= h(1)*U(end) +h(2)*U(end-1)+h(3)*U(end-2)+h(4)*(cos(h(5)*U(end)) +exp(-abs(U(end)))); e_c=d(i1)-yc(i1); SE(i1)=e_c*e_c'; %%% Objective Function Ic(i1)=mean(SE); yf(i1)=sum(diag(W_f*ED_cf'))+b_f; % d(i1)= h(1)*U(end) +h(2)*U(end-1)+h(3)*U(end-2)+h(4)*(cos(h(5)*U(end)) +exp(-abs(U(end)))); e_f=d(i1)-yf(i1); SEf(i1)=e_f*e_f'; %%% Objective Function If(i1)=mean(SEf); yst(i1)=sum(diag(W_st*ED_st'))+b_st; % d(i1)= h(1)*U(end) +h(2)*U(end-1)+h(3)*U(end-2)+h(4)*(cos(h(5)*U(end)) +exp(-abs(U(end)))); e_st=d(i1)-yst(i1); SEq(i1)=e_st*e_st'; %%% Objective Function Ist(i1)=mean(SEq); end hold on semilogy(Ic,'r') semilogy(If,'k') semilogy(Ist,'b') %saveas(gcf,strcat('Iq.png'),'png') figure plot(d,'cy') hold on plot(y,'r') plot(yf,'r') plot(yst,'b') legend('RBF','FRBF','desired','qRBF') %saveas(gcf,strcat('RBF.png'),'png') % save(strcat('MSE','.mat')); %save(['q_track','.mat'],'q_track'); % save(['y','.mat'],'y'); save(['yf','.mat'],'yf'); save(['d','.mat'],'d'); save(['yq','.mat'],'yq'); % save(['I','.mat'],'I');save(['Iq','.mat'],'Iq'); save comparison_test.mat graph_channel