BP_PID_PID参数整定_bppid自整定_BP神经网络pid控制_

%BP based PID Control clear all; close all; %model parameter x=[0,0,0]; du_1=0; %控制器输出变换量初始值 u_1=0;u_2=0;u_3=0;u_4=0;u_5=0;u_6=0; % y_1=0;y_2=0;y_3=0; %扰动部分的输入、输出初始化 u_1r=0; y_1r=0; u_1d=0;u_2d=0;u_3d=0;u_4d=0;u_5d=0; y_1d=0;y_2d=0; %前馈部分的输入 u_1b=0;u_2b=0;u_3b=0;u_4b=0;u_5b=0;u_6b=0;u_7b=0; y_1b=0;y_2b=0;y_3b=0;y_4b=0; ts=1; %采样周期 K=20; A=1+0.1*K; sys1=tf(0.0004*K,[30,(3+10*A),A,0],'inputdelay',3); %建立被控对象传递函数 dsys1=c2d(sys1,ts,'z'); %传递函数的离散化 [num1,den1]=tfdata(dsys1,'v'); %提取传递函数的分子分母 %扰动的一部分,惯性环节 B=0.01; sys2=tf(B,[8,1]); dsys2=c2d(sys2,ts,'z'); [num2,den2]=tfdata(dsys2,'v'); %扰动的另一部分 C=0.0003; sys3=tf(C,[8,1,0],'inputdelay',3); dsys3=c2d(sys3,ts,'z'); [num3,den3]=tfdata(dsys3,'v'); %扰动的前馈 K1=0.1;E=1; sys4=tf(0.0004*K1*K1,[30*E,(30+(3+10*A)*E),((3+10*A)+A*E),(A+1*E),0],'inputdelay',3); dsys4=c2d(sys4,ts,'z'); [num4,den4]=tfdata(dsys4,'v'); %NN set xite=0.20;%%%%学习速率 alfa=0.01;%%%%惯性因子 IN=3;H=2;Out=3; %NN Structure wi=0.50*rands(H,IN); %隐含层加权系数wi初始化 wi_1=wi;wi_2=wi;wi_3=wi; wo=0.50*rands(Out,H); %输出层加权系数wo初始化 wo_1=wo;wo_2=wo;wo_3=wo; Oh=zeros(H,1); %隐含层的输出 I=Oh; %隐含层的输入 error_2=0; error_1=0; for k=1:1:1000 time(k)=k*ts; if(k>=500) u_r(k)=-3; u_b(k)=-3; else u_r(k)=0; u_b(k)=0; end rin(k)=3.0; %Unlinear model y1out(k)=-den1(2)*y_1-den1(3)*y_2-den1(4)*y_3+num1(2)*u_4+num1(3)*u_5+num1(4)*u_6; y2out(k)=-den2(2)*y_1r+num2(2)*u_1r;%u_1r为扰动输入,y_1r为惯性环节下的扰动输出 y3out(k)=-den3(2)*y_1d-den3(3)*y_2d+num3(2)*u_4d+num3(3)*u_5d; y4out(k)=-den4(2)*y_1b-den4(3)*y_2b-den4(4)*y_3b-den4(5)*y_4b+num4(2)*u_4b+num4(3)*u_5b+num4(4)*u_6b+num4(5)*u_7b; yout(k)=y1out(k)+y2out(k)-y3out(k)+y4out(k);% %----------误差以及误差变化量-------------% error(k)=rin(k)-yout(k);%给定值为3 xi=[error(k),u_r(k),1];%增加一个输入, x(1)=error(k)-error_1;%比例元 x(2)=error(k);%积分元 x(3)=error(k)-2*error_1+error_2;%微分元 epid=[x(1);x(2);x(3)];%分别为比例元、积分元、微分元。 I=xi*wi';%隐含层的输入，即输入层输入*权值 for j=1:1:H Oh(j)=(exp(I(j))-exp(-I(j)))/(exp(I(j))+exp(-I(j))); %在激活函数作用下隐含层的输出 end K=wo*Oh; %输出层对应的输入，即隐含层的输出*权值 for l=1:1:Out K(l)=exp(K(l))/(exp(K(l))+exp(-K(l))); %输出层的输出，Getting kp,ki,kd end kp(k)=K(1);ki(k)=K(2);kd(k)=K(3);%PID参数值 %Kpid=[kp(k)+10,0.001,0]; %Kpid=[kp(k) ki(k) kd(k)]; Kpid=[10,0.001,2]; du(k)=Kpid*epid; %PID控制器的输出增量 u(k)=u_1+du(k); %PID控制器的输出 if (u(k)>50) %控制器的饱和环节 u(k)=50; end if (u(k)<-50) u(k)=-50; end dyu(k)=sign((yout(k)-y_1)/(du(k)-du_1+0.000001)); %y_1初始值为零 %Output layer for j=1:1:Out dK(j)=2/(exp(K(j))+exp(-K(j)))^2;%输出层激活函数的导数 end for l=1:1:Out delta3(l)=error(k)*dyu(k)*epid(l)*dK(l); end for l=1:1:Out for i=1:1:H d_wo=xite*delta3(l)*Oh(i);%+alfa*(wo_1-wo_2); end end wo=wo_1+d_wo+alfa*(wo_1-wo_2);%权值更新 %Hidden layer for i=1:1:H dO(i)=4/(exp(I(i))+exp(-I(i)))^2;%隐含层激活函数的导数 end segma=delta3*wo; for i=1:1:H delta2(i)=dO(i)*segma(i); end d_wi=xite*delta2'*xi; wi=wi_1+d_wi;%alfa*(wi_1-wi_2);%权值更新 %Parameters Update du_1=du(k); u_6=u_5;u_5=u_4;u_4=u_3;u_3=u_2;u_2=u_1;u_1=u(k); y_3=y_2;y_2=y_1;y_1=y1out(k); u_1r=u_r(k); y_1r=y2out(k); u_5d=u_4d;u_4d=u_3d;u_3d=u_2d;u_2d=u_1d;u_1d=u_r(k); y_2d=y_1d;y_1d=y3out(k); u_7b=u_6b;u_6b=u_5b;u_5b=u_4b;u_4b=u_3b;u_3b=u_2b;u_2b=u_1b;u_1b=u_b(k); y_4b=y_3b;y_3b=y_2b;y_2b=y_1b;y_1b=y4out(k); wo_3=wo_2; wo_2=wo_1; wo_1=wo; wi_3=wi_2; wi_2=wi_1; wi_1=wi; error_2=error_1; error_1=error(k); end figure(1); plot(time,rin,'r',time,yout,'b'); xlabel('time(s)');ylabel('rin,yout'); figure(2); plot(time,error,'r'); xlabel('time(s)');ylabel('error'); figure(3); plot(time,u,'r'); xlabel('time(s)');ylabel('u'); figure(4); plot(time,u_r,'r'); xlabel('time(s)');ylabel('u_r'); figure(5); %subplot(211); %plot(time,y2out,'r'); %xlabel('time(s)');ylabel('y2out'); %subplot(212); plot(time,y1out,'r'); xlabel('time(s)');ylabel('y1out'); figure(6); subplot(311); plot(time,kp,'r'); xlabel('time(s)');ylabel('kp'); subplot(312); plot(time,ki,'r'); xlabel('time(s)');ylabel('ki'); subplot(313); plot(time,kd,'r'); xlabel('time(s)');ylabel('kd');