matlab-使用Ziegler-Nichols闭环计算方法计算PID控制器的kp,ki,kd三个参数-源码

clc; clear; close all; warning off; addpath(genpath(pwd)); %% Transfer Function Definition %This is just an example D = 0.5; E = 48; R = 10; C = 100e-6; L = 100e-6; Vc = E/(1-D); IL = E/(R*(1-D)^2); A = [0 , -(1-D)/L ; (1-D)/C , -1/(R*C)]; B = [Vc/L ; -IL/C]; C = [0 , 1]; D = 0; s = tf('s'); TF = tf(ss(A,B,C,D)); %% Ku Ultimate Gain (WORK IN PROGRESS) %Still need to make sure that the Ku used is the correct one (as just %finding Ku this way might not be the only option) syms K CE = cell2mat(TF.Denominator) + K*cell2mat(TF.Numerator);%Characteristic Equation nonzero = find(cell2mat(TF.Numerator)~=0); Kvalues = zeros(1,length(CE)); for i = nonzero Kvalues(i) = solve(CE(i),K); end Ku = Kvalues(Kvalues> 0 & min(Kvalues));%The selection of this Ku value still needs to be checked %% Frecuency and Period System = feedback(Ku*TF,1); Poles = abs(pole(System)); Frecuency = Poles(1); Pu = 2*pi/Frecuency; %% For a P Controller Kp = 0.5*Ku; P = Kp; System = feedback(P*TF,1); %% For a P Controller Kp = 0.5*Ku; P = Kp; System = feedback(P*TF,1); step(48*System,'g') hold on %% For a PI Controller Kp = 0.45*Ku; Ti = Pu/1.2; Ki = Kp/Ti; PI = (Kp + Ki/s); System = feedback(PI*TF,1); step(48*System,'r') %% For a PID Controller Kp = 0.6*Ku; Ti = Pu/2; Ki = Kp/Ti; Td = Pu/8; Kd = Kp*Td; PID = (Kp + s*Kd + Ki/s); System = feedback(PID*TF,1); step(48*System,'b')