智能控制作业--专家pid

PID算法：就是“比例（proportional）、积分（integral）、微分（derivative）”，是一种常见的“保持稳定”控制算法。

import  numpy as np
import matplotlib.pyplot as plt
import control.matlab as ctl

def ndT(num,den):
    ret1 = (num[0])[0]
    ret2 = (den[0])[0]
    return ret1,ret2

ts = 0.001
sys = ctl.tf(5.235e005,[1,87.35,1.047e4,0])
dsys = ctl.c2d(sys,ts,'zoh')
num,den = ctl.tfdata(dsys)
num,den = ndT(num,den) #Matlab 和 python 的差别

#变量初始化
u_1, u_2,u_3  = 0, 0, 0
y_1, y_2,y_3  = 0, 0, 0
x = [0,0,0] # 3行1列 对应着课上的 当前偏差e(k),偏差增量 △e(k) 和 △e(k - 1)
x2_1 = 0

#比例 积分 微分 参数的初始值的设定
kp = 0.6
ki = 0.03
kd = 0.01

error_1 = 0

#控制算法的循环过程
i = 0
cycle = 500  #总周期为 500次的采样
r = [0 for i in range(cycle)]
u = [0 for i in range(cycle)]
y = [0 for i in range(cycle)]
error =[0 for i in range(cycle)]
time = np.linspace(0,2,cycle)

for k in range(500):


    u[k] = kp * x[0] + kd * x[1] + ki * x[2] #PID控制器的输出
    r[k] = 1
    #专家控制部分
    #规则1，效果为开环
    if abs(x[0]) > 0.8:
        u[k] = 0.45
    elif abs(x[0]) > 0.4:
        u[k] = 0.4
    elif abs(x[0]) > 0.2:
        u[k] = 0.12
    elif abs(x[0]) > 0.01:
        u[k] = 0.10

    # 规则2 偏差往绝对值增大方向变化或为某一固定值，根据与中间值的关系进行判断
    if x[0] * x[1] > 0 or (x[1] == 0):
        if abs(x[0]) >= 0.05:
            u[k] = u_1 + 2 * kp * x[0]
        else:
            u[k] = u_1 + 0.4 * kp * x[0]


    # 规则3 偏差往绝对值减小方向变化，或已达到平衡，保存控制器输出维持不变
    if (x[0] * x[1] < 0 and x[1] * x2_1 > 0) or x[0] == 0:
        u[k] = u[k]


    # 规则4 偏差处于极限状态，根据中间值进行判断
    if x[0] * x[1] < 0 and x[1] * x2_1 < 0:
        if abs(x[0]) >= 0.05:
            u[k] = u_1+2*kp*error_1
        else:
            u[k] = u_1 + 0.6 * kp * error_1


    # 规则5 偏差小于极小值
    if abs(x[0]) <= 0.001:
        u[k] = 0.5*x[0]+0.01*x[2]

    #对控制器输出进行限制
    if u[k] >= 10:
        u[k] = 10

    if u[k] <= -10:
        u[k] = -10

    y[k]= -den[1]*y_1-den[2]*y_2-den[3]*y_3+0*u[k]+num[0]*u_1+num[1]*u_2+num[2]*u_3;


    error[k] = r[k] - y[k]

    u_3 = u_2
    u_2 = u_1
    u_1 = u[k]

    y_3 = y_2
    y_2 = y_1
    y_1 = y[k]

    x[0] = error[k]
    x2_1 = x[1]
    x[1] = (error[k] - error_1) / ts
    x[2] = x[2] + error[k] * ts

    error_1 = error[k]

plt.figure('智能控制作业1 -- 专家控制PID')
plt.plot(time,y)
plt.plot(time,r)

plt.show()

import  numpy as np
import matplotlib.pyplot as plt
import control.matlab as ctl

def ndT(num,den):
    ret1 = (num[0])[0]
    ret2 = (den[0])[0]
    return ret1,ret2

ts = 0.001
sys = ctl.tf(5.235e005,[1,87.35,1.047e4,0])
dsys = ctl.c2d(sys,ts,'zoh')
num,den = ctl.tfdata(dsys)
num,den = ndT(num,den) #Matlab 和 python 的差别

#变量初始化
u_1, u_2,u_3  = 0, 0, 0
y_1, y_2,y_3  = 0, 0, 0
x = [0,0,0] # 3行1列 对应着课上的 当前偏差e(k),偏差增量 △e(k) 和 △e(k - 1)
x2_1 = 0

#比例 积分 微分 参数的初始值的设定
kp = 0.6
ki = 0.03
kd = 0.01

error_1 = 0

#控制算法的循环过程
i = 0
cycle = 500  #总周期为 500次的采样
r = [0 for i in range(cycle)]
u = [0 for i in range(cycle)]
y = [0 for i in range(cycle)]
error =[0 for i in range(cycle)]
time = np.linspace(0,2,cycle)

for k in range(500):


    u[k] = kp * x[0] + kd * x[1] + ki * x[2] #PID控制器的输出
    r[k] = 1
    
    #对控制器输出进行限制
    if u[k] >= 10:
        u[k] = 10

    if u[k] <= -10:
        u[k] = -10

    y[k]= -den[1]*y_1-den[2]*y_2-den[3]*y_3+0*u[k]+num[0]*u_1+num[1]*u_2+num[2]*u_3;


    error[k] = r[k] - y[k]

    u_3 = u_2
    u_2 = u_1
    u_1 = u[k]

    y_3 = y_2
    y_2 = y_1
    y_1 = y[k]

    x[0] = error[k]
    x2_1 = x[1]
    x[1] = (error[k] - error_1) / ts
    x[2] = x[2] + error[k] * ts

    error_1 = error[k]

plt.figure('智能控制作业1 -- 纯PID')
plt.plot(time,y)
plt.plot(time,r)

plt.show()

Matlab

clc;
ts=0.001;
sys=tf(5.235e005,[1,87.35,1.047e4,0]);     %Process TF
dsys=c2d(sys,ts,'z');
[num, den]=tfdata(dsys,'v');
disp(num)
disp(den)

 num 的值在python中是比Matlab 里少一个，少的就是第一个的0。