电机的PID转矩控制

Question

I'm trying to control the motor torque. This I'm trying to accomplish by implementing a PID control on motor current and hence on the PWM. Higher PWM means higher torque. I'm new to arduino-uno, hence would need help with the coding. I have written a code for the same but a bit unsure on its correctness. I'm still unsure about the value of the 'Integral' variable to be used though. Any help is much appreciated Thanks in advance

#include "RunningAverage.h"
 int m1 = 13;
int m2 = 12;
int me = 9;
int t = millis()+5000;
RunningAverage myRA(80);
int stat=0;
int pwmn=100;
int counter=0;
int kP;
int kI=0;
int kD=0;
int SetPt;
int Last;
int Actual;
int Error;
int Integral;
float P;
float I;
float D;
int Drive;
int ScaleFactor;


void motorRight(){
digitalWrite(m1,HIGH);
digitalWrite(m2, LOW);
}


void motorLeft(){
digitalWrite(m2,HIGH);
digitalWrite(m1, LOW);
}

void motorOff(){
digitalWrite(m2, LOW);
digitalWrite(m1, LOW);
}


void motorBrake(){
digitalWrite(m2, HIGH);
digitalWrite(m1, HIGH);
delay(10);
motorOff();
}

void setup() {
// put your setup code here, to run once:
pinMode(m1,OUTPUT);
pinMode(m2,OUTPUT);
pinMode(me,OUTPUT);
analogWrite(me,pwmn);
Serial.begin(115200);
motorRight();
}

void loop() {
// put your main code here, to run repeatedly:
myRA.addValue(analogRead(A1));
Serial.print(myRA.getAverage());
Serial.println(",500,600");
Actual = myRA.getAverage();
Error = SetPt - Actual;
P = Error*kP; // calc proportional term
I = Integral*kI; // integral term
D = (Last-Actual)*kD; // derivative term
Drive = P + I + D; // Total drive = P+I+D
Drive = Drive*ScaleFactor; // scale Drive to be in the range 0-255
//Serial.println(pwmn);
if(counter>10){
if(pwmn<250){
pwmn++;
}
counter=0;
}
counter++;
if (abs(Drive)>255) {
Drive=255;
}
analogWrite (me,Drive); // send PWM command to motor board
Last = Actual; // save current value for next time
// analogWrite(me,pwmn);
delay(50);
}' 

Answer 1

You are correct, the integral calculation is wrong. Integral (from integrate) means to add up. What you want is

I = I + kI*Error;

You should also limit the Integral term (I) to some maximum and minimum value to avoid "wind-up" error. This happens when the servo cannot reach is goal and since the I term is the sum of the errors it can get out of control.