PID control library forked from Aaron Berk. Make sure to read articles from www.controlguru.com to properly understand how to instantiate the PID loop.

Fork of PID by Aaron Berk

PID.cpp

Committer:
pHysiX
Date:
2014-05-12
Revision:
2:b55a16b5f05c
Parent:
0:6e12a3e5af19

File content as of revision 2:b55a16b5f05c:

/**
 * @author Aaron Berk
 *
 * @section LICENSE
 *
 * Copyright (c) 2010 ARM Limited
 *
 * Permission is hereby granted, free of charge, to any person obtaining a copy
 * of this software and associated documentation files (the "Software"), to deal
 * in the Software without restriction, including without limitation the rights
 * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
 * copies of the Software, and to permit persons to whom the Software is
 * furnished to do so, subject to the following conditions:
 *
 * The above copyright notice and this permission notice shall be included in
 * all copies or substantial portions of the Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
 * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
 * THE SOFTWARE.
 *
 * @section DESCRIPTION
 *
 * A PID controller is a widely used feedback controller commonly found in
 * industry.
 *
 * This library is a port of Brett Beauregard's Arduino PID library:
 *
 *  http://www.arduino.cc/playground/Code/PIDLibrary
 *
 * The wikipedia article on PID controllers is a good place to start on
 * understanding how they work:
 *
 *  http://en.wikipedia.org/wiki/PID_controller
 *
 * For a clear and elegant explanation of how to implement and tune a
 * controller, the controlguru website by Douglas J. Cooper (who also happened
 * to be Brett's controls professor) is an excellent reference:
 *
 *  http://www.controlguru.com/
 */

/**
 * Includes
 */
#include "PID.h"

PID::PID(float Kc, float tauI, float tauD, float interval)
{

    usingFeedForward = false;
    inAuto           = false;

    //Default the limits to the full range of I/O: 3.3V
    //Make sure to set these to more appropriate limits for
    //your application.
    setInputLimits(0.0, 3.3);
    setOutputLimits(0.0, 3.3);

    tSample_ = interval;

    setTunings(Kc, tauI, tauD);

    setPoint_             = 0.0;
    processVariable_      = 0.0;
    prevProcessVariable_  = 0.0;
    controllerOutput_     = 0.0;
    prevControllerOutput_ = 0.0;

    accError_ = 0.0;
    bias_     = 0.0;

    realOutput_ = 0.0;

}

void PID::setInputLimits(float inMin, float inMax)
{

    //Make sure we haven't been given impossible values.
    if (inMin >= inMax) {
        return;
    }

    //Rescale the working variables to reflect the changes.
    prevProcessVariable_ *= (inMax - inMin) / inSpan_;
    accError_            *= (inMax - inMin) / inSpan_;

    //Make sure the working variables are within the new limits.
    if (prevProcessVariable_ > 1) {
        prevProcessVariable_ = 1;
    } else if (prevProcessVariable_ < 0) {
        prevProcessVariable_ = 0;
    }

    inMin_  = inMin;
    inMax_  = inMax;
    inSpan_ = inMax - inMin;

}

void PID::setOutputLimits(float outMin, float outMax)
{

    //Make sure we haven't been given impossible values.
    if (outMin >= outMax) {
        return;
    }

    //Rescale the working variables to reflect the changes.
    prevControllerOutput_ *= (outMax - outMin) / outSpan_;

    //Make sure the working variables are within the new limits.
    if (prevControllerOutput_ > 1) {
        prevControllerOutput_ = 1;
    } else if (prevControllerOutput_ < 0) {
        prevControllerOutput_ = 0;
    }

    outMin_  = outMin;
    outMax_  = outMax;
    outSpan_ = outMax - outMin;

}

void PID::setTunings(float Kc, float tauI, float tauD)
{

    //Verify that the tunings make sense.
    if (Kc == 0.0 || tauI < 0.0 || tauD < 0.0) {
        return;
    }

    //Store raw values to hand back to user on request.
    pParam_ = Kc;
    iParam_ = tauI;
    dParam_ = tauD;

    float tempTauR;

    if (tauI == 0.0) {
        tempTauR = 0.0;
    } else {
        tempTauR = (1.0 / tauI) * tSample_;
    }

    //For "bumpless transfer" we need to rescale the accumulated error.
    if (inAuto) {
        if (tempTauR == 0.0) {
            accError_ = 0.0;
        } else {
            accError_ *= (Kc_ * tauR_) / (Kc * tempTauR);
        }
    }

    Kc_   = Kc;
    tauR_ = tempTauR;
    tauD_ = tauD / tSample_;

}

void PID::setKP(float KP)
{
    Kc_ = KP;
}

void PID::reset(void)
{

    float scaledBias = 0.0;

    if (usingFeedForward) {
        scaledBias = (bias_ - outMin_) / outSpan_;
    } else {
        scaledBias = (realOutput_ - outMin_) / outSpan_;
    }

    prevControllerOutput_ = scaledBias;
    prevProcessVariable_  = (processVariable_ - inMin_) / inSpan_;

    //Clear any error in the integral.
    accError_ = 0;

}

void PID::setMode(int mode)
{

    //We were in manual, and we just got set to auto.
    //Reset the controller internals.
    if (mode != 0 && !inAuto) {
        reset();
    }

    inAuto = (mode != 0);

}

void PID::setInterval(float interval)
{

    if (interval > 0) {
        //Convert the time-based tunings to reflect this change.
        tauR_     *= (interval / tSample_);
        accError_ *= (tSample_ / interval);
        tauD_     *= (interval / tSample_);
        tSample_   = interval;
    }

}

void PID::setSetPoint(float sp)
{

    setPoint_ = sp;

}

void PID::setProcessValue(float pv)
{

    processVariable_ = pv;

}

void PID::setBias(float bias)
{

    bias_ = bias;
    usingFeedForward = 1;

}

float PID::compute()
{

    //Pull in the input and setpoint, and scale them into percent span.
    float scaledPV = (processVariable_ - inMin_) / inSpan_;

    if (scaledPV > 1.0) {
        scaledPV = 1.0;
    } else if (scaledPV < 0.0) {
        scaledPV = 0.0;
    }

    float scaledSP = (setPoint_ - inMin_) / inSpan_;
    if (scaledSP > 1.0) {
        scaledSP = 1;
    } else if (scaledSP < 0.0) {
        scaledSP = 0;
    }

    float error = scaledSP - scaledPV;

    //Check and see if the output is pegged at a limit and only
    //integrate if it is not. This is to prevent reset-windup.
    if (!(prevControllerOutput_ >= 1 && error > 0) && !(prevControllerOutput_ <= 0 && error < 0)) {
        accError_ += error;
    }

    //Compute the current slope of the input signal.
    float dMeas = (scaledPV - prevProcessVariable_) / tSample_;

    float scaledBias = 0.0;

    if (usingFeedForward) {
        scaledBias = (bias_ - outMin_) / outSpan_;
    }

    //Perform the PID calculation.
    controllerOutput_ = scaledBias + Kc_ * (error + (tauR_ * accError_) - (tauD_ * dMeas));

    //Make sure the computed output is within output constraints.
    if (controllerOutput_ < 0.0) {
        controllerOutput_ = 0.0;
    } else if (controllerOutput_ > 1.0) {
        controllerOutput_ = 1.0;
    }

    //Remember this output for the windup check next time.
    prevControllerOutput_ = controllerOutput_;
    //Remember the input for the derivative calculation next time.
    prevProcessVariable_  = scaledPV;

    //Scale the output from percent span back out to a real world number.
    return ((controllerOutput_ * outSpan_) + outMin_);

}

float PID::getInMin()
{

    return inMin_;

}

float PID::getInMax()
{

    return inMax_;

}

float PID::getOutMin()
{

    return outMin_;

}

float PID::getOutMax()
{

    return outMax_;

}

float PID::getInterval()
{

    return tSample_;

}

float PID::getPParam()
{

    return pParam_;

}

float PID::getIParam()
{

    return iParam_;

}

float PID::getDParam()
{

    return dParam_;

}