Package com.marslib.util

Class OnlineFeedforwardEstimator

java.lang.Object

com.marslib.util.OnlineFeedforwardEstimator

public class OnlineFeedforwardEstimator
extends Object

A Zero-GC, realtime Feedforward extractor that continuously monitors mechanism telemetry (Voltage, Velocity, Acceleration) to compute empirical kV and kA values dynamically.

It uses a sliding-window Ordinary Least Squares (OLS) regression algorithm on a 2-variable system (Velocity & Acceleration) against Effective Voltage (Voltage - kS). By operating strictly over periods of contiguous movement (vel > threshold), it avoids static friction nonlinearities.

Students: You can use this to perfectly dial in your Feedforward models without running a formal SysId sequence.

Constructor Summary

Constructors

Constructor	Description
OnlineFeedforwardEstimator(String name, int windowSize, double ksVolts)	Constructs the Online Estimator.

Method Summary

Modifier and Type	Method	Description
void	addMeasurement(double appliedVoltage, double velocity, double acceleration)	Feeds the live telemetry into the sliding regression window.
double	getEstimatedKA()	Retrieves the latest fully-regressed Acceleration Feedforward constant.
double	getEstimatedKV()	Retrieves the latest fully-regressed Velocity Feedforward constant.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Details

OnlineFeedforwardEstimator

public OnlineFeedforwardEstimator(String name,
 int windowSize,
 double ksVolts)

Constructs the Online Estimator.

Parameters:

name - The AdvantageScope logging namespace (e.g. "SwerveDrive" or "Shooter").

windowSize - The number of contiguous 20ms ticks to calculate over (e.g. 500 = 10sec of data).

ksVolts - Theoretical or empirical kS (static friction voltage) to subtract from the model.

Method Details

addMeasurement

public void addMeasurement(double appliedVoltage,
 double velocity,
 double acceleration)

Feeds the live telemetry into the sliding regression window.

Parameters:

appliedVoltage - The total voltage passed to the motor.

velocity - The velocity of the mechanism (m/s or rad/s).

acceleration - The acceleration of the mechanism (m/s^2 or rad/s^2).

getEstimatedKV

public double getEstimatedKV()

Retrieves the latest fully-regressed Velocity Feedforward constant.

getEstimatedKA

public double getEstimatedKA()

Retrieves the latest fully-regressed Acceleration Feedforward constant.