lib/com.irurueta.android.navigation.inertial.processors.attitude/KalmanGyroscopePredictor

KalmanGyroscopePredictor

class KalmanGyroscopePredictor(phiMatrixMethod: PhiMatrixMethod = PhiMatrixMethod.APPROXIMATED, processNoiseCovarianceMethod: ProcessNoiseCovarianceMethod = ProcessNoiseCovarianceMethod.BETTER, quaternionStepIntegratorType: QuaternionStepIntegratorType = QuaternionStepIntegratorType.RUNGE_KUTTA, val gyroscopeNoisePsd: Double = DEFAULT_GYROSCOPE_NOISE_PSD)

Predicts next Kalman filter state based on current gyroscope measurement.

Parameters

phiMatrixMethod

method to estimate phi matrix relating previous and current filter state.

processNoiseCovarianceMethod

method to estimate process noise covariance matrix.

quaternionStepIntegratorType

method to integrate quaternions on each measurement.

Constructors

KalmanGyroscopePredictor

constructor(phiMatrixMethod: PhiMatrixMethod = PhiMatrixMethod.APPROXIMATED, processNoiseCovarianceMethod: ProcessNoiseCovarianceMethod = ProcessNoiseCovarianceMethod.BETTER, quaternionStepIntegratorType: QuaternionStepIntegratorType = QuaternionStepIntegratorType.RUNGE_KUTTA, gyroscopeNoisePsd: Double = DEFAULT_GYROSCOPE_NOISE_PSD)

Types

Companion

object Companion

Properties

gyroscopeNoisePsd

val gyroscopeNoisePsd: Double

gyroscope noise PSD.

kalmanErrorCovPostPredicted

val kalmanErrorCovPostPredicted: Matrix

Kalman filter error covariance matrix after prediction.

kalmanErrorCovPrePredicted

var kalmanErrorCovPrePredicted: Matrix

Kalman filter error covariance matrix before prediction.

kalmanStatePostPredicted

val kalmanStatePostPredicted: Matrix

Kalman filter state after prediction.

kalmanStatePrePredicted

var kalmanStatePrePredicted: Matrix

Kalman filter state before prediction.

nedBodyAngularSpeedTriad

var nedBodyAngularSpeedTriad: AngularSpeedTriad

Contains gyroscope data expressed in NED coordinates respect to body.

predictedBodyQ

val predictedBodyQ: Quaternion

Contains predicted attitude respect to body for next step. Because rotation is defined respect to body, rotation converts from body coordinates to leveled local NED coordinates respect to Earth: pn = previousBodyQ * pb

previousBodyQ

var previousBodyQ: Quaternion

Previous step attitude respect to body. Because rotation is defined respect to body, rotation converts from body coordinates to leveled local NED coordinates respect to Earth: pn = previousBodyQ * pb

val q: Matrix

Continuous time process noise covariance matrix.

qba

val qba: Matrix

Covariance noise matrix of the accelerometer bias

qbg

val qbg: Matrix

Covariance noise matrix of the gyroscope bias

val rg: Matrix

Covariance measurement matrix of the gyroscope measurements.

timeIntervalSeconds

var timeIntervalSeconds: Double

Time interval between measurements expressed in seconds (s).

Functions

predict

fun predict()

Predicts expected orientation, orientation error, gyroscope and accelerometer biases using current gyroscope measurements and current orientation estimation.

reset

fun reset()

Resets internal parameter.