Attitude and Heading Reference Systems Questions and Answers

Answer: C

Correct answer: Only three

Three pairs are right. An attitude and heading reference system outputs attitude and heading; an inertial navigation system adds velocity and position, because it integrates specific force as well as angular rate; and an air data computer outputs altitude, airspeed and Mach number from pressure and temperature.

The fourth is wrong. A magnetometer measures the direction of the magnetic field and so provides heading, not rate of turn. Rate of turn comes from the gyros - specifically, from the vertical component of the angular rate vector.

Answer: D

Correct answer: Only three

Three pairs are right. A complementary filter high-passes the gyro signal and low-passes the reference, summing the two; a Kalman filter weights measurements by estimated uncertainty; and bias estimation means treating the sensor offset as a state to be estimated rather than a fixed calibration constant.

The fourth is wrong. Median voting does not average the three units - it takes the middle value precisely to avoid averaging, since an outlier drags a mean but not a median. Its purpose is fault isolation, not noise reduction.

Answer: A

Correct answer: Position

An attitude and heading reference system produces attitude - pitch and roll - and heading. It does not produce position; that is the distinguishing feature between it and an inertial navigation system.

The reason is error growth. Attitude requires integrating angular rate once and is held bounded by the gravity and magnetic references, whereas position requires integrating specific force twice, so an accelerometer bias produces error growing as \( t^2 \) with nothing available to bound it. A bias of \( 0.001\ \mathrm{m/s^2} \) gives \( 1.8 \) m in a minute and hundreds of metres in an hour, which is why navigation-grade sensors and an external position fix are needed for that job.

Answer: B

Correct answer: A pitot tube

A pitot tube belongs to the air data system, not to a strapdown inertial unit. It senses total pressure, which with static pressure gives airspeed - an aerodynamic quantity, not an inertial one.

The three sensors that do make up a typical unit are the rate gyro triad, which measures angular rate about three axes; the accelerometer triad, which measures specific force along three axes; and the magnetometer, which supplies the absolute heading reference the other two cannot provide. The unit's outputs are attitude and heading, computed entirely from those.

Answer: C

Correct answer: Magnetic variation

Magnetic variation is the angle between true and magnetic north at a location - a property of the earth's field, not of any sensor. It is looked up from a stored model and applied as a correction; an instrument with no error at all would still need it.

The other three are genuine inertial sensor error mechanisms: bias, a non-zero output when stationary; angle random walk, the integrated effect of the sensor's white noise; and scale factor error, an error in gain that is proportional to the rate being measured.

Answer: D

Correct answer: The accelerometer triad

The accelerometer triad cannot establish heading. It senses the gravity vector, which is vertical, so it fixes pitch and roll but says nothing about direction in the horizontal plane - rotating the aircraft about its vertical axis leaves every accelerometer reading unchanged.

The other three all can. A magnetometer gives magnetic heading from the horizontal field component; gyro compassing finds true heading from the horizontal component of earth rate, given gyros quiet enough to resolve \( 15\cos\phi \) degrees per hour; and on a known heading the value can simply be entered.

Answer: A

Correct answer: Median voting between three units

Median voting is a redundancy management technique, not a fusion technique. It compares outputs from separate, complete units to decide which to believe, and its purpose is to survive a failure - it does not combine dissimilar sensors to improve an estimate.

The other three belong together. Complementary filtering blends the gyros with the accelerometers and magnetometer across a fixed crossover; Kalman filtering does the same adaptively; and weighting by estimated uncertainty is the principle on which the Kalman filter operates. Fusion exploits complementary error characteristics; voting exploits replication.

Answer: B

Correct answer: The acceleration was sensed as a tilt

This is the acceleration-versus-gravity ambiguity. Accelerometers measure specific force and cannot separate the aircraft's own acceleration from the reaction to gravity, so during a sustained take-off acceleration the sensed vector tilts backward by \( \arctan(a/g) \) and an unfiltered solution reads that as a pitch-up. Once the aircraft settles at a steady climb speed the acceleration ceases and the false tilt disappears - exactly the behaviour described.

A gyro failure or lost alignment would not resolve itself when the acceleration stopped, and a magnetometer disturbance would affect heading rather than pitch. The remedy is to reduce the filter's trust in the accelerometers while acceleration is detected.

Answer: C

Levelling depends on the only specific force being the reaction to gravity, which taxiing destroys. Everything computed afterwards is a change from this initial state, so a corrupted alignment would persist all flight.

Answer: D

Correct answer: Excluded the outlier by median voting

The three readings are \( 088.0, 088.3 \) and \( 093.5 \) degrees. Two agree closely and the third is more than five degrees away, so median voting takes the middle value, \( 088.3 \), and the deviant unit is excluded automatically - which is why operation continues with no crew action.

Averaging would give about \( 089.9 \), nearly two degrees from where the healthy units agree, and would conceal the fault rather than isolate it. Selecting the highest value would select the faulty unit. Reversion to standby instruments is unnecessary while two good units remain.