Classifications

• • G—PHYSICS
  • G01—MEASURING; TESTING
  • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
  • G01C19/00—Gyroscopes; Turn-sensitive devices using vibrating masses; Turn-sensitive devices without moving masses; Measuring angular rate using gyroscopic effects
  • G01C19/56—Turn-sensitive devices using vibrating masses, e.g. vibratory angular rate sensors based on Coriolis forces
• • G—PHYSICS
  • G01—MEASURING; TESTING
  • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
  • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
  • G01C21/10—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration
  • G01C21/12—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration executed aboard the object being navigated; Dead reckoning
  • G01C21/16—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration executed aboard the object being navigated; Dead reckoning by integrating acceleration or speed, i.e. inertial navigation
  • G01C21/165—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration executed aboard the object being navigated; Dead reckoning by integrating acceleration or speed, i.e. inertial navigation combined with non-inertial navigation instruments
• • G—PHYSICS
  • G01—MEASURING; TESTING
  • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
  • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
  • G01C21/10—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration
  • G01C21/12—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration executed aboard the object being navigated; Dead reckoning
  • G01C21/16—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration executed aboard the object being navigated; Dead reckoning by integrating acceleration or speed, i.e. inertial navigation
  • G01C21/183—Compensation of inertial measurements, e.g. for temperature effects
  • G01C21/188—Compensation of inertial measurements, e.g. for temperature effects for accumulated errors, e.g. by coupling inertial systems with absolute positioning systems
• • G—PHYSICS
  • G01—MEASURING; TESTING
  • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
  • G01C25/00—Manufacturing, calibrating, cleaning, or repairing instruments or devices referred to in the other groups of this subclass
  • G01C25/005—Manufacturing, calibrating, cleaning, or repairing instruments or devices referred to in the other groups of this subclass initial alignment, calibration or starting-up of inertial devices
• • G—PHYSICS
  • G01—MEASURING; TESTING
  • G01P—MEASURING LINEAR OR ANGULAR SPEED, ACCELERATION, DECELERATION, OR SHOCK; INDICATING PRESENCE, ABSENCE, OR DIRECTION, OF MOVEMENT
  • G01P21/00—Testing or calibrating of apparatus or devices covered by the preceding groups

Definitions

• the present invention relates to sensors, and in particular to the compensation for vibration-induced error affecting the operation of such sensors.
• Inertial sensors such as gyroscopes and accelerometers, are often adversely affected by vibrations. They may be very sensitive to such vibrations. In fact, vibration sensitivity can often be a dominant performance limiting error source. Many inertial sensors rely on micro-electromechanical sensor (MEMS) devices which utilize vibratory mechanisms for rate/acceleration sensing. These types of sensors have shown evidence of significant vibration sensitivity problems.
• MEMS micro-electromechanical sensor
• Inertial sensors are often used in environments that inherently subject the sensors to significant vibrations. These include use in aircraft and weapons, where the power plants and environment serve as significant sources of vibration.
• a commonly used resonant beam accelerometer that is often viewed as a one milli-g accuracy sensor can exhibit bias shifts on the order of several milli-g' s under typical operational vibration environments.
• the most common effect is a slowly varying low frequency error component which changes as a function of the applied vibration spectrum.
• the effect of the vibration induced bias shift can be trimmed or estimated such as by use of a Kalman filter.
• the vibration level or spectrum shape changes significantly, the vibration-induced error also changes, diminishing the value of any previously obtained calibration.
• an in-flight transfer alignment and calibration procedure is usually performed prior to release of the guided munition munitions.
• velocity differences (or related quantities) between the aircraft INS and the weapon INS are processed by a Kalman filter to initialize the attitude and heading of the weapon INS, and to estimate some of its inertial sensor errors, typically including the gyro and accelerometer bias error.
• the vibration spectrum present at the weapon inertial measurement unit is strongly driven by its captive-carry environment, such as a wing-store station or weapon bay.
• the weapon's inertial sensor bias errors will be affected by the vibration environment.
• gyro and accelerometer bias shifts due to a change in the vibration-induced errors.
• this change can negate any benefit of the pre-launch weapon inertial measurement unit calibration. There is a need to minimize the detrimental effects of such changes in vibration environment.
• a factory calibration procedure is performed by subjecting a sensor to suitable vibration stimulus, thereby exposing sensitivity coefficients relating vibration power levels at various frequencies to the bias-like error of the sensor.
• accelerometer samples are collected and used to characterize the actual vibration environment, specifically, the amount of vibration energy in various portions of the frequency spectrum.
• Predicted sensor vibration-induced error is calculated based on the measured vibration spectrum and the previously-obtained factory calibration coefficients.
• the predicted vibration-induced error is then applied as a compensation to the sensor output.
• high rate vibration measurement samples from the accelerometers are provided as the input to a power spectral density estimator.
• vibration power spectrum estimates are then combined with previously obtained calibration coefficients to generate real time bias correction.
• Vibration bias sensitivity coefficients that are estimated in a controlled environment to ensure their accuracy to compensate the sensor in real-time applications.
• the high data rate vibration samples are obtained using acceleration measurement devices such as accelerometers mounted to a rigid sensor assembly shared with the sensor to be compensated. Vibration bias estimates are then combined with outputs from both a plurality (typically three or more) of gyros and a plurality of accelerometers, which combined form an inertial measurement unit.
• acceleration measurement devices such as accelerometers mounted to a rigid sensor assembly shared with the sensor to be compensated.
• Vibration bias estimates are then combined with outputs from both a plurality (typically three or more) of gyros and a plurality of accelerometers, which combined form an inertial measurement unit.
• a digital signal processor is used to perform calculations to obtain the vibration power spectral density and then provide the sensor vibration-induced error estimates in real time at a high rate. Still further embodiments utilize software to perform the calculations, or further combinations of hardware, software and firmware.
• Figure 1 is a block diagram of a mechanism for estimating and correcting a sensor for vibration induced error.
• Figure 2 is a graph showing a vibration-induced error model for a simplified example.
• Figure 3 is a flowchart representing a computer implementable method for estimating and correcting a sensor for vibration induced error.
• Figure 4 is a block diagram showing the estimation of coefficients given reference accelerations. Detailed Description
• FIG. 1 a block diagram showing a mechanism for estimating and correcting for vibration induced errors is indicated generally at 110.
• a housing or other mounting device 120 in one embodiment contains a micro- electromechanical sensor (MEMS) device 124, for example, an inertial measurement unit (IMU).
• MEMS micro- electromechanical sensor
• IMU inertial measurement unit
• This type of sensor relies on vibratory mechanisms for rate/acceleration sensing.
• the sensor is a Coriolis-type vibratory gyro in one embodiment which provides a first output representative of directional changes.
• An accelerometer 126 is also mounted to housing 120 to provide a second output representative of acceleration.
• the second output is provided to a vibration power spectral density (PSD) estimator 130, which provides estimations of the power spectral densities of vibrations as a function of high frequency accelerometer 126 measurements.
• PSD vibration power spectral density
• the frequency of such measurements is at least twice the highest expected frequency of vibration to ensure sufficient accuracy of measurements.
• Sample rates may be on the order of 5KHz or higher.
• the PSD estimates are formed at a relatively low rate, for example 1 Hz, and are provided to a compensation module 134, which utilizes calibration coefficients 140 to provides vibration-induced bias estimates or compensation values, both for the gyroscope first output and the accelerometer second output.
• An adder 150 receives the first output, and a gyro vibration-induced bias estimate and provides a compensated gyroscope output on a line 155.
• An adder 160 receives the second output and an accelerometer vibration-induced bias estimate from module 134, and provides a compensated accelerometer output on a line 165. To ensure that the vibration measurements correspond to the sensor being compensated, the components are mounted to a rigid sensor assembly indicated at 170 and shared with the sensor to be compensated.
• vibration bias f(S ⁇ ( ⁇ ), Syy ( ⁇ ), S zz ( ⁇ )) where S xx ( ⁇ ), S yy ( ⁇ ), and S zz ( ⁇ ) represent the vibration power spectral densities in the sensor's x, y and z axes.
• these accelerometer axes are those that are customarily referred to as the input axis, output axis, and pendulous axis of the accelerometer.
• an accelerometer exhibits a sensitivity to input axis vibration, a common characteristic of devices with non-negligible scale factor nonlinearity.
• the vibration spectrum can be broken up into a set of frequency bins, and the sensitivity of the bias error to the vibration power within each bin is defined as illustrated in Figure 2, which is intended as a typical graph of the spectral density versus frequency.
• the vibration-induced bias is then expressed as a sum over the frequency bins: where gj 2 represents the vibration power (mean squared acceleration) in the i th frequency bin, and c, is the vibration-induced bias error sensitivity coefficient for that bin.
• gj 2 represents the vibration power (mean squared acceleration) in the i th frequency bin
• c is the vibration-induced bias error sensitivity coefficient for that bin.
• the coefficients are determined for a small number of sensors referred to as a family of sensors, and the typical coefficient values are applied as compensation coefficients for all sensors of the same type.
• the calibration process may thus be applied to a single manufacturing batch of sensors, or any other family quantum of sensors that are likely to be similarly affected by vibration. This eliminates the need to perform extensive factory calibration procedures on every production sensor, but may retain most of the accuracy of the vibration-induced error compensation.
• the frequency bins are of equal size and cover the frequency range at which vibration power may significantly affect the sensors.
• the frequency span of the bin is adjusted such that in frequency ranges where vibrations are most likely to affect the sensors, more bins are concentrated, and in other less significant frequency ranges, fewer bins are used.
• the bins are coupled to a bank of bandpass filters, each covering one the frequency bins. The mean squared acceleration at the output of each of the bandpass filters is used as a measure of the vibration power in each bin. A sine sweep vibration profile is used as test stimulus to provide observability of the model coefficients. The center frequency of the narrowband sinusoidal vibration stimulus is swept over the range of vibration frequency typically encountered by the sensor in its intended application.
• a block diagram flow chart is shown in Figure 3.
• Measured accelerations (dv/dt) for x, y and z axes are provided at 310 and sampled in each of the respective axes at 312, 314 and 316 over a desired time interval. Respective power spectral densities are then computed at 322, 324 and 326. For each axis, the vibration power is then computed for each desired bin at 332, 334 and 336. There are N bins shown, and the actual number of bins varies depending upon the desired accuracy of correction and computational power available. The vibration induced errors are calculated at 340 with predetermined vibration sensitivity coefficients provided at 345.
• FIG. 4 is a block diagram showing calibrating of a vibration-induced error model for a triad of accelerometers, including the dependence of each sensor's error as a function of vibration power in a set of three principal axes that are nominally aligned with the accelerometer's input axes. Measured accelerations in each axis are provided at 410 in response to actual reference or truth accelerations provided by a current test condition at 412. A difference for each axis is provided by adders 416, 418 and 420, which are provided to a vibration sensitivity coefficient estimator 430.
• the measured accelerations 410 are also provided to a buffer 435, which samples the accelerations in each axis over a predetermined calibration time at a rate at least twice the highest frequency desired to be sampled in order to obtain valid sampling data. Vibration power spectral densities are then computed at 440 along each axis and then the vibration power in each of N frequency bins is computed at 445. These values are then provided to the estimator 430.
• a test condition is defined by the axis of the applied vibration stimulus and the frequency bin in which most of the vibration energy is concentrated.
• a model for the vibration-induced error on the jth test condition is
• the measurement vector represents the vibration-induced acceleration errors present in the three accelerometer's outputs at each test condition.
• the vector of unknowns (c) represents the vibration sensitivity coefficients to be estimated.
• the matrix H is comprised of the measured vibration power levels in the three principal axes and the 3 N vibration frequency bins.
• vibration sensitivity coefficient estimates are also used to generate the vibration sensitivity coefficient estimates in different embodiments, including a weighted least squares estimator, maximum likelihood type estimators, and recursive estimators such as the Kalman filter.
• a global position system (GPS) aided inertial navigation system provides accurate measurements of position and velocity (or equivalently, pseudorange and/or deltarange for a plurality of GPS satellites).
• the GPS measurements expose the position and velocity errors of the navigation solution formed using the EMU gyroscopes and accelerometers, which are affected by host vehicle vibration.
• Processing the GPS measurements and the measured vibration spectrum in, for example, a GPS/inertial Kalman filter algorithm provides observability of the vibration-induced error sensitivity coefficients.
• the sensor is affected by only a single component of the three dimensional vibration stimulus
• a sensor may be influenced by the vibration spectra acting on all three axes.
• This three-dimensional vibration stimulus would be characterized by the vibration PSDs in the sensor's three principal coordinate axes, and by the associated cross power spectra.
• the vibration power spectra estimation function is performed using software algorithms hosted in a general purpose microprocessor.
• An alternative embodiment uses a dedicated digital signal processor type hardware device as a coprocessor tasked with generating the PSD estimates. Still further embodiments will be apparent.
• sets of calibration coefficients are provided for selected temperature ranges for sensors which exhibit temperature-dependent vibration sensitivity coefficients.
• This embodiment includes a temperature sensor located proximate to the sensor being compensated. An output of the temperature sensor is used by the compensation module to obtain the correct set of compensation coefficients.

Landscapes

• Engineering & Computer Science (AREA)
• Radar, Positioning & Navigation (AREA)
• Remote Sensing (AREA)
• Physics & Mathematics (AREA)
• General Physics & Mathematics (AREA)
• Automation & Control Theory (AREA)
• Manufacturing & Machinery (AREA)
• Navigation (AREA)
• Gyroscopes (AREA)
• Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
• Vibration Prevention Devices (AREA)
• Indication And Recording Devices For Special Purposes And Tariff Metering Devices (AREA)
• Other Investigation Or Analysis Of Materials By Electrical Means (AREA)

Priority Applications (5)

Applications Claiming Priority (2)

Publications (1)

Family

ID=23445187

Family Applications (1)

Country Status (7)

Cited By (4)

Families Citing this family (48)

Citations (2)

Family Cites Families (3)

• 1999
  • 1999-08-04 US US09/366,926 patent/US6498996B1/en not_active Expired - Lifetime
• 2000
  • 2000-08-04 AT AT00964899T patent/ATE287524T1/de not_active IP Right Cessation
  • 2000-08-04 EP EP04077186A patent/EP1471328B1/en not_active Expired - Lifetime
  • 2000-08-04 AT AT04077186T patent/ATE423959T1/de not_active IP Right Cessation
  • 2000-08-04 DE DE60041667T patent/DE60041667D1/de not_active Expired - Lifetime
  • 2000-08-04 DE DE60017569T patent/DE60017569T2/de not_active Expired - Lifetime
  • 2000-08-04 WO PCT/US2000/021396 patent/WO2001011318A1/en not_active Ceased
  • 2000-08-04 EP EP00964899A patent/EP1200802B1/en not_active Expired - Lifetime
  • 2000-08-04 JP JP2001515927A patent/JP2003506702A/ja active Pending
  • 2000-08-04 CA CA2381196A patent/CA2381196C/en not_active Expired - Fee Related

Patent Citations (2)

Also Published As

Similar Documents

Legal Events