Processor for an inertial measurement unit

• US 4,222,272 A
• Filed: 03/27/1978
• Issued: 09/16/1980
• Est. Priority Date: 03/27/1978
• Status: Expired due to Term

• Current Assignee: Sanders Associates, INC. (BAE Systems Plc)
• Original Assignee: Sanders Associates, INC. (BAE Systems Plc)
• Inventors: Mairson, Theodore
• Primary Examiner(s): Myracle, Jerry W.
• Application Number:
  US05/890,354
• Time in Patent Office: 904 Days
• Field of Search: 73/510, 73/514, 73/515, 73/516 R, 73/516 LM, 73/517 R, 73/517 B, 73/178 R, 364/453, 364/565, 364/566, 364/720, 364/454
• US Class Current: 73/504.03
• CPC Class Codes: G01P 15/18 in two or more dimensions
• CPC Class Codes: G01P 3/44 for measuring angular speed G01P3/56 takes precedence

Abstract

A processor is provided for an inertial measurement unit which computes its angular velocity and translational acceleration in terms of dynamic variables chosen because they uniquely define the motion of the inertial measurement unit in terms of linear combinations of the outputs of the transducers used in the inertial measurement unit. In one embodiment, the processor includes a demodulator for obtaining d.c., in-phase and quadrature sets of signals from the outputs of the transducers, a combining circuit for deriving some of the dynamic variables from the in-phase and quadrature sets of signals, and a microprocessor for transforming the dynamic variables into angular velocity and translational acceleration vectors. The processor is used with a small rugged, precise inertial measurement unit in which vector components are measured by the use of accelerometers fixed in a spinning rotor, at least one of which is off-set from the rotor axis and has a sensitive axis parallel to that of the rotor. This off-set/parallel accelerometer permits obtaining the signs of the angular velocities in addition to their magnitudes such that complete vector components are obtained. In an alternative embodiment, two orthogonally oriented rotor systems are utilized which permits all a.c. signal processing, thereby eliminating the necessity of d.c. measurements. Error isolation and correction are easily accomplished in a specialized combining circuit which simplifies initial alignment of the instrument. Another alternative embodiment employs three rotors and provides complete redundancy.

Claims

1. A processor for an inertial measurement unit that includes a number of linear acceleration transducers on a single rotor, at least one of said transducers mounted on said rotor such that the sensitive axis of said transducer is offset from and parallel to the rotor axis, and at least one other transducer having a sensitive axis, the center of sensitivity of which is located on the rotor axis, comprising:

means for demodulating the outputs of said transducers; and

means for processing the demodulated outputs of said transducers so as to obtain the translational acceleration and angular velocity of said unit in three mutually orthogonal directions, said processing means including means for deriving a set of dynamic variables which define said translational acceleration and said angular velocity in terms of linear combinations of selected demodulator outputs.

2. The processor of claim 1 wherein said dynamic variable deriving means includes a linear combining matrix.

3. The processor of claim 1, wherein said inertial measurement unit includes a drive unit which produces in-phase and quadrature reference signals, and wherein said demodulating means includes means utilizing said in-phase and quadrature reference signals for demodulating the output signals from selected transducers such that selected demodulated output signals represent in-phase and quadrature components of the corresponding original transducer output signals and wherein said dynamic variable demodulating means includes a linear combining matrix and means for coupling said selected in-phase and quadrature demodulated output signals to said linear combining matrix.

4. The processor of claim 1 wherein said rotor rotates about an axis denoted by x'"'"', wherein the angular velocity about three mutually orthogonal axes x'"'"', y'"'"', and z'"'"' defining the rotor coordinate system is denoted by p'"'"', q'"'"' and r'"'"', wherein the angular velocity of the unit about three mutually orthogonal axes x, y and z defining the unit coordinate system is denoted by p, q, and r, wherein the origins of the coordinate systems and the x and x'"'"' axes are coincident, wherein a_o, b_o, and c_o denote translational acceleration of said unit along axes x, y and z respectively, said set of dynamic variables including a_o, b_o, c_o, p'"'"'², pr-q, pq+r, p'"'"'q and p'"'"'r.

5. The processor of claim 1 wherein said rotor rotates about an axis denoted by x'"'"', wherein the angular velocity about three mutually orthogonal axes x'"'"', y'"'"' and z'"'"' defining the rotor coordinate system is denoted by p'"'"', q'"'"' and r'"'"', wherein the angular velocity of the unit about three mutually orthogonal axes x, y and z defining the unit coordinate system is denoted by p, q, and r, wherein the origins of the coordinate systems and the x and x'"'"' axes are coincident, wherein a_o, b_o and c_o denote translational acceleration of said unit along axes x, y and z respectively, wherein four transducers are utilized, wherein the output of said at least one transducer is a'"'"'₁.sbsb.dc, wherein the output of another of said transducers is c₄.sbsb.dc, wherein the transducer whose output is c'"'"'₄.sbsb.dc is located at a predetermined distance from the axis of the rotor with a sensitive axis normal thereto, wherein the speed of said rotor with respect to said unit is φ

, and wherein said dynamic variable deriving means includes means for obtaining the quantities a'"'"'₁.sbsb.dc, c'"'"'₄.sbsb.dc, -c_o, -p'"'"'q, p'"'"'r, b_o, φ and

said predetermined distance.

6. The processor of claim 5 wherein said processor includes a pair of transducers maintained equidistantly by an amount x to either side of the plane which contains the transducer whose output is c'"'"'₄.sbsb.dc, said pair of transducers having outputs c'"'"'₂ and c'"'"'₃, respectively, having parallel sensitive axes normal to the axis of the rotor and having centers of sensitivity coincident with the rotor axis, said at least one transducer being offset by a distance z'"'"'₁, and said dynamic variable deriving means including means for combining the outputs of said transducers in accordance with the following formulas:

##EQU13##

7. The processor of claim 1 wherein said dynamic variable deriving means includes a number of linear combining matrices.

8. A processor for an inertial measurement unit that includes a number of transducers mounted on a rotor, comprising:

means for demodulating the outputs of said transducers to produce d.c. signals; and

means for processing said d.c. signals so as to obtain the translational acceleration and angular velocity of said unit in three mutually orthogonal directions, said processing means including means for deriving a set of dynamic variables which define said translational acceleration and said angular velocity in terms of selected uncombined demodulated transducer outputs, linear combinations of selected demodulator outputs, and the rotational velocity of said rotor with respect to said unit.

9. The processor of claim 8 wherein said processing means includes a linear combining matrix.

10. A processor for an inertial measurement unit that includes a number of transducers mounted on a rotor, comprising:

means for processing the outputs of transducers so as to obtain the translational acceleration and angular velocity of said unit in three mutually orthogonal directions, said processing means including means for linearly combining the demodulated outputs of selected transducers thereby to permit the derivation of dynamic variables which define said translational acceleration and said angular velocity.

11. The processor of claim 10 wherein said linear combining means includes a linear combining matrix.

12. A processor for an inertial measurement unit that includes mutually orthogonal rotors, with each of said rotors having transducers mounted thereon, comprising:

means for demodulating the outputs of said transducers; and

means for processing the demodulated outputs of said transducers so as to obtain the translational acceleration and angular velocity of said unit in three mutually orthogonal directions, said processing means including means for deriving a set of dynamic variables which define said translational acceleration and said angular velocity in terms of linear combinations of selected demodulated transducer outputs, and the rotational velocities of said rotors with respect to said unit.

13. The processor of claim 12 wherein a first rotor rotates about an axis denoted by x'"'"', wherein the angular velocity about three mutually orthogonal axes x'"'"', y'"'"' and z'"'"' defining said first rotor coordinate system is denoted by p'"'"', q'"'"' and r'"'"', wherein the angular velocity of the unit about three mutually orthogonal axes x, y, and z defining the unit coordinate system as denoted by p, q, and r, wherein the origins of the coordinate systems for said first rotor and the x'"'"' and x axes are coincident, wherein a second rotor rotates about an axis orthogonal to said x'"'"' axis, wherein a, b, c denote translational acceleration along axes x, y and z respectively, wherein a'"'"', b'"'"' and c'"'"' denote translational acceleration along axes x'"'"', y'"'"' and z'"'"', respectively, wherein said first rotor includes three linear acceleration transducers, the outputs of which are a'"'"'₁, c'"'"'₂ and c'"'"'₃, wherein the angular velocity of said first rotor with respect to the case is φ

, wherein a second of said rotors includes three linear acceleration transducers having outputs b'"'"'₁, a'"'"'₂, and a'"'"'₃, wherein the rotational velocity of said second rotor with respect to said unit is θ

, and wherein said means for deriving a set of dynamic variables include combining network means operably connected to said demodulating means for deriving the quantities c_o, p'"'"', p'"'"'q, p'"'"'r, b_o, -a_o, q'"'"'r and q'"'"'p, where a_o, b_o and c_o denote the translational acceleration of said unit in inertial space.

14. In a system for measuring inertial parameters utilizing a unit containing a rotor and transducers carried thereon, a method for compensating for the non-ideal nature and alignments of the transducers while at the same time deriving the translational acceleration and the angular velocity of the unit comprising the steps of:

providing a matrix from demodulated outputs from selected non-ideal transducers carried on said rotor;

subtracting from this matrix the values which an ideal matrix should have, on an element by element basis;

inverting the resulting matrix; and

utilizing the inverted matrix to produce a linear combining matrix for the processing of signals from selected ones of said non-ideal transducers.

15. A processor for an inertial measurement unit that includes a number of transducers on a single rotor, at least one of said transducers mounted on said rotor such that the sensitive axis of said transducer is offset from and parallel to the rotor axis and at least one other transducer having a sensitive axis, the center of sensitivity of which is located on the rotor axis, comprising:

means for processing the outputs of said transducers so as to obtain the translational acceleration and angular velocity of said unit in three mutually orthogonal directions, said processing means including means for deriving the dynamic variables which define said translational acceleration and said angular velocity as linear combinations of demodulated transducer outputs.

16. In combination:

an inertial measurement unit including a number of transducers on a single rotor, at least one of said transducers mounted on said rotor such that the sensitive axis of said transducer is offset from and parallel to the rotor axis and at least one other transducer having a sensitive axis, the center of sensitivity of which is located on the rotor axis, and

means for processing the outputs of said transducers so as to obtain the translational acceleration and angular velocity of said unit in three mutually orthogonal directions.

17. The combination of claim 16 wherein said processing means includes means for deriving dynamic variables which define said translational acceleration and said angular velocity as linear combinations of demodulated transducer outputs.