Signal processor for inertial measurement using coriolis force sensing accelerometer arrangements

• US 4,870,588 A
• Filed: 10/21/1985
• Issued: 09/26/1989
• Est. Priority Date: 10/21/1985
• Status: Expired due to Fees

• Current Assignee: Sundstrand Corporation (Raytheon Technologies Corporation d/b/a RTX)
• Original Assignee: Sundstrand Data Control, Inc.
• Inventors: Merhav, Shmuel J.
• Primary Examiner(s): Lall, Parshotam S.
• Assistant Examiner(s): Trans, V. N.
• Application Number:
  US06/789,658
• Time in Patent Office: 1,436 Days
• Field of Search: 364/453, 364/454, 364/566, 73/504, 73/510, 73/511, 73/514, 73/517 R, 73/178 R
• US Class Current: 701/220
• CPC Class Codes: G01C 19/00 Gyroscopes; Turn-sensitive devices using vibrating masses; Turn-sensitive devices without moving masses; Measuring angular rate using gyroscopic effects
• CPC Class Codes: G01P 15/08 with conversion into electric or magnetic values

Abstract

Disclosed is a method and apparatus for processing signals supplied by accelerometer assemblies in which one or more accelerometers are cyclically displaced in a predetermined manner so that signals representing the specific force experienced by the accelerometers and the angular rate experienced by the accelerometers are produced. The signal processor separately estimates the signal components of the signal being processed and provides an error signal by subtracting the estimated signal components from the signal being processed. The error signal is fed back through circuitry that controls the magnitude of the estimated signal components so that the value of each estimated signal component rapidly converges to the value of the signal components of the signal being processed. In an arrangement for determining the angular rate of one or more pair of cyclically displaced accelerometers, the signal processor includes a signal component that is in-phase with the signal that oscillates the accelerometer pair, a signal component that is in phase quadrature with the signal that displaces the accelerometer pair and a signal component that corresponds to random unmodulated additive noise. In this arrangement, a signal summing unit subtracts estimates of each signal component from the signal being processed to supply an error signal equal to (a₁ -a₁) Cos ωt+(a₂ -a₂) Sin ωt+(a₃ -a₃), where a₁, a₂, and a₃, respectively represent the values of the in-phase, quadrature and random noise components of the signal being processed and a₁, a₂, and a₃ represent estimates of those signal component values. To obtain a₃, the error signal is scaled and integrated. To obtain the a₁ Cos ωt signal estimate, the error signal is multiplied by Cos ωt, and scaled to obtain a signal representative of the derivative with respect to time of a₁. This signal is then integrated and multiplied by Cos ωt. The signal component a₂ Sin ωt is obtained in a similar manner by multiplying the error signal by Sin ωt, scaling, integrating and multiplying the integrated signal by Sin ωt. In such an arrangement, angular rate is obtained by scaling the a₁ signal estimate. In addition, the signal estimates can be used in an inertial navigation system that employs the invention to eliminate misalignment of the accelerometer pairs and to improve system operation by eliminating phase shift between the signal source that oscillates the accelerometer pairs and the signals provided by the accelerometer pairs.

Claims

1. A signal processing apparatus for determining the amplitudes of components of an output signal obtained from a pair of accelerometers mounted on a moving body, the pair of accelerometers being mounted in alignment with an axis of a coordinate system associated with said moving body and being cyclically displaced in response to a drive signal, said output signal including a first periodic component having a first amplitude equal to an angular rate associated with the moving body, a second periodic component having a second amplitude, and a third component having a third amplitude representing an unmodulated random value sequence, said signal processing apparatus comprising:

(a) means for repetitively estimating the first, second and third amplitudes;

(b) means for repetitively producing an estimated signal corresponding to the output signal obtained from said pair of accelerometers using said estimated first, second and third amplitudes;

(c) means for repetitively determining an error signal by subtracting the estimated signal from the output signal obtained from said pair of accelerometers; and

(d) means for repetitively revising the three estimated amplitudes as a function of the error signal,

(e) the error signal approaching zero as the estimated amplitudes respectively approach the first, second and third amplitudes of the output signal obtained from said pair of accelerometers, the estimated first amplitude being then equal to the angular rate associated with the moving body.

2. The signal processing apparatus of claim 1, wherein the first amplitude includes an error component caused by a misalignment between the pair of accelerometers and a path along which said pair of accelerometers are cyclically displaced, and an error component caused by a phase difference between said drive signal and said output signal, and wherein said second amplitude includes an error component caused by the misalignment between the pair of accelerometers and the path along which said pair of accelerometers are cyclically displaced, said signal processing apparatus further comprising:

(a) means for constraining said moving body in a stationary condition;

(b) means for monitoring the level of the estimated amplitude of the second component used to produce the estimated signal; and

(c) means for aligning said pair of accelerometers with said path along which said accelerometers are cyclically displaced to achieve a minimum value for the estimated second amplitude signal.

3. The signal processing apparatus of claim 1, wherein the first component of the signal obtained from a pair of accelerometers includes a cosine function and the second component of said signal includes a sine function.

4. The signal processing apparatus of claim 3, further comprising:

(a) means for moving said moving body in a controlled manner;

(b) means for monitoring the level of the estimated amplitude of the second component used to produce the estimated signal; and

(c) means for adjusting said phase difference between said drive signal and said output signal obtained from said accelerometer pair to achieve a minimum value for the second estimated amplitude.

5. Apparatus for determining components of angular rate of a moving body, comprising:

(a) at least one pair of accelerometers, mounted in alignment with an axis of a coordinate system associated with said moving body, and cyclically displaced in response to a drive signal, producing an output signal of the form a₁ f₁ (ω

t)+a₂ f₂ (ω

t)+a₃, where f₁ (ω

t) is a periodic function of time representative of the cyclic oscillation of said accelerometer pair, f₂ (ω

t) is a periodic function of time that is phase quadrature with f₁ (ω

t), a₁ and a₂ respectively denote the amplitudes of said time varying functions f₁ (ω

t) and f₂ (ω

t), and a₃ represents an unmodulated random value sequence;

(b) means for reiteratively subtracting a signal form;

af₁ (ω

t)+a₂ f₂ (ω

t)+a₃ from said signal of the form a₁ f₁ (ω

t)+a₂ f₂ (ω

t)+a₃ to provide an error signal where a₁, a₂, and a₃ are signals estimating said amplitude factors a₁, a₂, and a₃, respectively;

(c) means for reiteratively multiplying said error signal by k₁ f₁ (ω

t), where k₁ is a predetermined constant, to obtain a signal that is a derivative with respect to time of a₁ ;

(d) means for reiteratively multiplying said error signal by k₂ f₂ (ω

t), where k₂ is a predetermined constant, to obtain a signal that is a derivative with respect to time of a₂ ;

(e) means for reiteratively multiplying said error signal by a predetermined constant k₃ to obtain a signal that is a derivative with respect to time of a₃ ;

(f) means for reiteratively integrating said signal that is a derivative with respect to time of a₁ to obtain a₁ ;

(g) means for reiteratively integrating said signal that is a derivative with respect to time of a₂ to obtain a₂ ;

(h) means for reiteratively integrating said signal that is a derivative with respect to time of a₃ to obtain a₃ ;

(i) means for reiteratively multiplying a₁ by f₁ (ω

t);

(j) means for reiteratively multiplying a₂ by f₂ (ω

t); and

(k) means for multiplying the estimated signal a₁ by a predetermined scale factor to produce a signal that corresponds to said angular rate of the moving body as the error signal approaches a minimum and estimated signals a₁, a₂, and a₃ approach the values of amplitudes a₁, a₂ and a₃, respectively, with successive iterations by the above-recited means (b) through (j).

6. The apparatus of claim 5, wherein f₁ (ω

t) is equal to cos (ω

t), and f₂ (ω

t) is equal to sin (ω

t).

7. The apparatus of claim 5, wherein said amplitude a₁ includes an error component caused by an unknown amount of misalignment between the pair of accelerometers that supply said signal and the path along which said accelerometers are cyclically displaced and an error component caused by a phase difference between said drive signal and said output signal obtained from said accelerometer pair, and wherein said amplitude factor a₂ includes an error component caused by misalignment between said accelerometer pair and the path along which said accelerometers are cyclically displaced;

said apparatus further comprising;

(a) means for constraining said moving body in a stationary condition;

(b) means for monitoring the estimated signal a₂ ; and

(c) means for aligning said accelerometer pair with said path along which said accelerometers are cyclically displaced to achieve a minimum value for a₂.

8. The apparatus of claim 7, further comprising:

(a) means for moving said moving body in a controlled manner;

(b) means for monitoring the estimated signal a₂ ; and

(c) means for adjusting said phase difference between said drive signal and said output signal obtained from said accelerometer pair to achieve a minimum value for a₂.

9. An adaptive signal processor for determining an angular rate component of a moving body from an output signal that is obtained from at least one pair of accelerometers that are mounted in alignment with one axis of a coordinate system associated with said moving body, each said pair of accelerometers being cyclically oscillated by a signal source that drives said one or more accelerometer pairs, said output signal obtained from at least one pair of said accelerometers being of the form a₁ f₁ (ω

t)+a₂ f₂ (ω

t)+a₃, where f₁ (ω

t) is a first periodic function of time representative of the cyclic oscillation each said accelerometer pair, f₂ (ω

t) is a second periodic function of time that is in phase quadrature with f₁ (ω

t), a₁ and a₂ respectively denote the amplitudes of said time varying functions f₁ (ω

t) and f₂ (ω

t) and a₃ represents an unmodulated random value sequence said signal processor comprising;

a signal summing unit, said signal summing unit including a first input terminal connected for receiving said output signal of the form a₁ f₁ (ω

t)+a₂ f₂ (ω

t)+a₃, a second input terminal for receiving a signal a₁ f₁ (ω

t), a third input terminal for receiving a signal a₂ f₂ (ω

t) and having a fourth input terminal for receiving a signal a₃ where a₁, a₂ and a₃ are estimates of said amplitude factors a₁, a₂ and a₃, said signal summing unit being configured and arranged for supplying a first error signal equal to the difference between the output signal applied to said first input terminal and the signals applied to said second, third and fourth input terminals;

first and second multiplier circuits, each of said first and second multiplier circuits having an input terminal connected for receiving said first error signal;

a first signal source for generating a signal f₁ (ω

t), said first signal source being connected for supplying said signal to a second input terminal of said first multiplier circuit;

a second signal source for supplying a signal f₂ (ω

t), said second signal source being connected for supplying said signal to a second input terminal of said second multiplier;

first and second scaling means, said first scaling means being connected for receiving the signal supplied by said first multiplier circuit, said second scaling means being connected for receiving the signal supplied by said second multiplier circuit;

first and second integrator circuits, said first integrator circuit being connected for receiving the signal supplied by said first scaling means, and being connected for supplying a signal representative of a₁, said second integrator circuit being connected for receiving the signal supplied by said second scaling means and being connected for supplying a signal representative of a₂ ;

third and fourth multiplier circuits, said third multiplier circuit being connected for receiving said signal a₁ from said first integrator circuit and being connected for receiving said signal supplied by said first signal source, said fourth multiplier circuit being connected for receiving said signal a₂ from said second integrator circuit and being connected for receiving said signal supplied by said second signal generator, said third and fourth multiplier circuits each having an output terminal, the output terminal of said third multiplier circuit being connected to said second input terminal of said signal summing unit, said output terminal of said fourth multiplier circuit being connected to said third input terminal of said signal summing unit;

a third scaling means connected for receiving said first error signal, said third scaling means being operative to amplify said first error signal by a predetermined constant k₃ ;

a third integrator circuit connected for receiving the signal supplied by said third scaling means and connected for supplying the output signal of said third integrator circuit to said fourth input terminal of said signal summing unit, wherein the signal summing unit reiteratively determines the first error signal, the first error signal approaching a minimum value with successive iterations, the signals a₁, a₂ and a₃ respectively approaching equivalency with amplitude factors a₁, a₂ and a₃.

10. The signal processor of claim 9 wherein said first periodic function of said signal obtained from said accelerometer pair is Cos ω

t and said second periodic function of said signal obtained from said accelerometer pair is Sin ω

t and wherein said signals supplied by said first and second signal sources are Cos ω

t and Sin ω

t, respectively.

11. The signal processor of claim 9 wherein two pairs of accelerometers are mounted in alignment with one axis of said coordinate system with said first pair of accelerometers supplying said signal of the form a₁ f₁ (ω

t)+a₂ f₂ (ω

t)+a₃ and said second pair of accelerometers supplying a signal of the form a₁^s f₂ (ω

t)-a₂^s f₁ (ω

t)+a₃^s where a₁^s and a₂^s respectively denote the amplitudes of said periodic functions f₁ (ω

t) and f₂ (ω

t) and a₃^s is an unmodulated random value sequence and wherein said signal processor further comprises;

first, second and third adder circuits, said first adder circuit having an input terminal connected for receiving the output signal of said first multiplier circuit and having an output terminal connected to the input of said first scaling means, said second adder circuit having an input terminal connected to the output of said second multiplier circuit and having an output terminal connected to the input of said second scaling means, said third adder circuit having an input terminal connected for receiving said first error signal and an output terminal connected to the input of said third scaling means;

a second signal summing unit, said summing unit having first, second, third and fourth input terminals, with said first input terminal being connected for receiving said signal supplied by said second pair of accelerometers, said second signal summing unit supplying a second error signal equal to the difference between said signal supplied to said first input terminal and signals supplied to said second, third and fourth input terminals, said second error signal being supplied to a second input terminal of said third adder circuit;

fifth and sixth multiplier circuits each having an input terminal connected for receiving said second error signal, said fifth multiplier circuit having an output terminal connected to a second input terminal of said first adder circuit, said sixth multiplier circuit having an output terminal connected to a second input terminal of said second adder circuit, said fifth multiplier circuit having a second input terminal connected for receiving said signal supplied by said second signal source;

seventh and eighth multiplier circuits, said seventh multiplier circuit having first and second input terminals respectively connected for receiving the signal supplied by said first integrator circuit and said signal supplied by said second signal source and having an output terminal connected to said second input terminal of said second signal summing unit, said eighth multiplier circuit having a first input terminal connected for receiving the signal supplied by said second integrator circuit and having an output terminal connected to said third input terminal of said second signal summing unit; and

a third signal source, said third signal source connected for supplying a signal -f₁ (ω

t) to a second input terminal of said sixth multiplier circuit and to a second input terminal of said eight multiplier circuit.

12. The signal processor of claim 11 wherein the said first periodic function of time is Cos ω

t and said second periodic function of time is Sin ω

t and wherein said first signal source supplies a signal Cos ω

t, said second signal source supplies a signal Sin ω

t and said third signal source supplies a signal -Cos ω

t.

13. The signal processor of claim 9 wherein two pairs of accelerometers are mounted in alignment with one axis of said coordinate system with said first pair of accelerometers supplying said signal of the form a₁ f₁ (ω

t)+a₂ f₂ (ω

t)+a₃ and said second pair of accelerometers supplying a signal of the form a₁^s f₂ (ω

t)-a₂^s f₁ (ω

t)+a₃^s where a₁^s and a₂^s respectively denote the amplitudes of said periodic functions of time f₁ (ω

t) and f₂ (ω

t) and a₃^s is an unmodulated random value sequence and wherein said signal processor further comprises;

a second signal summing unit, said summing unit having first, second, third and fourth input terminals, with said first input terminal connected for receiving said signal supplied by said second pair of accelerometers, said second signal summing unit supplying a second error signal equal to the difference between said signal supplied to said first input terminal and signals supplied to said second, third and fourth input terminals;

fifth and sixth multiplier circuits each having an input terminal connected for receiving said second error signal, said fifth multiplier circuit having a second input terminal connected for receiving said signal supplied by said second signal source;

a third signal source, said third signal source connected for supplying a signal -f₁ (ω

t) to a second input terminal of said sixth multiplier circuit;

seventh and eighth multiplier circuits, said seventh multiplier circuit having an input terminal connected for receiving the signal supplied by said second signal source and a second input terminal connected for receiving the signal supplied by said first integrator circuit and having an output terminal connected to said second input terminal of said second signal summing unit, said eighth multiplier circuit having a first input terminal connected for receiving said signal supplied by said third signal source and having an output terminal connected to said third input terminal of said second signal summing unit; and

fifth and sixth integrator circuits, said fifth integrator circuit connected for receiving the signal supplied by said sixth multiplier circuit and connected for supplying a signal to a second input terminal of said eighth multiplier circuit, said sixth integrator circuit connected for receiving said second error signal and connected for supplying a signal to said fourth input terminal of said second summing unit.

14. The signal processor of claim 13 wherein said first periodic function of time is Cos ω

t and said second periodic function of time is Sin ω

t and wherein said first signal source supplies a signal Cos ω

t, said second signal source supplies a signal Sin ω

t and said third generator supplies a signal -Cos ω

t.

15. An inertial measurement unit for measuring specific force and angular rate components of a moving body, said inertial measurement unit comprising:

an accelerometer assembly including at least three pairs of accelerometers, each of said pairs of accelerometers being mounted for measuring specific force along one axis of a Cartesian coordinate system that is fixed to said moving body and being mounted for cyclic displacement of the form f₁ (ω

t);

accelerometer drive means connected to said accelerometer assembly for cyclically dispacing each of said three pairs of accelerometers;

a preprocessor unit connected for receiving said signal supplied by each of said accelerometers, said preprocessor unit including means for adding the signals supplied by each said pair of accelerometers and means for subtracting the signals supplied by each said pair of accelerometers to generate first, second and third signals representative of the specific force along each coordinate axis of said coordinate system and to generate first, second and third angular rate signals of the form a₁ f₁ (ω

t)+a₂ f₂ (ω

t)+a₃, where f₂ (ω

t) is a periodic signal that is in phase quadrature with said signal f₁ (ω

t), and where a₁ and a₂ denote the amplitudes of said periodic signals f₁ (ω

t) and f₂ (ω

t) respectively and a₃ represents an unmodulated random value sequence with the amplitude a₁ of each of said signals being representative of the angular rate associated with a separate axis of said coordinate system of said moving body; and

first, second and third signal processing circuits, said first, second and third signal processing circuits being connected for receiving a different one of said angular rate signals supplied by said preprocessor unit and for supplying a signal representative of the angular rate for a different one of said axes of said coordinate system, each of said first, second and third signal processing circuits including

(a) a signal summing unit, said signal summing unit including a first input terminal connected for receiving said signal of the form a₁ f₁ (ω

t)+a₂ f₂ (ω

t)+a₃, a second input terminal for receiving a signal a₁ f₁ (ω

t), a third input terminal for receiving a signal a₂ f₂ (ω

t) and having a fourth input terminal for receiving a signal a₃ where a₁, a₂ and a₃ are estimates of said amplitudes a₁, a₂ and a₃, said signal summing unit being configured and arranged for supplying a first error signal equal to the difference between the signal applied to said first input terminal and the signals applied to said second, third and fourth input terminals;

(b) first and second multiplier circuits, each of said first and second multiplier circuits having an input terminal connected for receiving said first error signal;

(c) a first signal source for generating a signal f₁ (ω

t), said first signal source being connected for supplying said signal to a second input terminal of said first multiplier circuit;

(d) a second signal source for supplying a signal f₂ (ω

t), said second signal source being connected for supplying said signal to a second input terminal of said second multiplier;

(e) first and second scaling means, said first scaling means being connected for receiving the signal supplied by said first multiplier circuit, said second scaling means being connected for receiving the signal supplied by said second multiplier circuit;

(f) first and second integrator circuits, said first integrator circuit being connected for receiving the signal supplied by said first scaling means, and being connected for supplying a signal representative of a₁, said second integrator circuit being connected for receiving the signal supplied by said second scaling means and being connected for supplying a signal representative of a₂ ;

(g) third and fourth multiplier circuits, said third multiplier circuit being connected for receiving said signal a₁ from said first integrator circuit and being connected for receiving said signal supplied by said first signal source, said fourth multiplier circuit being connected for receiving said signal a₂ from said second integrator circuit and being connected for receiving said signal supplied by said second signal generator, said third and fourth multiplier circuits each having an output terminal, the output terminal of said third multiplier circuit being connected to said second input terminal of said summing unit, said output terminal of said fourth multiplier circuit being connected to said third input terminal of said signal summing unit;

(h) a third scaling means connected for receiving said first error signal, said third scaling means for amplifying said first error signal by a predetermined constant k₃ ; and

(i) a third integrator circuit connected for receiving the signal supplied by said third scaling unit and connected for supplying the output signal of said integrator circuit to said fourth input terminal of said summing unit.

16. The inertial measurement unit of claim 15 wherein said first periodic function of said signal obtained from said accelerometer pair is Cos ω

t and said second periodic function of said signal obtained from said accelerometer pair is Sin ω

t and wherein said signals supplied by said first and second signal sources are Cos ω

t and Sin ω

t, respectively.

17. The inertial measurement unit of claim 15 wherein the alignment of each pair of accelerometers relative to the direction of said cyclic displacement is adjustable in response to the a₂ of each of said first, second and third signal processors for aligning each pair of accelerometers with said direction of cyclic displacement.

18. The inertial measurement unit of claim 17 wherein the phase angle of the cyclic displacement of each of said three pairs of accelerometers is adjustable in response to a₂ of each of said first, second and third signal processors to eliminate phase shift error between the cyclic displacement of each said accelerometer pair and the output signal provided by that accelerometer pair.

19. The inertial measurement unit of claim 15 wherein two pairs of accelerometers are mounted in alignment with each axis of said coordinate system with the first pair of accelerometers supplying said signal of the form a₁ f₁ (ω

t)+a₂ f₂ (ω

t)+a₃ and the second pair of accelerometers supplying a signal of the form a₁^s f₂ (ω

t)-a₂^s f₁ (ω

t)+a₃^s where a₁^s and a₂^s respectively denote the amplitudes of said periodic functions f₁ (ω

t) and f₂ (ω

t) and a₃^s is an unmodulated random value sequence and wherein each said signal processing unit further comprises;

first, second and third adder circuits, said first adder circuit having an input terminal connected for receiving the output signal of said first multiplier circuit and having an output terminal connected to the input of said first scaling means, said second adder circuit having an input terminal connected to the output of said second multiplier circuit and having an output terminal connected to the input of said second scaling means, said third adder circuit having an input terminal connected for receiving said first error signal and an output terminal connected to the input of said third scaling means;

a second signal summing unit, said summing unit having first, second, third and fourth input terminals, with said first input terminal being connected for receiving said signal supplied by said second pair of accelerometers, said second signal summing unit supplying a second error signal equal to the difference between said signal supplied to said first input terminal and signals supplied to said second, third and fourth input terminals, said second error signal being supplied to a second input terminal of said third adder circuit;

fifth and sixth multiplier circuits each having an input terminal connected for receiving said second error signal, said fifth multiplier circuit having an output terminal connected to a second input terminal of said first adder circuit, said sixth multiplier circuit having an output terminal connected to a second input terminal of said second adder circuit, said fifth multiplier circuit having a second input terminal connected for receiving said signal supplied by said second signal source;

seventh and eighth multiplier circuits, said seventh multiplier circuit having first and second input terminals respectively connected for receiving the signal supplied by said first integrator circuit and said signal supplied by said second signal source and having an output terminal connected to said second input terminal of said second signal summing unit, said eighth multiplier circuit having a first input terminal connected for receiving the signal supplied by said second integrator circuit and having an output terminal connected to said third input terminal of said second signal summing unit; and

a third signal source, said third signal source connected for supplying a signal -f₁ (ω

t) to a second input terminal of said sixth multiplier circuit and to a second input terminal of said eighth multiplier circuit.

20. The inertial measurement unit of claim 19 wherein said first periodic function of time is Cos ω

t, said second periodic function of time is Sin ω

t and wherein said first signal source supplies a signal Cos ω

t, said second signal source supplies a signal Sin ω

t and said third signal source supplies a signal -Cos ω

t.

21. The inertial measurement unit of claim 20 wherein the alignment of each pair of accelerometers relative to the direction of said cyclic displacement is adjustable in response to the a₂ of each of said first, second and third signal processors for aligning each pair of accelerometers with said direction of cyclic displacement.

22. The inertial measurement unit of claim 21 wherein the phase angle of the cyclic displacement of each of said three pairs of accelerometers is adjustable in response to a₂ of each of said first, second and third signal processors to eliminate phase shift error between the cyclic displacement of each said accelerometer pair and the output signal provided by that accelerometer pair.

23. The inertial measurement unit of claim 15 wherein two pairs of accelerometers are mounted in alignment with one axis of said coordinate system with said first pair of accelerometers supplying said signal of the form a₁ f₁ (ω

t)+a₂ f₂ (ω

t)+a₃ and said second pair of accelerometers supplying a signal of the form a₁^s f₂ (ω

t)-a₂^s f₁ (ω

t)+a₃^s where a₁^s and a₂^s respectively denote the amplitudes of said periodic functions of time f₁ (ω

t) and f₂ (ω

t) and a₃^s is an unmodulated random value sequence and wherein said signal processing unit further comprises;

a second signal summing unit, said summing unit having first, second, third and fourth input terminals, with said first input terminal connected for receiving said signal supplied by said second pair of accelerometers, said second signal summing unit supplying a second error signal equal to the difference between said signal supplied to said first input terminal and signals supplied to said second, third and fourth input terminals;

fifth and sixth multiplier circuits each having an input terminal connected for receiving said second error signal, said fifth multiplier circuit having a second input terminal connected for receiving said signal supplied by said second signal source;

a third signal source, said third signal source connected for supplying a signal -f₁ (ω

t) to a second input terminal of said sixth multiplier circuit;

seventh and eighth multiplier circuits, said seventh multiplier circuit having an input terminal connected for receiving the signal supplied by said second signal source and a second input terminal connected for receiving the signal supplied by said first integrator circuit and having an output terminal connected to said second input terminal of said second signal summing unit, said eighth multiplier circuit having a first input terminal connected for receiving said signal supplied by said third signal source and having an output terminal connected to said third input terminal of said second signal summing unit; and

fifth and sixth integrator circuits, said fifth integrator circuit connected for receiving the signal supplied by said sixth multiplier circuit and connected for supplying a signal to a second input terminal of said eighth multiplier circuit, said sixth integrator circuit connected for receiving said second error signal and connected for supplying a signal to said fourth input terminal of said second summing unit.

24. The inertial measurement unit of claim 23 wherein said first periodic function of time is Cos ω

t, said second periodic function of time is Sin ω

t and wherein said first signal source supplies a signal Cos ω

t, said second signal source supplies a signal Sin ω

t and said third signal source supplies a signal -Cos ω

t.

25. The inertial measurement unit of claim 24 wherein the alignment of each pair of accelerometers relative to the direction of said cyclic displacement is adjustable in response to the a₂ and a₂^s of each of said first, second and third signal processors, for aligning each pair of accelerometers with said direction of cyclic displacement.

26. The inertial measurement unit of claim 25 wherein the phase angle of the cyclic displacement of each of said three pairs of accelerometers is adjustable in response to a₂ and a₂^s of each of said first, second and third signal processors to eliminate phase shift error between the cyclic displacement of each said accelerometer pair and the output signal provided by that accelerometer pair.