BACKGROUND OF THE INVENTION FIELD OF THE INVENTION The present invention relates generally to inertial navigation systems, and specifically to a system and method for improving the accuracy of pressure altitude determinations in an inertial navigation system.

DESCRIPTION OF RELATED ART Inertial navigation systems in aircrafts have typically employed accelerometers to provide position information to a navigation computer. It is well known that the vertical position (altitude) of the aircraft can be determined from a measured acceleration in the vertical direction by performing a double time integration of the measured vertical acceleration.

The double integration of acceleration in the vertical direction is unstable as acceleration bias can lead to exponential growth in the computed altitude, causing the estimated altitude calculation to have unbounded error due to several factors.

First, any vertical acceleration measurement errors from the accelerometers are directly integrated in subsequent calculations to cause both vertical velocity and

vertical position error. Second, in order to obtain the actual value for vertical acceleration from the measurement taken by the accelerometer, the effects of gravity must be subtracted from the vertical acceleration measurement. Erroneous acceleration measurements will cause incorrect values for gravity to be subtracted from the measured acceleration, which further compounds the error in the altitude determination causing an even faster growth in the altitude error. Thus, inertial navigation systems relying upon the integration of acceleration measurements to obtain an estimation of altitude are unstable systems.

To provide a more stable inertial navigation system, external references have been used either alone or in combination with inertial measurements to compute estimations of altitude. For instance, a barometric altimeter is a well known device for providing altitude information as a function of the value of barometric pressure based on the direct relationship between pressure and altitude. Barometric altitude, also known as pressure altitude, is determined as a function of pressure based on the standard day model for the atmosphere : where S is the pressure altitude, K1--44-342 [km], K2=0. 190263 [km], K3=45.395 [km], K4=14. 605 [km], Po=1013. 25 [mb], and PB=226. 32 [mb]. Since the barometric altitude determination is stable, it is typically used in a variety of mechanizations to

aid or bound the altitude estimations computed from the inertial measurements in the inertial vertical loop. The independent pressure altitude estimation that aids the inertial vertical loop is referred to as the slave altitude.

Differences between the altitude estimation and true altitude can result from the altitude estimation being based on the standard atmosphere, whereas actual atmospheric conditions encountered by a navigation system are usually nonstandard.

Furthermore, errors in the sensors providing the intertial vertical loop data and the pressure data used for the altitude estimation will produce differences between the acutal altitude and the estimated altitude. These errors have been modeled in a five state Kalman filter mechanization for the free inertial vertical loop in order to minimize their detrimental effects, as described in the article"A Kalman Filter Mechanization for the Baro-Inertial Vertical Channel"by J. Ausman in Proceedings of the Institute of Navigation Forty Seventh Annual Meeting, Williamsburg, VA, pp.

153-159,1991. The disclosure of this article is hereby incorporated by reference into the present application. This Kalman filter mechanization models five error states including three error states for the inertial vertical loop (inertial vertical acceleration error, inertial vertical velocity error, and inertial vertical position error) and two error states for the pressure altitude (barometric scale factor and barometric bias in the barometer).

The barometric bias and barometric scale factor estimate errors in the pressure altitude determinations made from pressure measurements taken by the

barometric altimeter. The barometric bias and scale factor error states essentially attempt to account for differences between the calculated pressure altitude and the calculated intertial altitude, so that the barometic scale factor and the barometric bias are actually modeling errors in altitude. However, the barometric altimeter does not directly measure altitude, rather it directly measures pressure and then mathematically converts the pressure measurement to a value for altitude. Thus, barometric scale factor and barometric bias are actually modeling altitude errors in the barometric altimeter in an artificial domain, since these error states are modeling errors in altitude instead of modeling errors in the actual pressure measurements taken. Noise and pressure offsets in the pressure sensor will result in erroneous pressure measurements which are, in turn, converted into erroneous altitude determinations. The effects of noise and offset on the pressure sensor offset error can deviate substantially at higher altitudes from purely an altitude error, so that merely modeling altitude errors will not always provide an accurate correction of the altitude determination.

Thus, there is clearly a need for a system and method for directly modeling the pressure sensor offset error and noise in the pressure measurements themselves in order to increase the accuracy of altitude determinations in an inertial navigation system.

SUMMARY OF THE INVENTION The present invention provides a system and method for improving the accuracy of altitude determinations in an inertial navigation system. The system utilizes pressure measurements which are taken by a barometric altimeter and converted into an estimated pressure altitude using any known pressure-to-altitude conversion. The estimated pressure altitude is then converted into a pressure correction value using a correction value generating formula that is a function of altitude. The pressure correction value is then multiplied by a pressure offset value for the barometric altimeter to generate a pressure offset error for the barometric altimeter. This pressure offset error is used in the present invention to modify the altitude estimation in order to generate an altitude determination having an improved accuracy. The present invention further determines an amount of observation noise in the barometric altimeter that is a function of pressure noise and altitude, where the altitude estimation is further modified to account for the observation noise. Thus, the system and method of the present invention improves the accuracy of altitude determinations in inertial navigation system over prior systems by directly accounting for errors in pressure measurements taken by a barometric altimeter due to offset and noise.

BRIEF DESCRIPTION OF THE DRAWINGS The features of the present invention, which are believed to be novel, are set forth with particularity in the appended claims. The present invention, both as to its organization and manner of operation, together with further advantages, may best be understood by reference to the following description, taken in connection with the accompanying drawings in which the reference numerals designate like parts throughout the figures thereof and wherein: FIG. 1 is an operational block diagram of a preferred method of improving the accuracy of pressure altitude determinations in an inertial navigation system of the present invention; FIG. 2 is a state diagram of six state Kalman filter mechanization of a preferred embodiment of the present invention; FIG. 3 is an operational block diagram of a preferred method of modeling the observation noise in accordance with the present invention; FIGS. 4A-4C are graphical illustrations of an example of the flight profile of an X- 33 aircraft ; FIG. 4D is a graphical illustration of the inertial velocity error for the flight profile of FIGS. 4A-4C; FIG. 4E is a graphical illustration of the vertical position error for the flight profile of FIGS. 4A-4C ; FIG. 5A is a graphical illustration of the inertial velocity error for the flight profile of FIGS. 4A-4C with the pressure offset error state of the present invention applied thereto; and FIG. 5B is a graphical illustration of the vertical position error for the flight profile of FIGS. 4A-4C with the pressure offset error state of the present invention applied thereto.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The following description is provided to enable any person skilled in the art to make and use the invention and sets forth the best modes contemplated by the inventors of carrying out their invention. Various modifications, however, will remain readily apparent to those skilled in the art, since the general principles of the present invention have been defined herein specifically to provide a system and method of improving the accuracy of pressure altitude determinations in an inertial navigation system.

In order to provide an accurate estimation of altitude using external references available to inertial navigation systems in an aircraft, external measurements, such as pressure, temperature and vertical acceleration, are often taken and converted into an estimation of altitude using a physical relationship between the external measurements and altitude. The pressure measurements are taken by a barometric altimeter and converted into an altitude estimation either alone or in combination with other external measurements. In order to account for errors in altitude estimations made from barometric altimeter measurements, past inertial navigation systems have accounted for errors in an artificial domain by estimating errors in the estimated altitude due to barometric scale factor or barometric bias. The present invention models errors directly in the pressure measurements themselves to account for pressure sensor offset and observation noise in the barometric altimeter.

Referring now to FIG. 1, an operational block diagram of a preferred method of improving the accuracy of pressure altitude determinations in an inertial navigation system of the present invention is illustrated. Initially in step 100, a pressure measurement from the barometric altimeter is obtained, and the pressure . measurement is converted into an estimated pressure altitude S using any known relationship between pressure and altitude. The standard day model for the atmosphere is one method commonly used for computing a pressure altitude value from a measured pressure value according to the following equation :

where P is the measured pressure, S is the pressure altitude, Kl=44. 342 [km], K2=0. 190263, K3=45.395 [km], K4=14. 605 [km], Po=1013. 25 [mb], and PB=226. 32 [mb]. The standard day model is only one possible manner of converting a measured pressure to an estimated altitude, where it is known to those skilled in the art that altitude can be computed from pressure measurements in a plurality of possible manners, including using a combination. of a plurality of external measurements, such as pressure and temperature or other measurable external conditions.

The altitude estimation S is then converted into a pressure correction value using a correction value generating fonnulaf (S) in step 102. The correction value generating formulae) is the derivative of the standard day pressure altitude and is represented by the equation:

where K5 = K2 K1/P@, and SB=11 [km].

Po The correction value generating formula f (S) is then multiplied by a value for the pressure offset PE for the barometric altimeter to generate a pressure offset error value ASp in step 104 according to the equation: #SP = PE#f(S) The pressure offset error value ASp is then used to modify the altitude estimation in step 106 to generate an altitude determination having an improved accuracy by directly accounting for the pressure offset error in the barometric altimeter. [NOTE : HOW IS THE PRESSURE OFFSET ERROR VALUE ACTUALLY USED TO MODIFY THE ALTITUDE ESTIMATION AND IS IT IMPORTANT TO THE PRESENT INVENTION ?] In order to account for the various errors associated with generating an altitude determination, these errors can be modeled in a state diagram mechanization.

Referring now to FIG. 2, a state diagram of six state Kalman filter mechanization of a preferred embodiment of the present invention is illustrated. The mechanization includes three inertial error states: a stabilized vertical acceleration bias error state S3

, which feeds a vertical velocity error state S2, which feeds a vertical position error state SI. The inertial error states are used to correct a inertial altitude P calculated from a measured vertical acceleration as commonly known to those skilled in the art.

Vertical acceleration bias can be difficult to model, since changes in attitude can result in other accelerometers tipping into the vertical channel. In order to avoid accelerometer bias modeling in all three axes, accelerometer switching can be accounted for by injecting a covariance into the single channel dependent on platform attitude.

The Kalman filter mechanization of FIG. 2 further contains three error states accounting for the barometric altimeter (or barometer): a barometer bias error state S4, a barometer scale factor error state S5, and a pressure offset error state S6. The observation element for the barometer scale factor error state S5 is driven by pressure altitude estimation S instead of the corrected inertial position P in order to minimize startup residuals. (NOTE : IT APPEARS FROM THE FIGURE FROM THE DASHED LINE THAT THE CORRECTED INERTIAL ALTITUDE DRIVES THE PRESSURE OFFSET ERROR, IS THAT CORRECT? DOESN'T THE PRESSURE ALTITUDE DRIVE THE PRESSURE OFFSET ERROR ?] The Kalman filter utilizes these six errors states to determine the Kalman gain and state correction value in determining portion 200, where this correction value is then used to modify the altitude estimation to produce a more accurate determination of altitude. A description of the error states including their typical initial covariances and typical correlation times T are shown in the following table: State No. Symbol State Definition Initial Correlation Covariance Time(hrs 1 S1 Vertical Position Error (2e3 m)2 N/A 2 S2 Vertical Velocity Error (0. 1 m/s) NUA 3 S3 Stabilized Vertical Accel (le-3 m/s2)2 1. 5 Bias 4 S4 Barometer Bias (100 m)2 5 5 S5 Barometer Scale Factor (0. 01) 2 5 6 S6 Pressure Offset (0. 125 mb) 20

[NOTE: HOW IMPORTANT IS THE ABOVE TABLE TO THE UNDERSTANDING OF THE PRESENT INVENTION ? ARE THESE JUST EXAMPLE VALUES OR DO ANY OF THE ABOVE VALUES HAVE ANY SIGNIFICANCE ? IS THE PRESENT INVENTION LIMITED TO THESE VALUES?] [NOTE: HOW IS THE KALMAN FILTER PHYSICALLY IMPLEMENTED ? IS IT AN ACTUAL HARDWARE DEVICE RECEIVING INPUTS FROM THE SENSORS ? OR IS IT JUST A MODEL FOR FUNCTIONS PERFORMED IN WITHIN A CPU IN THE INERTIA NAVIGATION SYSTEM ?] A barometric altimeter with a pressure offset produces an altitude error that grows with altitude. Using the equation for the standard day pressure altitude, for a pressure offset PE, the altitude error can be represented by the equation:

where K6 = 1-, K1=44. 342, K2=0. 190263, K4=14.605, PO = 1013.25 [mb], SB = 11 [km], and S is the pressure altitude estimation. The pressure offset PE may have a predetermined value or may be determined by calibration techniques or other similar techniques commonly known to those skilled in the art.

At higher altitudes as well as for large pressure offsets Pe, the correction value generating formula f (S) is not linear and the altitude error ASR can differ greatly from the pressure offset error ASp. For a typical value of the pressure offset PE approximately equal to 0.125 [mb] at a true altitude S=30 km, #SR/#SP # 4.1. The correction value generating formulae) produces a better estimate for the pressure offset error ASp when the altitude S is assumed to be less than SB. For S < SB, #SR/#SP # 2. 3 for typical values used in the equations. While a mismatch in the observation element which is a function of altitude may exist, the Kalman

filter mechanization of FIG. 2 will converge to the appropriate pressure offset PE to remove the pressure offset error ASp.

The pressure sensor in the barometric altimeter also has a pressure noise associated with it that the present invention models as an observation noise. For pressure sensor noise with a given standard deviation, the effects of this pressure sensor noise on the altitude determination are more pronounced at higher altitudes.

As an aircraft increases its altitude, a given change in pressure will result in a larger change in altitude. Thus, the error associated with the pressure sensor noise increases with altitude. The present invention accounts for this altitude dependent error by estimating the amount of pressure noise and converting it into an altitude noise that is a function of both altitude and pressure noise.

Referring now to FIG. 3, an operational block diagram of a preferred method of modeling the observation noise in accordance with the present invention is illustrated. Initially in step 300, a pressure noise having a standard deviation ap is estimated for the barometric altimeter. The standard deviation ap of the pressure noise is then converted into a standard deviation (SE for the observation error for the altitude noise in step 302. The standard deviation crE of the observation error is determined by substituting sP=PE and s=ASRinto the altitude error equation, so that:

where Kl=44. 342, K2=0.190263, K4=14. 605, PO = 1013. 25 [mb], SB = 11 [km], and S is the pressure altitude estimation. The observation error is thus a function of both pressure noise op and altitude S, which enables the altitude noise can be more closely approximated than previously achievable. By way of example only, for a pressure noise having a standard deviation aP=0. 025 [mb] and a pressure altitude S=30 [km], the standard deviation of the observation error due to pressure noise 6E=300 [m]. [NOTE: DOES THIS RESULT SHOW AN ADVANTAGEOUS FINDING, OR IS MERELY PLUGGING VALUES INTO THE EQUATIONS?] Finally in step 304, the observation error is utilized to modify the altitude estimation to account for the altitude noise. The system and method of the present invention models noisy barometric altimeters through the use of an additional pressure offset state and an observation noise that is a function of pressure noise and altitude. [WHAT ARE THE ADVANTAGES OF MODELING OBSERVATION NOISE AS A FUNCTION OF PRESSURE NOISE AND ALTITUDE?]

[NOTE: PLEASE ENSURE THAT THE FOLLOWING SECTION DESCRIBING THE GRAPHS IS NECESSARY TO THE PRESENT APPLICATION. IF SO, IT WILL NEED TO BE EXPLAINED IN GREATER DETAIL. FOR INSTANCE, THE RESULTS OF THESE GRAPHS SEEM TO MERELY SHOW AN IMPROVED INERTIAL VELOCITY ERROR. HOWEVER, THE REST OF THE APPLICATION DESCRIBES IMPROVING THE ACCURACY OF THE ALTITUDE DETERMINATION, AND THESE RESULTS DON'T SHOW THAT. A MORE PRECISE DESCRIPTION WILL BE REQUIRED IF THESE RESULTS ARE IMPORTANT IN SHOWING THE IMPROVED EFFECTS OF THE PRESENT INVETNION] : Simulations suggest that for typical scale factor, barometer offset, pressure offset, and pressure noise values, velocity errors for a variety of flights up to 100,000 feet are better than 0.15 m/s RMS. An example of an X-33 profile with multiple climbs and dives is shown in FIGS. 4A-4C. FIG. 4A illustrates the aircraft velocity, FIG. 4B illustrates the aircraft altitude, and FIG. 4C illustrates the platform attitude. The inertial velocity error for this profile is shown in FIG. 4D, while the vertical position error is shown in FIG. 4E. Dashed lines 400 and 402 in FIGS. 4D and 4E, respectively, illustrate the square root of the covariance for the given state. The same profile illustrated in FIGS. 4A-4C was simulated with a pressure offset error state, where the inertial velocity error and vertical position

error were found to be significantly improved, as illustrated in FIGS. 5A and 5B, respectively. On average, the pressure offset state for typical atmospheric conditions and high altitude scenarios improved the inertial velocity error 2.5/1.

The system and method of the present invention improves the accuracy of altitude determinations made using pressure measurements taken by a barometric altimeter by directly modeling a pressure offset error and by modeling an observation noise that is a function of pressure noise and altitude. The Kalman filter mechanization of the present invention may be implemented in hardware or by software in the inertial navigation system. [NOTE: IS THIS CORRECT?] As can be seen from the foregoing, a system and method for improving the accuracy of altitude determinations in an inertial navigation system formed in accordance with the present invention is provided by directly accounting for errors in measurements taken by the barometric altimeter itself in the form of pressure offset error and observation noise.

In each of the above embodiments, the different structures of the system for improving the accuracy of altitude measurements in an inertial navigation system of the present invention are described separately in each of the embodiments.

However, it is the full intention of the inventors of the present invention that the separate aspects of each embodiment described herein may be combined with the other embodiments described herein. Those skilled in the art will appreciate that various adaptations and modifications of the just-described preferred embodiment can be configured without departing from the scope and spirit of the invention. For instance, the teachings of the present invention are not intended to be limited to inertial navigation systems, where the teachings of the present invention can be extended to other applications which utilize pressure measurements to generate an altitude determination. Therefore, it is to be understood that, within the scope of the appended claims, the invention may be practiced other than as specifically described herein.