Method for determining a navigation solution of a navigation system having a terrain-navigation module and a navigation system

The invention concerns a method for determining a navigation solution of a navigation system with a terrain navigation module, as well as a navigation system.

In DH Titterton and JL Weston: Strapdown inertial 10 navigation technology, Peter Peregrinus Ltd. 1997, it is presented how it can be built with the help of a Strap Down algorithm, which integrates accelerations and rotational speeds, an inertial navigation system. This inertial navigation system represents the basis for a navigation system that determines the position, speed and location of a vehicle. Since the integration of inertial sensor data affected by errors does not allow a stable long-term navigation solution, other sensors are merged, such as the global positioning system (GPS), which allows an absolute position determination, with the solution of the inertial navigation system. In Mohinder S. Grewal and Lawrence R. Weill and Angus P. Andrews: Global Positioning Systems, Inertial Na-vigation, and Integration, John Wiley & Sons, Inc., 2001, 25 describes how to merge using a stochastic filter, especially a Kalman filter, the inertial navigation solution with other sensor data, especially GPS. Due to the dependence of GPS on external satellite signals, it is easily possible that this system can be disturbed intentionally or unintentionally. For this reason, other sensor signals are used to merge with the inertial navigation solution. A sensor signal of this class is generated by a

terrain navigation module based on distance measurements with respect to the ground (radar, laser, so-nar ...), in what follows called height measurement. This is done by comparing the height measurements of the terrain obtained from the 5 height measurements with a reference map containing the actual heights of the terrain. There are several approaches to how a terrestrial navigation system of this kind can be constituted. In F. Gustafsson, F. Gunnarsson, N. Bergman, U. Forssel, J. Jansson, R. Karlsson, P. Nordlund: 10 Particle Filters for Positioning, Navigation, and Trac-king, in: IEEE Transactions on Signal Processing, 50, 425 -435, 2002, a group based on non-linear stochastic filters, especially particle filters, is described. However, this group has some drawbacks with respect to its modularity (possibility of separate development of the navigation system and the terrain navigation module) and the acquisition performance (possibility of correcting large initial position errors) . The second group is based on comparison procedures that perform a direct comparison of measured values ​​of terrain height and reference map, and is described in JP Goldon: Terrain contour matching (TERCOM): a cruise missile guidance aid, in: Proceedings of the SPIE Image Processing for Missile Guidance, volume 238, 1980. 25 These comparison procedures are modular and have excellent acquisition and follow-up capabilities. The drawback is that this terrain navigation module does not provide any information on the quality of the own position assistance. Therefore, the 30 current procedures, as described in J. Metzger, O. Meister, GF Trommer, F. Tumbrägel, B. Taddiken: Coo-variance Estimation for Terrain Referenced Navigation with a Comparison Technique, in: Proceedings of the ION 60th

Annual Meeting, June 7-9, Dayton, Ohio, USA, 2004, employ heuristic procedures to determine the quality needed for the fusion. In particular, in a heuristic procedure of this class, the scabrosity of the land is evaluated and the expected quality of the position assistance is estimated from it. However, since, apart from scabrosity, other factors, such as the quality of the IMU, the quality of the reference map and the comparison algorithm itself, have an influence on the quality of the position assists, the 10 quality determined is only a suboptimal solution. A second procedure, such as that described in J. Metzger, GF Trommer: Improvement of Modular Terrain Navigatión Sys-tems by Measurement Decorrelation, in: Procedings of the ION 59th Annual Meeting, June 23-25, Albuquerque, New Mexico 15, USA, 2003, estimates the variance of position assists from several past position assists. However, this presupposes a stationary ergodic noise process for the errors of fixed position values, which is not applicable in most cases. Therefore, this description of quality is also only suboptimal. At this point the nucleus of the invention begins, which allows a clearly more accurate indication of the quality of the position assists. 25

The problem of the invention is to provide a navigation method and system for determining a navigation solution of a navigation system with a terrain navigation module, with which it is possible to determine the navigation solution. 30 tion, especially the position, speed and situation with improved quality.

This problem is solved with the characteristics of the independent claims. Other ways of rea-

They are indicated in the subordinate claims referring to them.

According to the invention, a procedure is provided for determining a navigation solution of a navigation system with a terrain navigation module with the following steps:

<dl>&lt;dt /&gt;<dd>- determination of a navigation solution, an assisted position solution and the relative path traveled between two height measurements with the help of navigation sensors, a Strap Down module and a navigation filter, </dd></dl>

<dl>&lt;dt /&gt;<dd>- obtaining a quality of the respective current assisted position from the respective current assisted position by means of a first quality function, 15 </dd></dl>

<dl>&lt;dt /&gt;<dd>- formation of a search sector on the basis of predetermined criteria and on the basis of a respective pre-determined assisted position solution and formation of predetermined positions within the search sector, 20 </dd></dl>

<dl>&lt;dt /&gt;<dd>- storage of a current measurement of the height of the land and the relative path traveled between the current measurement of the height of the land and the measurement of the height of the land recently stored for use in obtaining the quality of a respective path relatively traveled and of the quality of a respective measured storage of ground height, </dd></dl>

<dl>&lt;dt /&gt;<dd>- determination of a comparison position for a respective position within the search sector and a respective relatively retracted path stored, </dd></dl>

<dl>&lt;dt /&gt;<dd>- obtaining a quality - transformed by means of the reference map - of a respective path </dd></dl>

<dl>&lt;dt /&gt;<dd>Relatively stored path through a second quality function using a respective relatively stored path. </dd></dl>

<dl>&lt;dt /&gt;<dd>- obtaining the quality of a respective stored height measurement by means of a third quality function using the corresponding stored height measurement, </dd></dl>

<dl>&lt;dt /&gt;<dd>- obtaining a quality of a respective reference height for a respective comparison position from predetermined parameters and 10 of the corresponding reference height by means of a fourth quality function, </dd></dl>

<dl>&lt;dt /&gt;<dd>- determination of distribution functions of the error measures for a respective comparison position from a function of quality 15 of the respective current assisted position, the quality - transformed by means of the reference map - of a respective road relatively re-run, the quality of a respective measured storage of ground height and the quality of a respective reference height, </dd></dl>

<dl>&lt;dt /&gt;<dd>- obtaining a total error distribution function for a respective predetermined position within the search sector based on the distribution functions obtained from the error measures 25, </dd></dl>

<dl>&lt;dt /&gt;<dd>- obtaining the minimum probabilities of all positions used in the search sector by means of a function of the distribution functions of all total errors as a probability with which a respective position of the search sector presents the minimum total error, </dd></dl>

<dl>&lt;dt /&gt;<dd>- determination of the quality of position assistance through a function of all </dd></dl>

<dl>&lt;dt /&gt;<dd>minimum probabilities, </dd></dl>

<dl>&lt;dt /&gt;<dd>- assistance of the navigation solution obtained in the Strap Down module by means of the navigation filter with the help of the determined position assistance and the determined quality of the 5 position assistance. </dd></dl>

The quality of the respective current assisted position can be obtained from the respective current assisted position by means of a first quality function with the aid of a first distribution function.

The first quality function may be a Gaussian distribution function with a first variance.

In the formation of the search sector, a search sector centered around an assisted position solution can be used.

In the formation of the search sector, a magnitude of the search sector can be used that varies based on the inaccuracy of a respective assisted position solution. Alternatively, in the formation of the search sector, positions that are equidistant on a predefined frame can be used.

In storage, the current ground height measurement can be obtained from the current height measurement and the current assisted position. 25

In the storage of a current measurement of the height of the land and the relative path traveled, a preset number of measurements of ground height and relative paths traveled can be stored.

In order to obtain the quality - transformed by means of the reference map - a second distribution function can be used for the relatively re-run path stored. In order to obtain the transformed quality through the reference map, you can use

Mainly a Gaussian distribution function with a second variance.

In order to obtain the quality of a respective stored height measurement, a third distribution function can be used. In this obtaining of quality, a Gaussian distribution function with a third variance can be used especially.

In order to obtain the reference height quality, a fourth distribution function can be used. In obtaining this quality, a Gaussian distribution function with a fourth variance can be used.

The total error distribution functions can be obtained for a respective predetermined position within the search sector by folding the different distribution functions of the 15 error measures.

The quality of the position assistance can be determined through a function of all the minimum probabilities by means of a discrete distribution function that is defined by the minimum probabilities. Alternatively, the quality of position assistance can be determined through a covariance matrix that is calculated from the minimum probabilities and the actual position assistance. 25

The navigation solution obtained in the Strap Down module can be assisted by the following steps:

<dl>&lt;dt /&gt;<dd>- determination of position assistance through the calculation of the error measure for the respective comparison position, </dd></dl>

<dl>&lt;dt /&gt;<dd>- determination of the total error for the respective position in the search sector and </dd></dl>

<dl>&lt;dt /&gt;<dd>- conduct a search for the minimum total error </dd></dl>

<dl>&lt;dt /&gt;<dd>through all the total errors in the search sector for the identification of position assistance. </dd></dl>

Also, according to the invention, a navigation system is provided with a Strap Down module and a navigation filter and with a terrain navigation module to determine a navigation solution, where the power module is fed terrain navigation an assisted position solution and relative paths traveled between two height measurements, the ship-10 module covering the following functions:

<dl>&lt;dt /&gt;<dd>- a function for obtaining (M) a quality (14) of the respective current assisted position (1a) from the respective current assisted position (1a), 15 </dd></dl>

<dl>&lt;dt /&gt;<dd>- a function for the formation (H) of a search sector and for the formation of predefined positions (7) within the search sector, </dd></dl>

<dl>&lt;dt /&gt;<dd>- a function for the storage (J) of a current measurement of the height of the land and the road 20 relative path between the current measurement of the height of the land and the measurement of the height of the land recently stored (1b) a in order to use these to obtain the quality (15) of a respective relatively traveled path and the quality 25 (16) of a respective ground height measurement, </dd></dl>

<dl>&lt;dt /&gt;<dd>- a function for the determination (K) of a comparison position (8) for a respective position (7) within the search sector and of a respective 30 path relatively traveled stored (10), </dd></dl>

<dl>&lt;dt /&gt;<dd>- a function for obtaining (P) a quality - transformed by means of the reference map (6) </dd></dl>

<dl>&lt;dt /&gt;<dd>- of a respective relatively stored path (10) by means of a second quality function using a respective relatively stored path (10), </dd></dl>

<dl>&lt;dt /&gt;<dd>- a function for obtaining (Q) of a quality 5 (16) of a respective measurement of stored ground height (11) by means of a third quality function using the corresponding measurement of stored ground height (11) , </dd></dl>

<dl>&lt;dt /&gt;<dd>- a function for obtaining (R) of a quality 10 (17) of a respective reference height (6) for a respective comparison position (8) from predetermined parameters and the corresponding reference height (6 ) by means of a fourth quality function, 15 </dd></dl>

<dl>&lt;dt /&gt;<dd>- a function for the determination (S) of distribution functions (18) of the error measurements for a respective comparison position (8) from a function of the quality of the respective current assisted position (14), the quality - trans-20 formed through the reference map - of a respective relatively traveled path (15), the quality of a respective measurement of stored ground height (16) and the quality of a respective reference height ( 17), 25 </dd></dl>

<dl>&lt;dt /&gt;<dd>- a function for obtaining (T) of a distribution function (19) of the total error for a respective predetermined position within the search sector (7) based on the distribution functions obtained (18) of the measurements of 30 error, </dd></dl>

<dl>&lt;dt /&gt;<dd>- a function for obtaining (U) the minimum probabilities (20) of all positions used (7) in the search sector by means of </dd></dl>

<dl>&lt;dt /&gt;<dd>a function of the distribution functions (19) of all total errors as a probability with which a respective position of the search sector presents the minimum total error, </dd></dl>

<dl>&lt;dt /&gt;<dd>- a function for determining (V) the quality of position assistance (4b) through a function of all minimum probabilities (20), </dd></dl>

<dl>&lt;dt /&gt;<dd>- a function for the assistance of the navigation solution obtained in the Strap Down module (C) through the navigation filter (D) with the help of the determined position assistance (4a) and the determined quality of the assistance of position (4b). </dd></dl>

The essential advantage of the invention over current processes is based on the improvement of the acquisition and monitoring properties, the increase in the integrity of the navigation system and the increase in the modularity of the system. navigation and terrain navigation module. twenty

In the following the invention is described with the help of the attached figures, which show:

Figure 1, a functional representation of the functions of the navigation system with a terrain navigation module and

Figure 2, a development diagram of the essential functions of the navigation system according to the invention with the terrain navigation module.

The navigation system Z according to the invention comprises a terrain navigation module A with which it is possible to determine the navigation solution 1 of a flying device or of a water vehicle and especially of a submarine. This vehicle is abbreviated as a vehicle in the following. The solution of

Navigation necessarily includes the vehicle's own position. In addition, the navigation solution may contain speed and situation, but these are not necessarily necessary.

For this purpose, measurement signals or measurement values ​​for accelerations and rotational speeds 2 that are obtained by an inertial sensor system or an inertial measurement unit B are fed to the navigation system Z 5. These signals or input data are processed to provide a navigation solution 1 of the vehicle, an assisted position solution 1a and the relative path 1b traveled between two height measurements with the aid of navigation sensors A, B, Ga, Gb, a Strap Down C module and a D navigation filter. In the field navigation module, the solution of position 1a - merged from all sensor and assisted signals - is used here as a positional hypothesis for the calculation of position assistance 4a and the quality of the position assistance 4b. In addition, the respective relative recovery path 1b of the vehicle between two height measurements, which is determined by the Strap Down C module from the accelerations and rotation speeds 2, is processed in the terrain navigation module. .

For a long-term stable determination of the navigation solution 1, the processing of additional sensor signals is necessarily necessary. The fusion of the data of different sensors is achieved by means of the navigation filter D, which, based on the sensor signals, provides corrections 3 for the navigation solution 1 that is calculated in the Strap Down C module. 30 Typically, a Kalman filter is used for the navigation filter D, other non-linear stochastic filters can also be used without restrictions.

The terrain navigation module A provides as

Additional sensor signal 4a for the navigation filter D is the three-dimensional position of the vehicle, which is needed as position assistance for long-term stable navigation. In addition, the terrain navigation module provides information on the quality of position assistance 4b.

As input quantities, the position data of the navigation solution is used as the hypothesis for the actual position. Likewise, the relative navigation module A 10 relative retracted paths 1b is fed so that the road traveled by the vehicle between two height measurements can be determined. The sensor source for a field navigation is represented by a distance meter E (for example, a radar or laser altimeter for a flying device or a sonar for a water vehicle), which determines the distance 5 of the ground vehicle, in what follows designated as the height measurement.

Other sensors, such as GPS Ga or barometric G20 altimeters, can also be processed in the navigation filter, but are not necessarily necessary.

The terrain navigation module is broken down in Figure 2. The module is divided into an area to determine position assistance Aa and an area to determine the quality of position assistance Ab.

The storage J of a current ground height measurement and the relative path traveled between the current ground height measurement and the recently stored ground height measurement 30 1b is common to both areas. In storage J, the current ground height measurement is obtained from the current height measurement 5 and the current assisted position 1a. These stored values ​​are used in the following to obtain

the quality 15 of a respective relatively covered road and the quality 16 of a respective height measurement of stored land. In addition, stored values ​​are used for the determination of position assistance 4a through the calculation L of the error measure 12 5 for the respective comparison position 8 and for the determination M of the total error 13 for the respective position in the search sector 7 and for the search of the minimum total error N through all the total errors in the search sector for the identification of the position assistance 4a. The identification of the position assistance 4a can be carried out because the position in the search sector 7 presenting the minimum total error 13 corresponds to the maximum probability of the actual position. This is based on the fact that the 15 minimum total error 13 also leads to the maximum similarity between the ground height measurements 11 determined from the height and stored measurements 5 and the corresponding reference heights 6. For the following determination of the position assistance 4a and 20 of the quality of the position assistance 4b it is sufficient that a preset number of measurements is stored in the storage J of a current ground height measurement and the relative path traveled of height of land and relative paths 1b. 25

Also, the formation of the search sector is common to both areas. The formation H of a search sector on the basis of predetermined criteria and on the basis of a respective predetermined assisted position solution 1a and a formation of pre-completed positions 30 within the search sector is necessary to determine position assistance 4a and quality of position assistance 4b. The search sector de-fine positions 7 that are taken into consideration for

the identification of position assistance 4a. In particular, a search sector can be employed that is centered around an assisted position solution 1a and that allows search sector magnitudes that vary based on the inaccuracy of a respective assisted position solution 5a. Positions 7 that are equally arranged on a predefined frame can be used for the search sector.

For comparison between the measured values ​​of past ground height 11 and the corresponding reference heights 6, as well as for the determination of quality 17 of the corresponding reference heights 6, a comparison position is required 8 based on a respective position 7 in the search sector and on the corresponding path relatively traveled to 15 thru 10. For this reason, a comparison position 8 is determined for a respective position 7 within the search sector and a respective relatively traveled path 10.

The development of the quality determination of the position assistance Ab begins with the obtaining of a quality 14 of the respective current assisted position 1a from the respective current assisted position 1a by means of a first quality function. The quality function is needed to describe the influence of a hypothesis of affected position of error, that is, an assisted position 1a affected of error. Obtaining the quality function can be carried out especially with the aid of a first distribution function that can in turn be described especially by means of a first deposit. In the description by means of a first variation, the distribution function is limited to a Gaussian distribution.

For each position 7 in the search sector and for

each comparison position 8 based on this position uses the quality of a respective relatively traveled path 10 in order to determine the influence of the error of the corresponding relatively retracted path 10 and express it by means of a quality. 5 This is done through the determination P of a quality - transformed with the aid of the reference map 6 - of a respective relatively traveled path 10 by means of a second quality function using a respective relatively traveled path 10 10. The error of the corresponding relatively re-run stored path 10 occurs mainly due to the errors of the acceleration and rotation speed values ​​2 and the processing by the Strap Down C module. For quality - transformed with the help of the reference map 15 6 - of the relatively stored path 10 a second distribution function can be used in particular. It is also possible to approximate this distribution function through a Gaussian distribution and determine it through a second variance. For small errors of the relative paths stored 10, the determination of the transformed quality with the help of the reference map 6 takes place by means of the linearization of the reference map, and for larger errors said determination has place through a statistical evaluation of the reference heights depending on the errors of the relative paths traveled 10.

For each measurement of stored ground height 11 a quality is obtained to determine the influence of the errors of the ground height measurement on position assistance 4a and on the quality of position assistance 4b. This is done through the obtaining Q of a quality 16 of a respective measurement of

height of stored land 11 by means of a third quality function using the corresponding height measurement of stored land 11. This quality function of a respective height measurement of stored ground 11 can be especially described by means of a Third distribution function. In particular, by limiting the distribution function to a Gaussian distribution function with a third variance, the quality function can be determined.

For all comparison positions 8, the influence of the errors of the corresponding reference heights 6 on position assistance 4a and on the quality of position assistance 4b is determined by means of a quality function. This is done by determining R of a quality 17 of a respective reference height 6 for a respective comparison position 8 from predetermined parameters and the corresponding reference height 6 by means of a fourth function of quality. This quality function of a reference height can be described especially by means of a fourth distribution function. In particular, the quality function can be determined by limiting the distribution function to a Gaussian distribution function with a fourth variance.

From the quality functions determined 14, 25 15, 16 and 17, the influence of all errors of the input quantities 1a, 1b, 5 and 6 on the error of error measure 12 can be determined and can be expressed this influence by means of a fifth distribution function 18. This determination S of a fifth distribution function 18 of the error measurement for a respective comparison position 8 is made from a function of the quality of the respective current assisted position 14, the quality - transformed with reference map help

recency - of a respective relatively traveled path 15, the quality of a respective measurement of stored ground height 16 and the quality of a respective reference height 17. This means that for each comparison position 8 a distribution function is determined 18 5 of the error measure. For an error measure 12 based on the absolute difference and on the representation of the different quality functions by means of Gaussian distribution functions, a density of absolute value results for the distribution function 18 of the error measure. The absolute value density is described by making it zero for negative arguments and corresponding, for positive arguments, to the sum of the original Gaussian distribution and the Gaussian distribution reflected on the Y axis. 15

From the determination of the distribution function 18 of the error measurement, the way in which the error of the error measurement 12 is described stochastically is known. Corresponding to obtaining the total error 13, it can be determined from of this the sixth 20 distribution function 19 of the total error. The obtaining T of a sixth distribution function 19 of the total error for a respective predetermined position within the search sector 7 based on the fifth distribution functions obtained 18 of the error measurements can be carried out especially by folding the different -New distribution functions of error measures 15 when obtaining the total error 13 takes place as a weighted or unweighted sum of all error measures 12. The sixth distribution function 19 of the total error 30 describes the stochastic properties of the total error 13 which is directly used for the determination of the position assistance 4a. For this reason, stochastic properties and, therefore, the quality of assistance

The position 4b can also be deduced from the sixth distribution function 19 of the total error.

For the stochastic description of the quality of position assistance 4b, it is determined for each of the positions 7 in the search sector, from the 5 distribution functions 19 of the total error of all positions 7 in the search sector , the probability with which the corresponding position in the search sector presents the minimum total error 13 and, therefore, the maximum similarity between measurements of ground height al-10 stored 11 and reference heights 6. Obtaining U of these minimum probabilities 20 of all positions used 7 in the search sector by means of a function of the distribution functions 19 of all total errors as a probability with which a respective position of the sector of search presents the minimum total error represents a stochastic description of the position assistance 4a. The minimum probability 20 is determined especially by integrating the cumulative probability that all other error measures are greater than 20 than a parameter X, multiplied by the density function of the error measure itself in the point X. This is indicated in equation 0.1, where Pmin represents the minimum probability for position 7 with the index mn in the search sector, fQmn describes the distribution function of the total error 19 for position 7 with the index mn in the search sector and fQkl indicates all other distribution functions of the total error 19 for the positions with the index kl in the search sector. The magnitude of the search sector is indicated by means of M and N.

Therefore, a high minimum probability 20 means a high probability that the corresponding position 7 within the search sector corresponds to the actual position.

The determination of the quality of the assistance of 5 position 4b with the aid of a function of all minimum probabilities 20 can be carried out especially by means of a discrete distribution function that is defined by the minimum probabilities 20. This discrete distribution function can be processed directly on a navigation filter D. Likewise, it is possible to represent the quality of position assistance 4b with the help of a covariance matrix that is calculated from the minimum probabilities 20 and the actual position assistance 4a. This covariance matrix represents an approximation 15 of quality as it can be directly processed in a special navigation filter D, a Kalman filter.

The assistance of the navigation solution obtained in the Strap Down C module through the navigation filter-20 tion D with the help of the determined position assistance 4a and the determined quality of the position assistance 4b closes the recursion and It allows long-term stable navigation.