DE102009042969A1 - Method and device for determining the position of a watercraft - Google Patents

Abstract

The invention relates to a method for determining the position of a watercraft by comparing depth values 14 measured with a sonar system with known depth values, a known depth value being assigned to a grid point 36 of a rasterized digital terrain model of a known sea area Grid point 36 of a grid model of a terrain section assigned 16, the grid width of which is adapted to the grid width of the digital terrain model. To determine a match, a plurality of parameters are defined 30, with a match measure characterizing the match being determined 32 for each parameter. Depending on the match measures, overall match measures are formed 58, which have at least one extreme value indicating a maximum match, its associated position in the digital terrain model is determined 60 as the position or a possible position of the watercraft. The invention also relates to a device for carrying out such a method.

Description

The invention relates to a method and a device for determining the position of a watercraft according to the preamble of claim 1 or 6.

The position determination in the case of surface water vehicles is based primarily on satellite navigation systems, such as, for example, the Global Positioning System (GPS), in which the transit times of the satellite signals between the navigation satellites and the receiving antenna are determined at a known propagation speed. This results in a distance of the navigation system to the respective satellite. Due to the known position of the satellites, a position determination can be unambiguous.

A disadvantage of a position determination by means of GPS, however, is that the position determination of a satellite navigation receiver can be disturbed at least temporarily by known deception systems for transmitting counterfeit satellite signals to foreign satellite receivers and a correct position determination is thus prevented.

Furthermore, the position determination by means of GPS has the disadvantage that the GPS signal is absorbed very quickly under water and thus is not capable of propagation. As a result, submerged vessels such as submarines, AUVs or underwater bodies are out of contact with GPS sensors that would allow them to accurately locate them.

Known navigation systems for underwater vehicles generally have an inertial measuring device, which, however, allows accurate navigation only for a short time. Therefore, a support by additional sensors is necessary such as. A Doppler log and a heading sensor. The disadvantage of this navigation, however, lies in the summation of the error. The error that grows over time prevents a precise determination of the position of the vessel.

The invention is therefore based on the problem to improve the position determination of vessels, regardless of known navigation systems, in particular satellite navigation systems.

The invention solves the problem by a method having the features of claim 1 and a device having the features of claim 6.

If a watercraft has a sonar system for measuring depth values, it is possible to determine the position of the vessel by means of a comparison of measured depth values with known depth values of a sea area.

The known depth values are taken, for example, from a depth map, in particular a bathymetric map or a sea chart, or are present as previously measured comparison depth values. In the following she will be called terrain data. The depth values measured directly by the sonar system are referred to below as search pattern data.

The terrain data obtained by digitizing the depth lines of a map or the measured comparative depth values are present as x, y, z triples distributed irregularly in space. In order to obtain an overview of the entire sea area to be considered, a nationwide representation, for example, in the form of a digital terrain model (DTM) is needed.

A DTM is a digital, numerical model of terrain elevations or terrain depths and terrain forms. It approximates the terrain surface by a limited number of regularly arranged bases, so-called x, y, z-triples. These are usually given as a square grid with equal distances in the x and y directions. Two definitions of a spatial reference are possible. An x, y, z triplet either represents one surface of a grid cell of the DTM or refers to the intersections of the grid lines, the grid points. In the following, as spatial reference, an x, y, z triple is always assigned to a grid point. The square grid has the advantage that logical and algebraic operations are easy to perform. However, the representation as a square grid is not mandatory. Further, for example, a division of the DGM in triangular, general square or hexagonal grid is conceivable.

By means of known interpolation methods, the terrain data are adapted to the square grid of the DGM. Each grid point of the DGM is assigned a depth value. The DGM thus represents a so-called reference model.

The search pattern data of the sonar system represent a terrain cutout or a part of the surface of the river bottom. This terrain cutout is as small as possible the known sea area of the reference model and must be within the known sea area. A so-called raster model is created for this terrain clipping by assigning a measured depth value to each raster point. The raster width of the raster model, which indicates the spacing of the raster points, becomes the grid width of the DTM customized. This adaptation takes place in such a way that the spatial position specification of the raster model defined by rows and columns matches the position specification of the DTM.

The comparison of the terrain data with the search pattern data is carried out for a plurality of defined parameters, such as mean depth, soil structure, deviation from the soil structure, standard deviation, bottom slope in west-east direction and / or bottom slope in south-north direction. The parameters can be maintained for a plurality of position determinations or arbitrarily re-selected for each position determination.

For each parameter, a match score indicative of the compliance of the DTM with the raster model is determined for each predetermined number of raster points of the DTM. The degree of agreement has the form of a percentage agreement, which ranges from 0 to 100 percent.

The position determination of the watercraft preferably does not take place on the basis of a single parameter but on the basis of a combination of a plurality of parameters with respect to overall agreement dimensions. The overall agreement dimensions are formed as a function of the conformity measures. This is done, for example, by means of multiplication of the individual, possibly weighted, correspondence measures and subsequent possible radication or by averaging. Any application-related dependencies are conceivable.

Each halftone dot of the DGM, to which one transmission measure per parameter is assigned, is assigned a total agreement measure. Thus, it is possible to advantageously generate logical links among the conformity measures which allow an operator to recognize to what extent, for example, both parameter 1 and parameter 2 and parameter 3 are involved in determining the position of the watercraft.

From the overall agreement measures present in matrix form, at least one extreme value is determined, which indicates a maximum agreement of the DGM with the raster model. Depending on the predefined dependence of the matching measures, the extreme value may be a minimum or a maximum. Based on the spatial position indication by the row and column system, a position corresponding to the extreme value in the DGM of the known sea area can be determined as the position of the vessel. If more than one extreme value is present, these indicate in each case a possible position of the vessel.

The selection of the at least one extreme value for determining the position of the watercraft can be carried out manually by an operator or automatically by means of suitable data processing. As a result, an automatic position determination is advantageously possible.

In a further embodiment of the invention, the search pattern data is split into a DC component and a variable component. Advantageously, only the variable component is used for the comparison of terrain data with search pattern data. This creates the opportunity to perform the comparison without knowing the absolute depth. Consequently, the comparison of the depth values is independent of the tides of the water.

In a further preferred embodiment of the invention, a part of the matching criteria is determined by means of the linear regression. The results of this regression analysis are a slope of the regression line, a possibly existing Y-offset and a standard deviation. The slope of the regression line thereby gives a measure of how much of the search pattern data is represented in the terrain data, the possibly existing Y offset is a measure of the mean depth and the standard deviation is a measure of the deviation of the terrain cutout from the straight examined sea area. By using such a regression analysis, the advantage of independence of position determination from the absolute depth arises.

In a further preferred embodiment of the invention, the correspondence measures of all parameters and / or the overall correspondence measures are entered in a display grid. This advantageous visual, areal presentation provides the operator with a very clear picture of whether there is a unique location or a variety of possible locations that are as likely as possible.

In a further preferred embodiment of the invention, a measure of the reliability of this position is given in addition to the position of the vessel. For this purpose, those grid points are determined, to which a total agreement measure above a predetermined threshold or value is associated. The smaller the number of these detected halftone dots, the greater the measure of the reliability of the determined position. This provides the operator with an advantageous decision aid to the determined position of the vessel. If, for example, a position determination is carried out in a completely flat terrain, the degree of agreement is almost the same in the entire area and thus provides a low degree of reliability of the determined Position. Furthermore, it is advantageous to provide this reliability value for further processing in order, for example, to create a history of the reliability values.

In a further embodiment of the invention, the depth values measured by means of the sonar system are detected selectively, linearly or areally. Accordingly, various embodiments of the sonar system are conceivable. Depending on which detection of the terrain is required, simple echo sounders or so-called fan solders are used. The search pattern data in the form of measured depth values can therefore be one-dimensional or correspond to a true two-dimensional pattern.

In a further embodiment of the invention, after a unique unambiguous position determination, a search range is defined for subsequent position determinations, wherein the range is determined which is theoretically achievable by the vessel in the time between the position determinations. This advantageously reduces the computational effort.

In a further embodiment of the invention, the position over ground and the speed over ground are determined by a continuous position determination in addition to the position. This has the advantage that all data relevant for navigation is available.

Further advantageous embodiments will become apparent from the dependent claims and from the closer explained with reference to the accompanying drawings embodiments. In the drawing show:

1 a block diagram illustrating the process steps of the invention;

2A -B a terrain data DGM and a raster model of the search pattern data;

3 a representation for illustrating the determination of the conformity measures;

4A -C the division of the search pattern data into a DC component and a variable component;

5A -B the determination of the correspondence measures by means of a linear regression.

1 shows a block diagram for illustrating the method steps for determining the position of a watercraft.

First of all, the depth lines of a known sea area taken from an entrained map are digitized by storing the points associated with the depth lines at specific intervals and providing them with the associated depth values as an attribute. Since these points are distributed unevenly, a DGM is created for an area-wide representation 10 , If the known depth values already exist in the form of a DGM, then corresponding processing is omitted.

The following explanations refer to a DGM in the form of a square grid. However, they are also applicable to other geometry models.

In raster-based DTM, a grid cell forms a primitive. It is limited by grid lines whose intersections or grid points are each assigned a depth value. By means of known interpolation methods, the points of the depth lines are adapted to the square grid. The accuracy of such a geometric description depends on the basic size of the raster cell, a so-called raster width. The grid width is chosen so narrow that prominent structures of the terrain surface are captured by the grid.

Since there are no coordinates defined by coordinates in a square grid, an origin of the grid, an orientation of the grid and a screen ruling must be defined to define the geometry of the grid cells. For geometrical position description, so-called index tuples (i, j) are used, which describe the position of a raster cell with respect to the origin (1, 1) of the raster. Taking these definitions into account, the terrain data is read from the DTM 12 ,

From the watercraft, a systematic hydrographic survey of a watercourse of a sea area is carried out by means of a sonar system.

Sonar (Sound Navigation and Ranging) is a sound measurement technique for locating and surveying objects under water. An active sonar survey is based on an emitted by its own watercraft and then received by the transducer acoustic signal. From the transit time measurement of the sound waves and the propagation speed, the depth is determined. Echosounders, for example, work according to the active sonar principle.

For the measurement of watercourses, however, fanlight systems are usually used instead of echo sounders with single jet. In this case, water sound signals are radiated and received by means of a receiving antenna, which has a plurality of electro-acoustic or opto-acoustic transducer for receiving sound waves and generating electrical reception signals, directionally selective. The receiving antenna is also a Direction generator downstream in which the received signals of the converter, depending on the arrangement of the converter on the antenna delayed and combined into group signals. This ensures that the receiving arrangement has a fan of a plurality of mutually pivoted directional characteristics in a fixed receiving sector. Thus, the measurement of the watercourse becomes more accurate and efficient. The depth values of the individual directional characteristics obtained in this way result in a comprehensive survey.

The inventive method is not limited to a comprehensive survey. Furthermore, it is possible to store the measurement data of a single-jet echo sounder as a punctiform or linear measurement.

The data obtained by means of the sonar system are in accordance with 1 stored as measured depth values 14 , In a further processing step, a raster model is created from these measured depth values, the so-called search pattern data 16 , For this purpose, each grid point is assigned a depth value which is either present directly, was determined by means of known interpolation methods or by means of approximation or filtering. Since the measured values of the sonar system, for example, due to the opening angle of the fan solder, are often not grid-shaped with the same distance, an attribute value is predicted by interpolation at a point at which this value was not measured.

Since there are no points in a raster model that can be described by coordinates, an origin of the raster, an orientation of the raster, and a screen ruling are defined to define the geometry of the raster.

It should be noted that the shape and size of the grid cells, which are defined by the definition of the grid, are adapted to the grid cells of the DGM when the grid model is created. This adaptation takes place in such a way that the grid points of the DGM and the grid model to be compared represent one and the same terrain point. It is conceivable, for example, to increase the raster width of the raster model by a factor of 2 and to use only every second raster point for a comparison.

2A -B each show a grid with a drawn depth line 20 , In 2A is exemplary the DGM of the terrain data and in 2 B the raster model of the search pattern data shown.

To define the geometry of grid cells 22 becomes an origin 24 and set an orientation of the grid. For this purpose, an index tuple (i, j) is defined which indicates the position of a raster cell 20 sets. The row index i is plotted vertically from top to bottom, and the column index j horizontally from left to right. Further, to define the geometry of the grid cells 22 set the screen ruling. This is chosen so narrow that striking structures of the terrain surface are detected by means of the grid. Since the terrain patch represented by the search pattern data is smaller than the sea area, the raster model of the search pattern data is also correspondingly smaller than the DTM of the terrain data. In order to detect coincidence of search pattern data with terrain data, a reference position becomes in the raster model of the search pattern data 26 established.

Based 1 is the next step, a determination of the parameters 30 for comparing the terrain data with the search pattern data. However, this method step is optional, ie the parameters can be maintained for a plurality of position determinations and need not be redefined at each new position determination. The parameters can be, for example, an average depth, a soil structure, deviations from the soil structure, a standard deviation, ground slope in the west-east direction and / or soil gradient in the south-north direction, but also other parameters not previously mentioned are conceivable. If, for example, the absolute depth of the vessel is not known, the soil structure is increasingly used for the comparison of the terrain data with the search pattern data.

Then the match sizes are calculated 32 , This is done for each previously defined parameter. The compliance metrics of the terrain data with the search pattern data are determined for a predetermined number of halftone dots of the DGM, and preferably are in the form of a percent match ranging from 0 to 100 percent.

3 Figure 11 is a diagram illustrating the determination of the correspondence measures for each grid point 36 , In the illustrated DGM, the terrain data as well as the search pattern data to be compared are exemplified for j-columns and i-rows.

For a comparison of the terrain data with the search pattern data will now be to each grid point, starting at the origin 24 of the DGM, a measure of agreement of the value of the reference position 26 of the search pattern data with the value of the grid point of the DTM of the terrain data.

The inventive method is not, however, to a comparison by means of a reference position 26 limited in the search pattern data. Furthermore, for example, a comparison of the terrain data with a contiguous sequence of raster cells or corresponding averaging of the search pattern data is conceivable.

For the determination of the degree of conformity, independent of the absolute depth of the vessel, it is advantageous to split the search pattern data. 4A -C illustrates the division of the search pattern data 40 into a DC component 42 and a variable portion 44 , It is the local course of the depth values over the depth 46 applied.

4A shows an exemplary course of the search pattern data 40 , In order to compare the measured depth values with the known depth values, independent of the absolute depth of the vessel, the in 4B illustrated average depth in the form of a DC component 42 from the search pattern data 40 subtracted. This results in a progression of the variable component 44 the depth value, which is compared with the terrain data and in 4C is shown.

5A 1B show another embodiment for determining the correspondence measures by means of a linear regression. Linear regression is a statistical method to examine relationships between two variables. In this case, the two variables are the depth values of the DGM Z1 and the depth values of the raster model Z2.

In 5A a section of the square grid of the DGM is shown. Each grid point 36 are assigned two values (Z1_i, Z2_i). The halftone dots 36 are numbered from left to right, starting in the first line, with the numbers from 1 to 6. In the second line, the numbers from 7 to 12 again follow from left to right.

In 5B is a selection of the value pairs of the first twelve grid points 36 plotted graphically. The horizontal axis 50 are assigned the depth values of the DGM Z1 and the vertical axis 52 the depth values of the raster model Z2. It is possible to establish a linear relationship between the values. Thus, it is possible a regression line 54 to form according to the regulation Z1 = a · Z2 + b

The parameter a corresponds to the slope of the regression line 54 and the parameter b the intercept. Using linear regression, the parameters a and b can be estimated. They are so determined that the straight line 54 the value points optimally adapted. An optimal solution is the straight line 54 for which the sum of the square vertical distances of the individual points from the line 54 becomes minimal. The thus determined parameters a and b provide values for the correspondence measures of certain parameters. The slope a is a measure of how much of the search pattern data is represented by the terrain data, and the intercept b indicates a measure of the mean depth. Furthermore, a measure for the scattering of the value points around the regression line 54 determined. The associated standard deviation is thus a measure of the uncertainty of these value points and indicates the deviation of the terrain cut-off from the sea area currently being investigated.

According to 1 a formation of overall conformity measures takes place 58 depending on the measures of conformity, by combining the correspondence measures of the parameters. In the combination, for example, the individually weighted agreement measures of the parameters are multiplied. The amount of data resulting from this multiplication can be reduced, for example, by means of radication. However, other possible combinations of the parameters are conceivable, for example an averaging.

As the last method step, the position of the watercraft is determined 60 by means of the overall agreement dimensions. For this purpose, an extreme value of the overall agreement measures is determined, which indicates a maximum match and whose corresponding position in the DGM of the known sea area is determined as the position of the vessel. On the basis of the previously made definitions of the square grid of the DGM with respect to the origin and the raster width, namely, the positions of the individual raster cells can be converted into a Cartesian coordinate system. It is also conceivable to use the spatial reference systems specially defined from a cartographic point of view for indicating the position of the watercraft. If there are several extreme values or values indicating a maximum match, the position of the vessel determined from this is a possible position.

In a variant of the method, a measure of the reliability of this position information is determined in addition to the position of the watercraft. The smaller the number of total coincidence amounts that is above a predetermined value, the greater the measure of the reliability of the position determination. This measurement is displayed to the operator together with the determined position. Furthermore, it is conceivable to store this measure for further processing in order, for example, to refer to areas after repeated positioning in which the specified position was determined with less reliability.

In order to provide the operator with an advantageous decision-making aid, the method described above is linked to a visual representation of the correspondence measures of all parameters and / or of the overall agreement parameters. The representation of these values takes place in a display grid which is adapted to the number of values by assigning to each grid point a measure of conformity or a total degree of agreement. The rows and columns represent the geometric position of the assigned value. Such a representation provides the operator with a very clear picture of whether there is a unique location or a variety of possible locations.

All features mentioned in the aforementioned functional description, in the claims and in the introduction to the description can be used both individually and in any combination with one another. The invention is thus not limited to the described or claimed feature combinations. Rather, all feature combinations are to be regarded as disclosed.

Claims (10)

Method for determining the position of a vessel, which comprises a sonar system for measuring depth values ( 14 ), by means of a comparison of measured depth values with known depth values, whereby a rasterized digital terrain model ( 10 ) with a plurality of grid points ( 36 ) of a known sea area, wherein in each case one grid point ( 36 ) is assigned a known depth value, characterized in that a raster model of a terrain cutout is created ( 16 ), each with a grid point ( 36 ) of the raster model is assigned a measured depth value and the raster width of the raster points ( 36 ) of the raster model to the raster width of the raster points ( 36 ) of the digital terrain model, a plurality of parameters for determining a match are defined ( 30 ), for each parameter, a match measure indicative of the compliance of the digital terrain model with the raster model for each predetermined number of raster points ( 36 ) of the digital terrain model ( 32 ), Overall agreement measures are formed as a function of the degree of agreement of the parameters ( 58 ), at least one extreme value of the overall agreement measures is determined, which indicates a maximum match and whose corresponding position in the digital terrain model of the known sea area is determined as the position or a possible position of the vessel ( 60 ). Method according to Claim 1, characterized in that for at least one parameter the measured depth values are converted into a DC component ( 42 ) and a variable component ( 44 ) and from the variable component ( 44 ) of the measured depth values the conformity measures are determined ( 32 ). A method according to claim 1 or 2, characterized in that for at least one parameter, the correspondence measures are determined by means of a linear regression ( 32 ). Method according to one of the preceding claims, characterized in that the conformity measures of all parameters and / or the overall agreement measures are entered into a display grid, in particular for the visual representation of the conformity measures and / or the overall agreement measures. Method according to one of the preceding claims, characterized in that a measure of a reliability of the determined position of the vessel is specified. Device for determining the position of a vessel which has a sonar system for measuring depth values ( 14 ), by means of a comparison of measured depth values with known depth values, whereby a rasterized digital terrain model ( 10 ) with a plurality of grid points ( 36 ) of a known sea area, wherein in each case one grid point ( 36 ) a known depth value can be assigned, characterized in that a raster model of a terrain cutout can be created ( 16 ), each with a grid point ( 36 ) of the raster model, a measured depth value can be assigned and the raster width of the raster points ( 36 ) of the raster model to the raster width of the raster points ( 36 ) of the digital terrain model, a plurality of parameters for determining a match are defined ( 30 ), for each parameter, a match measure indicative of the compliance of the digital terrain model with the raster model for each predetermined number of raster points ( 36 ) of the digital terrain model can be determined ( 32 ), Overall agreement dimensions can be formed as a function of the degree of agreement of the parameters ( 58 ), at least one extreme value of the overall agreement measures can be determined, which indicates a maximum match and whose corresponding position in the digital terrain model of the known sea area can be determined as the position or a possible position of the vessel ( 60 ). Apparatus according to claim 6, characterized in that for at least one parameter, the measured depth values in a DC component ( 42 ) and a variable component ( 44 ) are splittable and from the variable component ( 44 ) of the measured depth values the conformity measures can be determined ( 32 ). Device according to one of claims 6 to 7, characterized in that for at least one parameter, the correspondence measures can be determined by means of a linear regression ( 32 ). Device according to one of claims 6 to 8, characterized in that the conformity measures of all parameters and / or the overall agreement measures can be entered in a display grid, in particular for visual representation of the conformity measures and / or the overall agreement dimensions. Device according to one of claims 6 to 9, characterized in that a measure of a reliability of the determined position or the possible positions of the watercraft can be specified.

Images