FIELD OF THE INVENTION 0001 The invention relates to a method for correcting aiming errors between a sensor device and an effector device controlled by the sensor device via a servo device by means of a correction of an aiming error vector (B). The invention also relates to a device for executing this method.

BACKGROUND OF THE INVENTION 0002 A method for correcting aiming errors between gun carriages and devices arranged thereon is known from EP 0 314 721 B1, wherein the devices can be fire control systems and weapons installations. The method is executed by using device correction values of the rough position of the installed devices, measured with the fire control systems and weapons installations at rest, and by taking them into account in the servo controls of the gun carriages. The correction values for the devices are known at the factory and/or are determined from measured values.

OBJECT AND SUMMARY OF THE INVENTION 0003 It is the object of the present invention to improve such a method and to propose a device for executing it. 0004 This object is attained by the invention in an advantageous manner by means of a method in accordance with claim 1 and a device in accordance with claim 10. 0005 By means of this it is possible to take system deviations from a defined ideal geometry into account in order to increase accuracy during firing when calculating the control values for the gun carriage servos. 0006 Other advantageous embodiments of the invention ensue for the further dependent claims. 0007 The invention will be explained in greater detail by way of example in what follows by means of the drawings.

BRIEF DESCRIPTION OF THE DRAWINGS 0008 FIG. 1 is a schematic representation of the mutual linkage of the positions of the sensor devices and effector arrangements, 0009 FIG. 2 shows an individual observation in the course of the precision measurement in accordance with the invention, 0010 FIG. 3 shows the result of an individual observation in accordance with FIG. 2, 0011 FIG. 4 is a representation explaining the coordinate system used, 0012 FIG. 5 shows the result of the entire set of observed values, and 0013 FIG. 6 shows the result of the values corrected in accordance with the invention.

DETAILED DESCRIPTION 0014 FIG. 1 shows an arrangement with a total of five devices, namely two sensor devices in the form of fire control devices T1 and T2 and three computer-controlled effector arrangements in the form of guns G1, G2, G3. The sensor devices and the effector arrangements could be located aboard ship or on land. All these arrangements T1, T2, G1, G2, G3 are placed in gun carriages or emplacements and are at least mechanically roughly aligned. 0015 For example, a helicopter 10 and a single mounting arrangement with a sensor device T and an effector arrangement G are represented in FIG. 2. The sensor device T can be, for example, a fire control or aiming device, also identified by T, for controlling the gun G. The gun G can be provided with a TV sensor Sg, for example. The fire control device T controls the gun G via data or signal lines 11. The gun G, as well as the aiming device T, are aimed at a common measurement target K, for example a sphere also identified by K, which has been attached to a support cable 12 of the helicopter 10. 0016 By means of this arrangement it is intended to define a correction of the aiming error vector B, or of several aiming error vectors Bjk, in FIG. 1, for example B11, B12, B21, B22, B31, B32. It is assumed here that the aiming error vector B, or the aiming error vectors Bjk, are base vectors, which are known from rough position measurements, measurements at the factory, etc. and have been stored. 0017 A precision measurement is performed by means of the method in accordance with the invention in order to improve these known values of the aiming error vectors B, or Bjk, in several steps, or following several measurements. Therefore, after a number i of steps, the following applies to an aiming vector value B, which was corrected by a calculated correction vector Pn, wherein i represents whole number values from 1 to n:

B(new)B(old)Pn.

0018 After a number n of measurements it can be assumed that Pn Ps, wherein Ps corresponds to the real or correct, but unattainable per se, value for the correction of the system as such. 0019 If, for example, the sensor devices, or effector arrangements are located on a ship, in case of a change of the weight of the ship on account of its cargo, the fuel present, or a change in the shape of the ship's hull, etc., a new value for the correction value Ps results, which can again be approximately determined in the form of a new Pn value by measurements performed with the aid of the sphere K fastened to the helicopter 10. Very small changes in the shape of the hull of the ship, for example by bending or torsion, particularly following an explosion, cause a relatively large change in the reference angles. It is one aim of the invention to take these very small changes into consideration. 0020 The display shown in FIG. 3 represents how the TV sensor Sg sees a sphere K, for example the measurement target K, or the sphere K, namely in the actual position generally with a certain amount of deviation from an intersection point O of the crosshairs of the display. This deviation, which can be directly observed by the TV sensor Sg, is a positional error, which is the result of all system errors of whatever type, such as mechanical inaccuracies as the result of manufacturing tolerances or wear, residual errors in the rough position measurements, changes in the shape of the ship's hull, measuring noise. The deviation can be considered to be an aperture vector Di with two components which, when transposed, can be represented as follows:

i DidyidziT,

0021 wherein dyi and dzi are the components in the axes y, or z of the aperture vector Di. The value d of the length of the aperture vector Di can be calculated in accordance with FIG. 3 as

d(dyi2dzi2){fraction (1/2)}.

0022 The display in accordance with FIG. 3 is calibrated in accordance with a predetermined distance so that the components dyi and dzi, which actually are angles, can be represented by lengths, or distances. The following equation applies to the aperture vector Di:

DiMi*PsRiDicRi, wherein Riresidual error.

0023 Factors which affect the residual error Ri are, besides the thermal noise, inter alia the motion of the sea, inaccuracies of the servo system, and the fact that the operator cannot place a mark represented in FIG. 3 exactly on the measuring target in its instantaneous position Ki. 0024 A coordinate system in accordance with FIG. 4 is defined in the area of the aiming device T and the gun G. If, for example, the aiming device T and the gun K are located on the ground, the X-axis is oriented to the north, the Y-axis to the east and the Z-axis to the center of the earth, for example. If the aiming device T and the gun K are located on a ship, the X-axis, for example, is the longitudinal axis of the ship, the Y-axis the transverse axis, and Z-axis a right-hand axis which is orthogonal in respect to the X-axis and the Y-axis. In the coordinate system which is defined by the X-, Y-, and Z-axes, every position which the measuring target Ki acan assume, is determined by three coordinates xK, yK and zK. However, in ballistics the angle values K and K are used as coordinates for practical reasons, wherein the azimuth angle is identified by k, and the elevation angle by k. Therefore the values K and K are redundant. The coordinates xK, yK, zK and K are considered to be the components of a target vector OKi, wherein the azimuth angle or the elevation angle can also be calculated from these coordinates. The projection of the vector OK on the plane X-Y in FIG. 4 defines a straight line g, and a straight line also located in the plane X-Y and intersecting the straight line g perpendicularly in the zero point O is selected as the X-axis. 0025 The previously mentioned recursively calculated vector Pi preferably has four components, as follows:

Pixiyizii

0026 wherein xi, yi, zi and i are small angle values and wherein 0027 xi is a rotation around the X-axis, 0028 yi a rotation around the Y-axis, 0029 zi a rotation around the Z-axis, and 0030 i a rotation around the -axis. 0031 These rotations or tiltings result because the plane of rotation of the effector arrangement, i.e. the gun G, is not parallel with the plane of rotation of the sensor device, i.e. the aiming device T. 0032 The error resulting therefrom has two degrees of freedom, and therefore can be corrected by the two rotations xi around the X-axis and yi around the Y-axis. However, the rotation zi around the Z-axis also includes the rotation of the azimuth . Therefore a transformation matrix Mi exists for each position of a target defined by a target vector OKi, or for each process step i, which is defined as follows: 1

M_{i} = \begin{matrix} - \cos_{i} \sin_{i} & - \sin_{i} & \sin_{i} - \cos_{i} & 0 \\ - \sin_{i} & \cos_{i} & 0 & 1 \end{matrix}

0033 wherein i1, 2, 3, . . . n. 0034 A co-variance matrix Si also exists for each process step i, as follows: 2

S_{i} = S_{i - 1} - \frac{S_{i - 1} * M_{i}^{T} * M_{i} * S_{i - 1}}{(M_{i} * S_{i - 1} * M_{i}^{T} + I)}

0035 wherein I is an uniform matrix. 0036 Finally, an error vector E (equation error) is defined by the following equation:

EiDiMi*Pi1.

0037 The calculation is initialized with the following values: 3

P_{o} = {\begin{matrix} 0 & 0 & 0 & 0 \end{matrix}}^{T}

and

S_{o} = \begin{matrix} 1 & 0 & 0 & 0 \\ 0 & 1 & 0 & 0 \\ 0 & 0 & 1 & 0 \\ 0 & 0 & 0 & 1 \end{matrix} * C

0038 wherein C is a constant. 0039 The recurrence starts with initial values Po and So and with calculated values for Mi and measured values of Didyi dziT wherein i starts with 1. From this, the values of Ei and Si are determined in accordance with the above noted recurrence equations, as well as subsequently Pi in accordance with the following recurrence equation:

PiPi1Si*MiT*Ei wherein i1, 2, 3 . . . n.

0040 This recursive algorithm minimizes the following quality index J(p) (performance):

J(p)sum (i1, 2, . . . , n) (DiMi*Pi)T*(DiMi*Pi)

0041 The algorithm in accordance with the present invention is based on a special application of the method of the least error squares, wherein the most advantageous values are obtained in that the sum of squares of the respective difference between the observed value for Di and the calculated value for DicMi*Pn results in a minimum. 0042 The calculated correction vector Pi is transformed into the vector Di, or the components xi, yi, zi and i into the components dyi, dzi, by means of the transformation matrix Mi. A matrix S is used in order to avoid ambiguities in the plane of observation (FIG. 3). The matrix S is the above indicated co-variance matrix which, in particular for orthogonal-symmetrically laid out measurements, leads to a diagonal-symmetrical matrix with vanishing values in the diagonal, i.e. the track Sp or the convergence number tends toward 0. Tests in regard to the point position of this convergence number have shown that it is advantageous to select the value 49.25 or 492.5, etc. for the constant C. At C49.25, the value of the track of the co-variance matrix Sn decreases from 99.99 . . . at the start to approximately 0.03 with a sufficiently large number n of measurements, or steps. However, the constant C can also be 1, or have any arbitrary value. After a number n of measurements, or recurrence steps, for example 25<n<400, preferably n . 200, the value of Pn tends toward the searched-for value Ps. 0043 FIG. 5 shows by means of a cross a number of actual positions of the measurement target K carried by the helicopter 10. The correspondingly corrected values of these positions are represented in FIG. 6. If the X-Y-Z coordinate system is based on a ship, the helicopter 10 preferably flies on a circular track with a radius of an order of magnitude of 1.5 km, but helically, or with increasing elevation Ti, Ti, Ti around the ship. On the basis of the data determined by the aiming device T, and taking into consideration parameters so far known, in particular parallaxes between the aiming device T and the gun G, the sensor aiming line O of the gun G (instead of the line of fire, or the weapon tube axis, offset at a small parallax, of the gun G) is aimed as best as possible at a target Ki, preferably automatically. The intersection point of the crosshairs of the aiming line of the sensor (Sg) (FIG. 3) points in the direction in which the measurement target Ki is expected. 0044 Therefore every point in FIG. 5 identified by a cross respectively relates to a measured value of Ki or Ki, i.e. to the azimuth angle or the elevation angle of the gun G, which correspond to the respective position Ki of the target K carried by the helicopter 10. In contrast thereto, FIG. 6 corresponds to the theoretical values of or , which, following the corrections in accordance with the present method, would be measured under exactly the same conditions, if such a further measurement would be practically possible at all. In actuality, it is impossible to repeat the measurements with the helicopter 10 in exactly the same positions as during previous measurements, and under the same ship conditions, etc. 0045 Theoretically, due to the correction made, all points should coincide with the zero point in FIG. 6. However, because of the residual errors Ri in the system, as represented in FIG. 6, which are unavoidable, the points do not coincide with the zero point 0, i.e. statistically diverging deviations from the zero point 0 result, whose distribution, however, is free of average values, i.e. the mean value of the divergences of the points is zero on both axes. 0046 In comparison with other algorithms operating with different passages for computers of similar systems, the algorithm in accordance with the present invention has been shown to be particularly advantageous in view of the fact that initialization in accordance with the invention is completely problem-free, and that singularities (determinant0) never occur, so that no derailing of the system needs to be feared. Such derailments could occur, for example, if an attempt is made in each passage to match measured values to a predetermined curve, such as a sine curve. 0047 As in the system in accordance with patent document EP 0 314 721 B1, the correction data based on measurements, with which the aiming error vectors are corrected, have an effect which corrects the wrong aiming in real time. The measurements can again be performed from time to time, for example after four or six weeks, in order to match the correction data to changing conditions, for example those of a ship. This means that the measurement values gained from time to time can be integrated into the system and are therefore inherent in the system and therefore respectively correspond to an error which cannot be directly observed. 0048 The sensor device T (tracker) can be a sensor, an aiming device, a radar, laser or infrared device, etc., or several such devices can be combined. Not only conventional guns, such as cannons, for example, are used as effector devices G (guns), but also rocket-firing devices or laser guns. The measurements can be performed for different G/T pairs B11, B12, B21, B22, . . . (see FIG. 1), wherein one sensor device T can also control several effector devices G. 0049 The installation described by means of the drawings can have the required controls, computer means, or hardware, and programs, or software, in order to make possible the various methods, or partial methods, in accordance with the claimed variants, or in any arbitrary combination thereof.