Method and device for controlling a simulator

Abstract

The present invention relates to a filter for controlling a simulator for the representation of movements of a simulated vehicle, the rotational and translational control commands required to control the simulator being calculated from a difference between the angle of rotation and the apparent vertical angle, taking into account a physiological rotation rate limitation.

Claims

1. A method for actuating a simulator for simulating translational and rotational movements of a vehicle, wherein, in relation to a three vehicle axes, a rotational rate about a first vehicle axis and specific forces respectively acting along a second vehicle axis and a third vehicle axis are provided from a movement model that simulates the vehicle and converted into translational and rotational control commands for actuating the simulator, comprising the following steps: calculating, using a control unit, a rotational angle from the rotational rate about the first vehicle axis, calculating, using the control unit, from the specific forces, an apparent perpendicular angle between a vertical axis as the third vehicle axis and the apparent perpendicular arising on account of the specific forces acting along the second vehicle axis and third vehicle axis, calculating, using the control unit, an apparent perpendicular angle difference between a rotational angle and the apparent perpendicular angle and ascertaining a high-frequency difference component of the apparent perpendicular angle difference that is intended to be compensated by a translational movement of the simulator, depending on the apparent perpendicular angle difference, calculating, using the control unit, translational control commands for actuating the simulator for a translational movement of the simulator along the second vehicle axis, depending on the ascertained high-frequency rotational angle component of the apparent perpendicular angle difference, calculating, using the control unit, a compensation angle that corresponds to an acceleration value to be simulated along the second vehicle axis by inclining the simulator in relation to perpendicular to the Earth, depending on the rotational angle, the high-frequency difference component of the apparent perpendicular angle difference and the apparent perpendicular angle, calculating, using the control unit, a limited compensation angle from the compensation angle by means of a physiological rotational rate limitation, which restricts an inclination of the simulator below a perception threshold, and calculating, using the control unit, rotation control commands for actuating the simulator for a rotational movement of the simulator about the first vehicle axis, depending on the limited compensation angle and the rotational angle.

2. The method as claimed in claim 1, wherein the high-frequency difference component of the apparent perpendicular angle difference is ascertained by means of a high-pass filter.

3. The method as claimed in claim 1 wherein the translational control commands for translational movement of the simulator along the second vehicle axis is calculated from a product of the gravitational acceleration and a sine of the ascertained high-frequency difference component of the apparent perpendicular angle difference.

4. The method as claimed in claim 1 wherein the compensation angle is calculated from a sum of the rotational angle and the high-frequency difference component of the apparent perpendicular angle difference minus the apparent perpendicular angle.

5. The method as claimed in claim 1 wherein the rotational control commands for a rotational movement of the simulator about the first vehicle axis are calculated from a sum of the limited compensation angle and of the rotational angle.

6. The method as claimed in claim 1 wherein the limited compensation angle is calculated by means of a physiological rotational rate limitation, depending on the compensation angle and the rotational rate that is provided from the movement model, wherein a restriction value for limiting the compensation angle is ascertained depending on a predetermined minimum restriction value and a high-frequency rotational rate component of the rotational rate.

7. A simulator for simulating translational and rotational movements of a vehicle, which comprises: a platform that is movable in relation to a stationary reference plane by means of actuators; and a control unit that is configured to carry out the method as claimed in claim 1 for actuating the simulator.

8. The simulator as claimed in claim 7, wherein the simulator is a hexapod.

9. A computer program comprising program code means, configured to carry out the method as claimed in claim 1 when the computer program is executed on a computer.

Description

(1) The invention is explained in an exemplary manner on the basis of the attached figures. In detail:

(2) FIG. 1 shows a basic illustration of the classic washout filter algorithm (prior art);

(3) FIG. 2 shows a basic illustration of the apparent perpendicular filter according to the invention for a lateral movement;

(4) FIG. 3 shows the basic structure of the physiological rotational rate limitation for a lateral movement;

(5) FIG. 4 shows a schematic illustration of the apparent perpendicular during level flight with constant roll angle (suspended area);

(6) FIGS. 5a to 5d show illustrations of various signal curves for the flight state of the hanging area;

(7) FIG. 6 shows a schematic illustration of the apparent perpendicular during a curve on the ground;

(8) FIGS. 7a to 7d show signal curves of the flight state of the curve on the ground;

(9) FIG. 8 shows a schematic illustration of the apparent perpendicular during a coordinated curve during flight;

(10) FIGS. 9a to 9d show schematic illustrations of the signal curves of the flight state of the coordinated curve.

(11) FIG. 1 shows the basic structure of a classic washout algorithm, as is known from the prior art. Here, high-frequency translational accelerations are directly converted into a translational movement of the simulator, while low-frequency translational acceleration values are mapped by way of an inclination of the simulator cabin (tilt coordination channel).

(12) FIG. 2 shows the basic illustration of the apparent perpendicular filter according to the invention for a lateral movement. In the general form, the input variables for the apparent perpendicular filter are the specific forces

(13) $\begin{array}{cc}{\stackrel{.fwdarw.}{f}}_{\mathrm{aa}}={\stackrel{.fwdarw.}{a}}_{\mathrm{aa}}-{\stackrel{.fwdarw.}{g}}_{\mathrm{PA}}=\left(\begin{array}{c}{f}_{\mathrm{aa},x}\\ f_{\mathrm{aa},y}\\ f_{\mathrm{aa},z}\end{array}\right)& \left(1\right)\end{array}$
and the rotational accelerations

(14) $\begin{array}{cc}\stackrel{\stackrel{.fwdarw.}{.}}{}=\left(\begin{array}{c}{\stackrel{.}{}}_{\mathrm{aa},x}\\ {\stackrel{.}{}}_{\mathrm{aa},y}\\ {\stackrel{.}{}}_{\mathrm{aa},z}\end{array}\right)& \left(2\right)\end{array}$
or, alternatively, the rotational speeds

(15) $\begin{array}{cc}\stackrel{.fwdarw.}{}=\left(\begin{array}{c}{}_{\mathrm{aa},x}\\ {}_{\mathrm{aa},y}\\ {}_{\mathrm{aa},z}\end{array}\right)& \left(3\right)\end{array}$
of the simulated aircraft.

(16) In the aircraft, the apparent perpendicular angle is composed of the component as a consequence of additional forces and the component as a consequence of an angular position

(17) $\begin{array}{cc}{\stackrel{.fwdarw.}{}}_{\mathrm{PA}}={\left(\begin{array}{c}{}_x\\ {}_y\end{array}\right)}_{\mathrm{PA}}={\stackrel{.fwdarw.}{}}_{t,\mathrm{PA}}+{\stackrel{.fwdarw.}{}}_{,\mathrm{PA}}.& \left(4\right)\end{array}$

(18) In the individual rotational axes, the apparent perpendicular angles emerge via

(19) $\begin{array}{cc}{\left(\begin{array}{c}\tan \left({}_x\right)\\ \tan \left({}_y\right)\end{array}\right)}_{\mathrm{PA}}={\left(\begin{array}{c}\frac{f_{\mathrm{aa},x}}{f_{\mathrm{aa},z}}\\ \frac{f_{\mathrm{aa},y}}{f_{\mathrm{aa},z}}\end{array}\right)}_{\mathrm{PA}}& \left(5\right)\end{array}$
to form

(20) $\begin{array}{cc}{\stackrel{.fwdarw.}{}}_{\mathrm{PA}}={\left(\begin{array}{c}{}_x\\ {}_y\end{array}\right)}_{\mathrm{PA}}={\left(\begin{array}{c}\arctan \left(\frac{f_{\mathrm{aa},x}}{f_{\mathrm{aa},z}}\right)\\ \arctan \left(\frac{f_{\mathrm{aa},y}}{f_{\mathrm{aa},z}}\right)\end{array}\right)}_{,\mathrm{PA}}.& \left(6\right)\end{array}$

(21) The first term from formula 4 emerges from the translational acceleration of the aircraft. These values are not explicitly available when transferring the specific forces. However, by rewriting formula 4 as
{right arrow over ()}.sub.t,PA={right arrow over ()}.sub.PA{right arrow over ()}.sub.,PA,(7)
they are establishable and can be determined with the aid of the overall angle
{right arrow over ()}.sub.PA={right arrow over ()}.sub.aa.(8)

(22) The second component emerges from the angular position of the aircraft and the gravitational acceleration:

(23) $\begin{array}{cc}{\stackrel{.fwdarw.}{f}}_{,\mathrm{PA}}={\left(\begin{array}{c}{f}_x\\ f_y\\ f_z\end{array}\right)}_{,\mathrm{PA}}=\left(\begin{array}{c}\sin \left({}_{\mathrm{PA}}\right)\\ -\sin \left({}_{\mathrm{PA}}\right).Math.\cos \left({}_{\mathrm{PA}}\right)\\ -\cos \left({}_{\mathrm{PA}}\right).Math.\cos \left({}_{\mathrm{PA}}\right)\end{array}\right).Math.g,& \left(9\right)\end{array}$
from which the apparent perpendicular angles

(24) $\begin{array}{cc}{\stackrel{.fwdarw.}{}}_{,\mathrm{PA}}={\left(\begin{array}{c}\\ \end{array}\right)}_{,\mathrm{PA}}={\left(\begin{array}{c}\arctan \left(\frac{f_x}{f_z}\right)\\ \arctan \left(\frac{f_y}{f_z}\right)\end{array}\right)}_{,\mathrm{PA}}& \left(10\right)\end{array}$
are establishable. Hence, the two input variables are completely available.

(25) Equivalently, the aforementioned equations could also be established for the simulator cabin in order to determine the difference from the apparent perpendicular of the aircraft and the forces that act on the pilot.

(26) $\begin{array}{cc}{\stackrel{.fwdarw.}{}}_{\mathrm{PS}}={\left(\begin{array}{c}{}_x\\ {}_y\end{array}\right)}_{\mathrm{PS}}={\stackrel{.fwdarw.}{}}_{t,\mathrm{PS}}+{\stackrel{.fwdarw.}{}}_{,\mathrm{PS}}& \left(11\right)\end{array}$

(27) Both terms are available for the simulators. From this, the specific forces in the follow-up rotation of the simulator cabin emerge as

(28) 0$\begin{array}{cc}{\stackrel{.fwdarw.}{f}}_{,\mathrm{PS}}={\left(\begin{array}{c}{f}_x\\ f_y\\ f_z\end{array}\right)}_{,\mathrm{PS}}=\left(\begin{array}{c}\sin \left({}_{\mathrm{PS}}\right)\\ -\sin \left({}_{\mathrm{PS}}\right).Math.\cos \left({}_{\mathrm{PS}}\right)\\ -\cos \left({}_{\mathrm{PS}}\right).Math.\cos \left({}_{\mathrm{PS}}\right)\end{array}\right).Math.g,& \left(12\right)\end{array}$
from which the apparent perpendicular angles in the follow-up rotation of the simulator cabin

(29) $\begin{array}{cc}{\stackrel{.fwdarw.}{}}_{,\mathrm{PS}}={\left(\begin{array}{c}\\ \end{array}\right)}_{,\mathrm{PS}}={\left(\begin{array}{c}\arctan \left(\frac{f_{\mathrm{PS},x}}{f_{\mathrm{PS},z}}\right)\\ \arctan \left(\frac{f_{\mathrm{PS},y}}{f_{\mathrm{PS},z}}\right)\end{array}\right)}_{}& \left(13\right)\end{array}$
can be determined. The following applies to the translational acceleration in the coordinate system that is fixed relative to the cabin:

(30) $\begin{array}{cc}{\stackrel{.fwdarw.}{f}}_{t,\mathrm{PS}}={\left(\begin{array}{c}{f}_x\\ f_y\\ f_z\end{array}\right)}_{t,\mathrm{PS}}={\left(\begin{array}{c}\stackrel{\mathrm{.Math.}}{x}\\ \stackrel{\mathrm{.Math.}}{y}\\ \stackrel{\mathrm{.Math.}}{z}\end{array}\right)}_f.& \left(14\right)\end{array}$

(31) From this, the following follows for the apparent perpendicular angles in the simulator cabin:

(32) $\begin{array}{cc}{\stackrel{.fwdarw.}{}}_{t,\mathrm{PS}}={\left(\begin{array}{c}\\ \end{array}\right)}_{t,\mathrm{PS}}={\left(\begin{array}{c}\arctan \left(\frac{f_x}{f_z}\right)\\ \arctan \left(\frac{f_y}{f_z}\right)\end{array}\right)}_{t,\mathrm{PS}}.& \left(15\right)\end{array}$

(33) Using the relationships shown above, the input variables are initially converted either into an angle or into a specific force. For a rotational speed at the input, this is brought about by a simple integration, and by a double integration in the case of a rotational acceleration. Now, the obtained angle can be converted into a specific force by means of formula 10, provided that the apparent perpendicular filter should be set up as an equivalent with the aid of the forces.

(34) Here, FIG. 2 shows a basic structure of the present invention for the lateral movement. The longitudinal movement can be implemented in a manner analogous thereto without problems. The relationships between rotational position and translational acceleration, which are yet to be discussed below, apply both to the lateral movement and to the longitudinal movement in this case, and so the relationship between roll position and lateral acceleration is analogous to the relationship between pitch angle and longitudinal acceleration.

(35) In principle, three relevant flight states for the lateral movement which are relevant to the actuation of the simulator can be identified:

(36) Case 1: Level flight with the constant roll position (suspended area)

(37) Case 2: Driving a curve on the ground

(38) Case 3: Coordinated curve in flight with simultaneous roll position of the aircraft

(39) In case 1, an apparent perpendicular angle in the aircraft is established due to transverse force on the basis of the roll position in level flight. In case 2, an apparent perpendicular angle is likewise present, said apparent perpendicular angle being based on the centrifugal force when driving through the curve on the ground. In the case 3, there is, in principle, no apparent perpendicular angle since the transverse forces cancel on account of the coordinated curve with the simultaneous roll position of the aircraft.

(40) In principle, two degrees of freedom, the roll angle and the lateral acceleration, are usable for representing apparent perpendicular angles. Both are subject to restrictions. Thus, an acceleration in the lateral direction quickly leads to the arrival at the boundary of the movement space, whereas, by contrast, a representation by way of the roll angle at low frequencies is possible, but perception thresholds may possibly need to be observed, the latter preventing a quick representation of a transverse force.

(41) Since, both the aircraft and the simulator cabin simply have to be rotated about the longitudinal axis for case 1, this case is, in principle, exactly representable by the simulator. Thus, the first component for the roll angle of the simulator cabin is the roll angle of the simulated aircraft.

(42) By contrast, no transverse forces and hence no apparent perpendicular angles arise in this case for the roll position of the aircraft for case 3. If the simulator cabin in this case also follows the roll angle of the aircraft, the difference between the two transverse forces would lead to an apparent perpendicular angle in the simulator cabin which does not exist in the aircraft. Accordingly, this difference must be subtracted again from the first component of the roll angle. If this were carried out without further measures, the simulator cabin would remain stationary.

(43) Hence, the apparent perpendicular angle would be represented correctly. However, the perception of the rotational speed during the entry into the curve is lacking in this case. A roll movement is initiated by way of a suitable manipulation of the apparent perpendicular angle difference in order to excite said perception, for example by way of a high-pass filter. However, in the case of the lateral movement, this roll movement leads to a lateral acceleration impression, the sign of which is in the wrong direction. This can be compensated by an opposing lateral acceleration of the simulator cabin while observing the boundaries of the movement space. Since the latter is quickly reached, the roll speed must be terminated again quickly. As a result of the corresponding movement in the lateral direction and in the roll direction, this can be carried out without the incorrect perception of movement. At the beginning, the rotational rates are deliberately above the perception threshold. By contrast, the return rotation must occur below the said perception threshold. As long as this is successful, the apparent perpendicular angle in the simulator cabin corresponds to that of the aircraft. Only the rotational rates differ and the best possible correction of the roll angle has been found.

(44) Initially, no solution is possible for case 2 in the case of a sudden increase in the lateral force as a consequence of an occurring centrifugal force. Since the aircraft on the ground only carries out very small roll movements about the longitudinal axis, a sudden rotation of the simulator cabin must lead to a significant deviation between the acceleration impression in the aircraft and in the simulator. In the present invention, this deviation is determined by the difference of the already set roll angle and the apparent perpendicular angle in the aircraft on account of a lateral force. This difference must correspond to the component of the centrifugal acceleration on the apparent perpendicular angle. This difference is implemented by a limitation function.

(45) On account of the system design, there is no ideal solution in the simulator to the problem of representing quickly occurring and long-term transverse forces due to a centrifugal force. At best, an adaptation of the filter properties to the maneuvers to be carried out and, ideally, the avoidance of perceptions with the wrong sign are possible. In the present invention, the optimization can be carried out in the limitation function. This is expedient since the necessary rotational rate limitation represents a cause of the error. Here, there are far-reaching options for adapting the system reaction, for example by way of increasing the rotational rate limitation after fast maneuvers or in the case of high workloads on the pilots. Consequently, a good compromise between observing the perception threshold and a fast representation of the acceleration can be found for every maneuver. Since the two other cases are already represented correctly as far as this is possible, there is no feedback of the settings found here on other cases. Hence, the last component for the roll angle is also found.

(46) The functionality of the present invention is explained on the basis of an exemplary embodiment of FIG. 2 for the transverse movement. The basic structure can also be transferred to the longitudinal movement without restrictions. All that changes are the relationships between angular positions and accelerations.

(47) For the following exemplary embodiment in the transverse direction, the specific forces along the vertical axis (z-direction/E3) and along the transverse axis (y-direction/E2), and the roll rate (E1) of the simulated aircraft are provided. The component of the apparent perpendicular angle in the plane spanned by the vertical axis and transverse axis as a consequence of translational accelerations is then calculated as:

(48) $\begin{array}{cc}{}_f=\arctan \left(\frac{f_{\mathrm{aa},y}}{f_{\mathrm{aa},z}}\right),& \left(16\right)\end{array}$
wherein the apparent perpendicular angle represents a deviation between the vertical axis of the vehicle in the coordinate system that is stationary in relation to the vehicle and the perceived acceleration vector. In FIG. 2, the calculation is represented by M1/T1.

(49) Furthermore, the incoming roll rate at the pilot seat is integrated, represented by E1, and so the roll angle of the simulated aircraft can be determined by means of:
.sub.=.sub.aa,xdt.(17)

(50) After all input signals were converted into angles, the actual filter is set up. The angular position .sub.sim (A1) and the translational acceleration of the simulator cabin y.sub.pp,sim (A2) are obtained as output signals.

(51) The angular position of the simulator in this case corresponds to the sum (S4) of the positional angle of the aircraft .sub. and a limited compensation angle .sub.A,lim. The former corresponds to the converted rotational speed input signal .sub.aa,x. The latter corresponds to the limited compensation angle with which the simulator cabin should be tilted for representing long-term translational accelerations in relation to the perpendicular to the Earth.

(52) Without further filter components, this signal can map the first case, i.e. the state of a roll position with suspended areas, in a complete and error-free manner.

(53) By contrast, for case 3, i.e. for representing the acceleration impressions in a coordinated curve, the rotational angle that builds up in the simulator cabin .sub. as result of the roll movement must be compared to the apparent perpendicular angle of the aircraft .sub.f (S1). In this case, the two angles will deviate from one another since the apparent weight force points directly in the direction of the z-axis (vertical axis) in the aircraft on account of centrifugal forces. This force is missing in the simulator. The difference of both (S1) could now simply be added to the roll position and the apparent perpendicular is correct at all times. However, the perceivable roll movement in the simulator is also missing in this case since the cabin would simply remain still.

(54) On the other hand, an increasing roll angle in the cabin leads to a transverse force with an inverted sign. This can easily be seen from formula 10. Therefore, the only possibility is that of returning the roll angle again that arises after a brief roll movement with the correct sign and thus of compensating, to the greatest possible extent, the error arising as a result thereof on account of the increasing roll angle by way of an appropriate translational movement or acceleration. This is carried out by way of a high-pass filter (HP1), by means of which the apparent perpendicular angle difference (S1) is manipulated. The result is a high-frequency difference component .sub.ypp of the apparent perpendicular angle difference in the direction of the roll speed of the aircraft .sub.aa,x. By way of example, these can be input into a third order high-pass filter, which returns the roll position of the simulator cabin back into the zero position.

(55) An apparent perpendicular with an incorrect sign also arises in the simulator cabin during this maneuver. The cabin is accelerated in the opposite direction in order to compensate this. Here, the movement space in the translational direction is the limiting factor. Thus, the aforementioned high-pass filter must be designed in such a way that these boundaries are not infringed. In addition to the high-pass filter, the effects of a signal may likewise be manipulated by way of a proportionality factor. Moreover, it is conceivable to use further options in the translational movement space to the best possible extent.

(56) Now, a back conversion into translational accelerations by means of
y.sub.pp,sim=g.Math.sin(.sub.ypp)(18)

(57) (T2/F1) is still required for the output signal y.sub.pp,sim (A2). As a result of the corresponding signals in the roll direction and y direction, the apparent perpendicular in the simulator cabin continues to be represented correctly if the boundary of the movement space is observed. Hence, in this case, the sum in S2 of the roll angle of the simulator cabin as a consequence of a roll movement of the aircraft (.sub.) and the roll angle of the simulator cabin as compensated by a translational acceleration along the transverse axis .sub.ypp corresponds to the apparent perpendicular angle of the aircraft, as ascertained by formula 16.

(58) For the second case, a curve on the ground, the rotation of the apparent perpendicular does not correspond to a roll movement of the aircraft. As a consequence, an apparent perpendicular angle is present directly after entering the curve. The simulator cabin could follow said apparent perpendicular angle but the rotational movement would be detected by the pilot. Since the acceleration perception would then be contrary to the correct positional information items, dizziness or far-reaching consequences would be expected.

(59) Since the apparent perpendicular in the aircraft arises without a rotation about the roll axis, a difference is obtained at the first comparison point (S1) in a manner analogous to case 3. This once again leads to a non-high-frequency difference component .sub.ypp at the output of the high-pass filter HP1, followed by a compensation function in the translational direction .sub.ypp,sim.

(60) However, this time, the sum of the roll angle .sub. and the high-frequency difference component .sub.ypp at the output of S2 does not correspond to the apparent perpendicular angle in the aircraft .sub.f. The difference between the two variables is formed in the difference unit S3 and the former corresponds to the compensation angle .sub.a. Now, a correction is possible if the perception threshold is observed. Unlike in the first case and the third case, there cannot be a conversion of the signals with the correct apparent perpendicular apart from movements that can be converted below the perception threshold. All that can be attempted is to obtain the best possible result by way of a skillful limitation (L), the result of which is the limited compensation angle .sub.a,lim.

(61) FIG. 3 shows, in an exemplary manner, an embodiment of the rotational rate limitation. The limitation has an obligatory input E4, by means of which the compensation angle .sub.a is provided, and the input E5 for transferring the roll speed .sub.aa,x. The latter only serves to adapt the limit values and can, as desired, be replaced with other available information items for optimizing the rotational rate limitation or else be omitted.

(62) Moreover, the two parameters P1 for the smallest limit value L.sub.p,norm and P2 for the current system time T are included in the function.

(63) The rotational rate limitation is increased for a certain time after the event in the case of quickly changing rotational speeds by way of the high-pass filter HP2 and the factor F2. By calculating the magnitude at T5, the absolute limit value L.sub.p,alt is obtained. On the other hand, P1 sets a minimum value L.sub.p,norm for the perception threshold. In principle, this parameter, too, is changeable depending on the situation. By way of the comparison at S6 and the selection T8, the respectively larger value is selected as an absolute value for the maximum admissible rotational rate L.sub.p. The length of the time step t is obtained by forming the difference between the current system time T and the preceding system time T at T6. Multiplying said time step by the admissible rotational rate (M2) leads to the upper limit of the current angular change L.sub.UP and, after multiplication by the factor 1 (F3), to the lower limit of the current angular change L.sub.DN.

(64) The accompanying compensation angle .sub.a is compared at T3 with, firstly, the value from the preceding step such that the result is angle change .sub.a calculated for the current time step. Secondly, .sub.a is compared to the entire angle difference between the current value and the overall value .sub.A,lim,old calculated for the preceding time step (S5). The result is the angle .sub.A,old that remains from preceding time steps. The latter is added to the current angle step (S7), and so the still to be traveled, unlimited angle difference .sub.uni is available. At T4, the latter is restricted to the admissible value range using the limit values L.sub.UP and L.sub.DN in .sub.lim. This limited angle difference is added to the overall angle .sub.A,lim,old in S8. The result is the limited compensation angle .sub.A,lim. The latter is, in turn, stored in T7 for the next time step.

(65) Case 1: Level Flight with Constant Roll Angle (Suspended Area)

(66) FIG. 4 schematically shows the apparent perpendicular and apparent weight vector in level flight with constant roll angle from various perspectives. Shown on the left is the aircraft as it would look to an external observer in this flight state if it were standing on the ground. Here, the apparent perpendicular corresponds to the perpendicular to the Earth. However, a different picture emerges for the pilot from the view of the aircraft cabin. As a result of the constant roll angle, the effect on the pilot in his aircraft cabin reference system is as if a transverse force were to act on the pilot in the direction of the roll angle by way of said constant roll angle.

(67) In order to map this by the simulator, it is sufficient if the simulator cabin is rotated in accordance with the rotational angle, with there being an appropriate adaptation of the interior display. Since the roll movement of the aircraft can be traced by the simulator without losses, an identical movement is possible here, without an incorrect movement impression arising.

(68) FIG. 5a shows the corresponding input signals for this example. Here, the upper image shows the rotational rate until the final angle is reached, followed by the specific forces along the transverse axis and the specific forces along the vertical axis.

(69) The output signal for actuating the simulator in accordance with the present invention is shown in FIG. 5b. Here, FIG. 5c shows a conversion of the output signal for the translational movement into position (upper presentation), speed (central presentation), and acceleration (lower presentation). It is possible to see that no translational acceleration or movement of the simulator for simulating the flight state is necessary in this case.

(70) Finally, FIG. 5d shows the output signal as a rotational angle (upper image), rotational rate (central image), and rotational acceleration (lower image). It is possible to identify, in particular from the upper presentation, that a continuous change in angle over time can be approached without losses.

(71) Case 2: Curve on the Ground

(72) FIG. 6 schematically shows the case of a curve on the ground in the same way. In the case of a curve on the ground, centrifugal forces which rotate the apparent perpendicular vector in a manner similar to a curved flight arise in addition to the gravitational acceleration. However, virtually no roll angle can be set, leading to a transverse force that is perceivable in the aircraft cabin. In simulators, this transverse force is, in turn, represented by a rotation of the cabin. This leads to a difference in the roll movement between aircraft and simulator cabin, which, ideally, must be implemented below the perception threshold. In addition to the general problem of the load multiplication, this leads to errors during entry and exit processes. There may even be acceleration impressions with the wrong sign, in particular as a result of the fast return of the acceleration by means of a filter function on account of the restricted movement space.

(73) On the left, the aircraft is once again shown from the outside, said aircraft being aligned horizontally but the weight vector deviating from the perpendicular to the Earth. This can be identified in the central image, which shows the forces from the view of the aircraft cabin.

(74) In order to implement this in the simulator cabin, the simulator cabin is rotated below the perception threshold while the entire display of the instruments and the external representation in the cabin continues to maintain a horizontal alignment. The result is the perception of a centrifugal force that is based on gravitational acceleration.

(75) In a manner analogous to FIGS. 5a to 5d, FIG. 7a shows the input signals in the filters, FIG. 7b shows the output signals, FIG. 7c shows the translational movements, and FIG. 7d shows the rotational movements.

(76) Case 3: Coordinated Curve.

(77) In the third case, the aircraft is in a sideslip-free curved flight. A roll angle ensures the compensation of the rotation of the apparent perpendicular caused by centripetal forces that act in addition to the weight force. For the pilot, the latter points in the direction of the vertical axis. The load multiplication that occurs cannot be represented for the aforementioned reasons. In principle, the simulator cabin could remain at rest during the entire maneuver, but this leads to an error during the entry and exit of the curve as no roll movement is perceivable. Therefore, at least a small rotation about the roll axis, which is exited again as unnoticeably as possible, is desirable.

(78) If the aircraft is observed from the outside, a deviation arises here between the apparent perpendicular and in the perpendicular to the Earth, with the apparent perpendicular in this case corresponding to the vertical axis from the view of the aircraft cabin (central presentation). Accordingly, the simulator cabin remains unchanged, apart from the entry into and exit from the curve, which are compensated by a short roll movement.

(79) FIGS. 9a to 9d show the corresponding input and output signals of the filter in an analogous manner.

LIST OF EMPLOYED INDICES AND SIGNS

(80) TABLE-US-00001 Meaning Index PA Position of the pilots in the aircraft PS Position of the pilots in the flight simulator Sign {right arrow over (a)}.sub.aa Vector of the acceleration at the pilot seat in the aircraft {right arrow over (f)}.sub.aa Vector of the specific forces at the pilot seat in the aircraft f.sub.aa, x Specific force in the longitudinal direction at the pilot seat in the aircraft f.sub.aa, y Specific force in the lateral direction at the pilot seat in the aircraft f.sub.aa, z Specific force in the vertical direction at the pilot seat in the aircraft f.sub.PS, x Specific force in the longitudinal direction in the flight simulator f.sub.PS, y Specific force in the lateral direction in the flight simulator f.sub.PS, z Specific force in the vertical direction in the flight simulator {right arrow over (f)}.sub.t, PS Vector of the specific force as a consequence of translational acceleration in the flight simulator f.sub.x, t, PS Specific force in the longitudinal direction in the flight simulator as a consequence of longitudinal acceleration f.sub.y, t, PS Specific force in the lateral direction in the flight simulator as a consequence of lateral acceleration f.sub.z, t, PS Specific force in the vertical direction in the flight simulator as a consequence of vertical acceleration {right arrow over (f)}.sub., PA Vector of the specific forces at the pilot seat in the aircraft as a consequence of a position angle {right arrow over (f)}.sub., PS Vector of the specific forces at the pilot seat in the flight simulator as a consequence of a position angle f.sub.x, , PA Specific force in the longitudinal direction at the pilot seat in the aircraft as a consequence of a position angle f.sub.y, , PA Specific force in the lateral direction at the pilot seat in the aircraft as a consequence of a position angle f.sub.z, , PA Specific force in the vertical direction at the pilot seat in the aircraft as a consequence of a position angle f.sub.x, , PS Specific force in the longitudinal direction at the pilot seat in the flight simulator as a consequence of a position angle f.sub.y, , PS Specific force in the lateral direction at the pilot seat in the flight simulator as a consequence of a position angle f.sub.z, , PS Specific force in the vertical direction at the pilot seat in the flight simulator as a consequence of a position angle g Gravitational acceleration {right arrow over (g)}.sub.PA Vector of the gravitational acceleration at the pilot seat {umlaut over (x)}.sub.f Longitudinal acceleration in the coordinate system that is stationary in relation to the cabin .sub.f Lateral acceleration in the coordinate system that is stationary in relation to the cabin y.sub.pp, sim Lateral acceleration at the pilot seat in the flight simulator {umlaut over (z)}.sub.f Vertical acceleration in the coordinate system that is stationary in relation to the cabin .sub.PA Roll angle in the pilot seat of the aircraft .sub.PS Roll angle in the pilot seat of the flight simulator .sub.PA Pitch angle in the pilot seat of the aircraft .sub.PS Pitch angle in the pilot seat of the flight simulator .sub.x, PA Rotation of the apparent perpendicular about the transverse axis in the aircraft .sub.x, PS Rotation of the apparent perpendicular about the transverse axis in the flight simulator .sub., PA Rotation of the apparent perpendicular about the transverse axis as a consequence of a pitch position in the aircraft .sub., PS Rotation of the apparent perpendicular about the transverse axis as a consequence of a pitch position in the flight simulator .sub.t, PS Rotation of the apparent perpendicular about the transverse axis as a consequence of translational acceleration in the flight simulator .sub.f Rotation of the apparent perpendicular as a consequence of specific forces in the lateral direction .sub.ypp Equivalent apparent perpendicular angle about the longitudinal axis for a lateral acceleration .sub.y, PA Rotation of the apparent perpendicular about the longitudinal axis in the aircraft .sub. Rotation of the apparent perpendicular about the longitudinal axis as a consequence of a roll position .sub.t, PS Rotation of the apparent perpendicular about the longitudinal axis as a consequence of a translational acceleration in the flight simulator {right arrow over ()}.sub.PA Vector of the rotation of the apparent perpendicular at the pilot seat in the aircraft {right arrow over ()}.sub.PS Vector of the rotation of the apparent perpendicular at the pilot seat in the flight simulator {right arrow over ()}.sub.t, PA Vector of the rotation of the apparent perpendicular as a consequence of a translational acceleration at the pilot seat in the aircraft {right arrow over ()}.sub.t, PS Vector of the rotation of the apparent perpendicular as a consequence of a translational acceleration at the pilot seat in the flight simulator {right arrow over ()}.sub., PA Vector of the rotation of the apparent perpendicular as a consequence of a position angle at the pilot seat in the aircraft {right arrow over ()}.sub., PS Vector of the rotation of the apparent perpendicular as a consequence of a position angle at the pilot seat in the flight simulator .sub.aa Vector of the rotational accelerations in the aircraft {right arrow over ()}.sub.aa Vector of the rotational speeds in the aircraft {dot over ()}.sub.aa, x Rotational acceleration around the longitudinal axis in the aircraft {dot over ()}.sub.aa, y Rotational acceleration around the transverse axis in the aircraft {dot over ()}.sub.aa, z Rotational acceleration around the vertical axis in the aircraft .sub.aa, x Rotational speed about the longitudinal axis in the aircraft .sub.aa, y Rotational speed about the transverse axis in the aircraft .sub.aa, z Rotational speed about the vertical axis in the aircraft