Method of functional control of an inertial platform of a moving craft

25595/09 U] ll73 *?1ϋ ΠΊ3ΊΓΚ ΓΠ]Π9ϋ>9 ΓΓ^ΝΙΤΈΡJ19 ϊΊϋΊΐ?[ΐΓ? ΠϋΊϋ]

METHOD OF FUNCTIONAL CONTROL OF AN INERTIAL PLATFORM OF A MOVING CRAFT 1

METHOD OF FUNCTIONAL CONTROL OF AN INERTIAL PLATFORM OF A MOVING CRAFT

Field of the invention

The field of the present invention is that of verifying the functioning of an inertial unit for a moving body, the unit being mounted on a movement simulator.

Prior art

Generally speaking, a vehicle such as an aircraft, a rocket, or any other type of moving body is equipped with a navigation and control system comprising an onboard computer, an inertial unit, and control means to provide three-dimensional location, control, and guidance of the moving body.

More, particularly, the control means (for example aerodynamic control surfaces) direct the moving body toward a designated destination or along a particular path by means of commands received from the onboard computer, which receives inertial information from the inertial unit. From this inertial information, the onboard computer estimates the three-dimensional location of the moving body and, as a function thereof, sends instructions or commands to the control means so that the moving body follows its route to its destination.

However, if the inertial unit is suffering from a fault, the computer produces its estimates of the three-dimensional location of the moving body from inaccurate inertial information. The onboard computer therefore makes errors at each calculation step and sends erroneous commands to the control means. Consequently, the true path of the moving body is very different from the path estimated by the onboard computer. Accordingly, if the inertial unit is suffering from a serious fault, the instructions sent by the onboard computer are so 2 inappropriate that there is the risk of the moving body becoming destabilized.

Thus movement simulators are often used to verify inertial units. More particularly, hybrid simulation is 5 used employing real subsystems (for example an inertial unit, a movement simulator, an onboard computer of the moving body, certain other elements of the moving body, etc.) together with mathematical models of other subsystems (for example, further elements of the moving 10 body, the atmosphere, etc.).

At present, in hybrid simulation using inertial units, two situations can arise with regard to information coming from the accelerometers of the inertial unit. 15 In the first situation, the information from the accelerometers is not used and is replaced by information from a mathematical model. This is associated with the fact that the movement simulator cannot reproduce movements in translation. Accordingly, the information 20 from the accelerometers of the inertial unit is incomplete and includes no information concerning the linear displacement of the moving body. Consequently, the onboard computer of the moving body cannot use information from the accelerometers of the inertial unit 25 to determine the location of, to control, and to guide the moving body. The anomalies potentially present in this information are consequently not detected by the hybrid simulation. In other words, the contribution of the accelerometers of the real inertial unit is absent, 30 which rules out detecting even the slightest potential fault affecting the accelerometers.

In the second situation, the information from the accelerometers is complemented by information representing movement in translation calculated by means 35 of a mathematical model. However, the information from the accelerometers of the inertial unit is measured at a fixed point corresponding to the coordinates of the 3 simulation laboratory and is therefore not entirely representative of the information that the accelerometers would provide, for the same physical origin, as the moving body travels along its path around the Earth. For 5 example, the gravitational field at the laboratory fixed point is invariant, whereas that sensed by the accelerometers of the inertial unit onboard the moving body traveling around the Earth varies as a function of altitude and latitude. This difference therefore 10 falsifies the path of the moving body obtained using the inertial unit at the laboratory fixed point and makes the results difficult to interpret. Because of this, that method can detect only a crude fault affecting one or more accelerometers of the inertial unit. 15 Moreover, information from the gyros of the inertial unit is not entirely representative of what this information would be during real movement of the moving body around the Earth.

The way the Earth's rotation is resolved about the 20 axes of the gyros differs depending on whether the inertial unit is located at a point with fixed coordinates (as in hybrid simulation in the laboratory) or is onboard a moving body moving around the Earth. The impact of this incomplete representation complicates the 25 analysis of the hybrid simulation results obtained.

Moreover, the movement simulator must have angular dynamic performance better than or equal to requirements along the whole of the path of the moving body, which can require a very sophisticated and very costly movement 30 simulator.

Object and summary of the invention

The present invention consists in a method of verifying the functioning of an inertial unit for a 3 5 moving body, the unit being mounted on a movement simulator and supplying in real time measurement inertial 4 data representing a movement applied by said movement simulator, said method comprising: • theoretical modeling of the inertial unit on the movement simulator, said theoretical modeling being fed 5 in real time with kinematic data actually executed by the movement simulator and supplying in real time theoretical inertial data representing said measurement inertial data deemed to be measured by the inertial unit on the movement simulator; 10 · simulation modeling comprising modeling of the inertial unit in a real navigation environment, the simulation modeling being fed in real time with control commands and supplying in real time simulation inertial data representing output data from the inertial unit in 15 said real navigation environment, said simulation modeling taking account of said control commands to compute a path of the moving body using the inertial unit; calculating said control commands as a function of 20 the measurement inertial data, the simulation inertial data, and the theoretical inertial data; and • validating the inertial unit by comparing the path of the moving body obtained using the inertial unit with a predetermined reference path. 25 The method of the invention produces a path of the moving body that is very close to reality and is not affected by the problem associated with the measurements effected at a fixed point, thus enabling verification of whether the inertial unit has characteristics conforming 30 to requirements, thereby implying an enhancement in terms of quality and cost.

The measurement inertial data advantageously includes accelerometer information from accelerometers of said inertial unit, and the theoretical inertial data 35 advantageously includes theoretical accelerometer information. 5

Thus any faults emanating from the accelerometers of the inertial unit can be detected.

The measurement inertial data advantageously further includes gyro information from gyros of said inertial 5 unit, the theoretical inertial data advantageously includes theoretical gyro information, and the simulation inertial data includes simulation gyro information and simulation accelerometer information.

Thus the accelerometer and gyro information used to 10 calculate control commands is representative of the information that would be provided by the inertial unit onboard the moving body when traveling along its path around the Earth. This is because this accelerometer and gyro information comprises accelerometer and gyro 15 information from the inertial unit and complementary information calculated in real time by theoretical modeling and simulation.

In one particular implementation of the present invention, said control commands are calculated as a 20 function of inertial data I defined as the sum of the measurement inertial data R plus the simulation inertial data T2 minus the theoretical inertial data T1 (in other words, I = T2 - Tl + R).

Thus an underspecified movement simulator can be 25 used for highly dynamic transients of the path. By virtue of the principle I = T2 - Tl + R, since R and Tl depend on the movements accomplished by the movement simulator, they remain consistent, even if the simulator does not execute the command correctly. This computes an 30 accurate path of the moving body at lower cost.

The inertial unit can be considered valid when the difference between the inertial data I and the simulation inertial data T2 is bounded by a predetermined threshold value (in other words when |l - T2| ε) .

Thus the inertial unit can be validated efficiently and with very high accuracy. 35

<img img-format="tif" img-content="drawing" file="IL197865AD00021.tif" id="idf0001" />

6

The method of the invention further includes phase advance modeling fed in real time with input kinematic commands from simulation modeling and supplying output kinematic commands to the movement simulator to 5 compensate an execution delay inherent to said movement simulator.

Thus execution delays can be compensated, enabling all parameters to remain in phase so as to guarantee that use of the inertial unit and any other sensors 10 simultaneously on the movement simulator is representative.

Said input kinematic commands advantageously have a profile synchronous with that of the kinematic data and the amplitudes of said kinematic data and said input 15 kinematic commands are advantageously consistent.

Thus the kinematic data conforms to the movements actually executed by the movement simulator, enabling the real inertial unit to be simulated in phase with the kinematics of the moving body. This guarantees that use 20 of the inertial unit is representative and assures synchronism relative to the kinematics of the moving body for any other sensors on the movement simulator at the same time as the inertial unit.

The method of the invention advantageously further 25 includes masking modeling downstream from the simulation modeling and upstream from the phase advance modeling, so that the masking modeling is fed in real time with the input kinematic commands, this masking modeling supplying masked kinematic commands to the phase advance modeling 30 to mask at least some of the phases of the movement.

Thus a path can be produced for which the angular range of movement of the moving body is greater than that authorized by the movement simulator.

According to one feature, at least some of the 35 masked kinematic commands depend on a law internal to the masking modeling that is independent of the input kinematic commands .

<img img-format="tif" img-content="drawing" file="IL197865AD00022.tif" id="idf0002" />

7

This enables a path on which the moving body can execute a number of loops or circuits to be computed using a movement simulator having a limited range of angular movement. 5 The invention is also directed to a system for verifying the functioning of an inertial unit for a moving body, the unit being mounted on a movement simulator and supplying in real time measurement inertial data representing a movement applied by said movement 10 simulator, said system comprising: • a theoretical model comprising modeling of the inertial unit on the movement simulator, the theoretical model being fed in real time with kinematic data actually executed by the movement simulator and supplying in real 15 time theoretical inertial data representing said measurement inertial data deemed to be measured by the inertial unit on the movement simulator; • a simulation model comprising modeling of the inertial unit in a real navigation environment, the 20 simulation model being fed in real time with control commands and supplying in real time simulation inertial data representing output data from the inertial unit in said real navigation environment, said simulation model taking account of said control commands to compute a path 25 of the moving body using the inertial unit; • calculation means for calculating said control commands as a function of the measurement inertial data, the simulation inertial data, and the theoretical iner t i a1 data; and 30 · validation means for validating the inertial unit by comparing the path of the moving body obtained using the inertial unit with a predetermined reference path.

The calculation means are adapted to determine said control commands as a function of the inertial data I 35 defined by the sum of the measurement inertial data R plus the simulation inertial data T2 minus the theoretical inertial data Tl. 8

The validation means are adapted to consider the inertial unit valid when the difference between the inertial data and the simulation inertial data is bounded by a predetermined threshold value.

The system further comprises a phase advance model fed in real time with input kinematic commands from the simulation model and supplying output kinematic commands to the movement simulator for compensating an execution delay inherent to said movement simulator

The system further comprises a masking model between said simulation model and upstream said phase advance model, so that the masking model is fed in real time with the input kinematic commands, the masking model supplying masked kinematic commands to the phase advance model to mask at least some of the phases of the movement.

The invention is also directed to a computer program comprising code instructions for executing steps of a control method having at least one of the above features when it is executed on a computer.

Brief description of the drawings

Other features and advantages of the device and method of the invention emerge more clearly from a reading of the description given below by way of nonlimiting illustration and with reference to the appended drawings, in which: • Figure 1 shows diagrammatically a function verification system of the invention for verifying the functioning of an inertial unit for a moving body; • Figure 2 shows highly diagrammatically one example of the function verification system from Figure 1; • Figures 3 to 5 show different embodiments of the invention; and • Figure 6 shows diagrammatically one example of the path of a moving body. 9

Detailed description of embodiments

Figure 1 shows diagrammatically a function verification system 1 for verifying the functioning of an inertial unit for a moving body. This system 1 comprises 5 a movement simulator 3 connected to an electronic data processing device or computer 5 used to execute the instruction codes of an electronic data processing program adapted to execute the method of the invention.

The movement simulator 3 can produce angular 10 displacements about a roll axis Al, a pitch axis A2, and a yaw axis A3. Accordingly, by accommodating an inertial unit 7, the movement simulator 3 can apply thereto angular movements about the roll axis Al, the pitch axis A2, and the yaw axis A3. Note that the inertial unit 7 15 can be mounted on the movement simulator 3 on its own, in the moving body (this is not shown), or in at least part of the moving body.

The function verification system 1 further comprises an onboard computer 9 of the moving body that is 20 connected to the inertial unit 7 and to the computer 5. Note that the various connections between the computer 5, the movement simulator 3, the onboard computer 9, and the inertial unit 7 can be provided by electrical or optical cables, by radio, or by other 2 5 means .

Figure 2 illustrates highly diagrammatically one example of a function verification system 1 of the invention for verifying the functioning of an inertial unit 7 for a moving body (not shown). Note that Figure 2 30 also illustrates the main steps of the control method of the invention.

This method or system is used in a real closed-loop hybrid simulation using real subsystems (the inertial unit 7, the movement simulator 3, the onboard computer 9 35 of the moving body, and possibly at least part of the moving body) and numerical models of other subsystems (for example the propulsion subsystem of the moving body) 10 and the environment (the atmosphere) . The output of the real subsystems is input for the numerical models and the output of the numerical models is input for the real subsystems.

The inertial unit 7 is mounted on the movement simulator 3 (on its own or in at least part of the moving body) and it supplies in real time inertial measurement data (i.e. real inertial data) R representing movement that is applied by the movement simulator 3 and reproduces the angular movements of the moving body in terms of range of movement, speed, and acceleration.

This measurement inertial data R comes from gyros (rate gyros or free gyros) and accelerometers (not shown) of the inertial unit 7 and generally enables the onboard computer 9 of the moving body to determine its location, to control it, and to guide it.

This control system 1 further comprises a theoretical model Ml, a simulation model M2, calculation means 11, and validation means 13. The theoretical model Ml and the simulation model M2 can be included in the electronic data processing device 5 from Figure 1. Moreover, the calculation means 11 include the onboard computer 9 of the moving body.

The hybrid simulation proceeds in real time. Thus calculations and exchanges of data between the onboard computer 9 of the moving body and the theoretical model Ml, the simulation model M2, and the inertial unit 7 are effected at the real frequency timing the operations performed by the onboard computer 9 of the moving body and advantageously within a time period that is less than the period corresponding to that frequency; this applies to each computation loop.

The theoretical model Ml models the onboard inertial unit 7 on the movement simulator 3 and is fed in real time by the movement simulator 3 with kinematic data DI (positions, speeds, and angular accelerations actually executed by this movement simulator 3). Consequently, 11 the theoretical model Ml supplies the calculation means 11 in real time with theoretical inertial data Tl representing measurement inertial data R deemed to be measured by the inertial unit 7 onboard the movement simulator 3.

The simulation model M2 models the inertial unit 7 in a real navigation environment and is fed in real time with control data or commands D2 by the calculation means 11. In return, the simulation model M2 supplies in real time simulation inertial data T2 representing the output data of the inertial unit 7 in a real navigation environment.

Thus the simulation model M2 takes account of the control commands D2 in order to compute a path of the moving body using the inertial unit 7.

The calculation means 11 calculate the control commands D2 as a function of the measurement inertial data R, the simulation inertial data T2 , and the theoretical inertial data Tl.

The validation means 13 can then validate the inertial unit 7 by comparing the path of the moving body obtained using the real inertial unit 7 with a predetermined reference path.

The predetermined reference path can be computed by entirely numerical simulation, for example by the simulation model M2 alone, including modeling of all the real equipment including the onboard computer and the inertial unit. This is known in the art.

The predetermined reference path is obtained in a manner that is known in the art by an entirely numerical simulation model in which all the real equipment including the computer and the inertial unit are rendered numerically.

Thus the validation means 13 compare off-line the predetermined reference path produced entirely numerically with the path of the moving body produced by the function verification system 1 using in real time the 12 real inertial unit 7 and the real onboard computer 9. Comparative examination of the results determines either that the inertial unit 7 is valid, i.e. that it has characteristics conforming to requirements, or that the 5 inertial unit is not valid. In particular, if the two paths are superposed the inertial unit 7 can be considered to be perfect.

The measurement inertial data advantageously includes accelerometer information from the 10 accelerometers of the inertial unit in addition to gyro information from the gyros of the inertial unit 7. Similarly, the theoretical inertial data T1 includes theoretical accelerometer information and theoretical gyro information. 15 The information from the accelerometers of the real inertial unit 7 contributes to location, control, and guidance of the moving body by the onboard computer 9. Thus the anomalies potentially present in this information can be detected by comparing the 20 accelerometer information from the real inertial unit 7 and that from the theoretical model Ml.

Furthermore, the simulation inertial data T2 includes simulation gyro information and simulation accelerometer information. 25 Thus the accelerometer and gyro information used to calculate the control commands is representative of that which would be supplied by the inertial unit onboard the moving body as it travels this path around the Earth. This is because this accelerometer and gyro information 30 includes, in addition to information from the inertial unit located at a point with fixed coordinates, additional accelerometer and gyro information calculated in real time by the theoretical model Ml and the simulation model M2 . 35 The information delivered by the real inertial unit 7 consists of measurements effected at the laboratory fixed point and is not completely representative of the

<img img-format="tif" img-content="drawing" file="IL197865AD00023.tif" id="idf0003" />

13 information that would he delivered by the inertial unit onboard the moving body traveling around the Earth.

More particularly, the theoretical inertial data Tl calculated using the theoretical model Ml corresponds to 5 the theoretical expression of the measurements that the real inertial unit 7 at the laboratory fixed point is deemed to have carried out, and is highly representative, since this calculation uses the kinematic data applied to the real inertial unit 7. Moreover, the simulation 10 inertial data T2 is representative of the measurements effected by the inertial unit 7 onboard the moving body when traveling around the Earth.

Taking account of the measurement inertial data R and the corresponding theoretical inertial data Tl 15 eliminates any error caused by effecting the measurements at the fixed point. Consequently, adding the contribution of the simulation inertial data T2 to the measurement inertial data R and the theoretical inertial data Tl yields a path representative of reality and 20 unaffected by the problem associated with effecting the measurements at the fixed point.

Figure 3 shows diagrammatically one embodiment of a function verification system of Figure 2. In this embodiment, the calculation means 11 include first and 25 second arithmetic operation means Ila and lib applying arithmetical operations to the measurement inertial data R, the simulation inertial data T2, and the theoretical inertial data Tl to calculate inertial data I.

Accordingly, the first arithmetical operation means 30 11a calculate the difference between the simulation inertial data T2 and the theoretical inertial data Tl.

The second arithmetic operation means lib add the resulting difference (T2-T1) to the measurement inertial data R to express the inertial data I according to the 35 formula I = T2 + R - Tl.

The control commands D2 are then calculated as a function of this inertial data I defined as the sum of 14 the measurement inertial data R and the simulation inertial data T2 less the theoretical inertial data Tl.

This embodiment is based on the hybrid simulation by the function verification system 1 using a difference 5 between the measurement inertial data R supplied by the inertial unit 7 onboard the movement simulator 3 and the theoretical inertial data Tl supplied by the theoretical model Ml.

Given the formula I = T2 - Tl + R, since R and Tl 10 depend on the movements accomplished by the movement simulator 3, they remain consistent, even if the simulator does not execute the command correctly. This means that an accurate path of the moving body can be computed using a movement simulator 3 of lower cost. 15 Moreover, when the validation means 13 compare the predetermined reference path and the path of the moving body produced by the function verification system 1 offline, the validity of the inertial unit 7 is a function of the difference between R and Tl. 20 If R = Tl, then I = T2, which corresponds to the two paths being superposed, and the inertial unit 7 is then perfect. The path of the moving body obtained with this kind of inertial unit 7 is then very close to the results obtained using only the real onboard computer 9 in the 25 hybrid simulation.

If R is close to Tl {for example if |R - Tl| = ε where ε is a low value), then a difference is perceived between the two paths but the result remains within acceptable limits and the path can be considered to 30 conform to expectations. The inertial unit 7 then has non-nominal characteristics but remains within the limits or tolerances set out in its specification.

If R is not close to Tl, then the two paths are clearly different, the moving body does not reach the 35 target {true path) but the onboard computer 9 considers it has reached the target {estimated path). This 15 indicates that the inertial unit 7 is suffering from a serious fault.

In other words, the inertial unit 7 can be considered valid if the difference between the inertial 5 data I and the simulation inertial data T2 is bounded by a predetermined threshold value { 11 - T2 | ε) .

Moreover, by calculating the theoretical expression T1 of the measurements that the real inertial unit 7 is deemed to have carried out at the fixed point, 10 subtracting Tl from R, and adding the contribution of the simulation inertial data T2 (representing the measurements effected by the inertial unit onboard the moving body moving around the Earth), a path is obtained representative of reality. It should be noted that 15 subtracting Tl from R removes the contribution of the real inertial unit to the nearest η where η = |R - Tl| and the value of η is directly proportional to the imperfection of the inertial unit 7.

In this example, paths can be computed with a 20 simulation model M2 taking account of the various tolerances of the inertial unit 7. For example, a particular number N of paths can be computed by modulating the variables within their tolerances to obtain a tangle of paths. If the "true" path, i.e. that 25 obtained with the real inertial unit 7, is included in the tangle obtained by numerical simulation, then the inertial unit 7 can be considered valid. If not, it is suffering a fault.

Figure 4 shows diagrammatically that the function 30 verification system 1 further includes a phase advance model M3 that is fed in real time with input data or kinematic commands Cl by the simulation module M2 and supplies output kinematic commands C2 to the movement simulator 3 to compensate an execution delay inherent to 35 the movement simulator 3.

Generally speaking, the system for navigating and controlling a moving body includes a "control" function 16 that is intended to produce instructions for operating control means compatible with the characteristics of the moving body. This is an automatic function, using inertial information or data and possibly information or data of a different kind (if there is an additional, non-inertial unit), judging the accomplishment of the instruction previously commanded in real time, and adapting the level of the next command instruction as required. If the control means do not react quickly enough, the control function will realize this and give stronger commands to compensate this execution delay.

The input kinematic commands Cl advantageously have a profile synchronous with that of the kinematic data DI. Moreover, the amplitudes of the kinematic data Dl and the input kinematic commands Cl are consistent. The simulation of the inertial unit is then synchronized with the angular kinematics of the moving body, guaranteeing that the use of the inertial unit and any other sensors onboard on the movement sensor at the same time is representative.

Figure 5 shows diagrammatically that the function verification system from Figure 4 further includes a masking model M4 , which in this example is downstream from the simulation model M2 and upstream from the phase advance model M3 . Accordingly, the masking model M4 is fed in real time by the simulation model M2 with the input kinematic commands Cl and supplies masked kinematic commands C3 to the phase advance module M3 for masking at least some of the phases of movement of the moving body.

The phase advance model M3 then supplies to the movement simulator 3 output kinematic commands C4 that take account of the masked kinematic commands C3 to produce a path for which the range of angular movement and/or the dynamics of the moving body exceed that authorized by the movement simulator 3.

Thus an angular movement of the movement simulator 3 or cessation of that movement, commanded intentionally 17 from the masking model M4, is taken into account in the measurement inertial data R and the theoretical inertial data T1 but does not affect the control commands D2 .

According to the Figure 3 example, and by virtue of 5 the formula I = T2 - Tl + R, the inertial data I is not affected by any angular movement of the movement simulator 3 or by the stopping of such movement at the command of the masking model M4 .

Moreover, a surveillance function implemented in the 10 masking model M4 limits the commanded range of angular movement before the hardware or software stops of the movement simulator 3 are reached.

It should be noted that the theoretical model Ml, the simulation model M2, the phase advance model M3, 15 and/or the masking model M4 can be included in the electronic data processing device 5 from Figure 1.

Figure 6 shows diagrammatically one example of computing a path around an obstacle.

For example, using an underspecified movement 20 simulator 3 having a capacity in terms of the range of angular movement about the yaw axis of ± 100 degrees, which corresponds to an angular excursion of 200 degrees, a path can be computed having an angular excursion of 270 degrees. Accordingly, in this example, the moving body 25 21 can follow a path 23 from a start point 25 to an arrival point 27 circumventing an obstacle 29 by means of an angular excursion of 270 degrees.

During computation of the path 23, the angular position of an axis of the movement simulator 3 can be 30 modified. For example, the yaw axis A3 of the movement simulator 3 can be controlled independently of the yaw of the moving body 21.

At least some of the masked kinematic commands C3 can advantageously depend on a law internal to the 35 masking model M3. In particular, the internal law is independent of the input kinematic commands Cl. Accordingly, a movement command can be injected into the 18 movement simulator 3 from the masking model M4 independently of input kinematic commands Cl coming from the simulation module M2. This enables a path to be computed in which the moving body can execute a plurality 5 of loops with a movement simulator 3 having a limited range of angular movement.

Moreover, the path can have a dynamic range greater than that authorized by the movement simulator 3.

For example, should the input kinematic commands Cl 10 for a transient event command a speed greater than the speeds that the movement simulator 3 is capable of generating, the characteristics of the kinematic data Dl actually executed by the movement simulator 3 are reduced relative to the input kinematic commands Cl. This 15 insufficient movement is taken into account in the measurement inertial data R and the theoretical inertial data Tl. Nevertheless, by virtue of the formula I = T2 -Tl + R, the inertial data I is not affected by the movement being insufficient. Thus the path is not 20 affected by the underspecification of the movement simulator 3.

The invention is also directed to a computer program downloadable from a communications network and comprising program code instructions for executing steps of the 25 control method of the invention when it is executed on the computer 5. This computer program may be stored on a computer-readable medium and may be executable by a microprocessor.

This program may use any programming language and 30 take the form of source code, object code or code intermediate between source code and object code, such as a partially-compiled form, or any other desirable form.

The invention is also directed to a computer-readable information medium containing instructions of 35 the above computer program.

The information medium can be any device capable of storing the program. For example, the medium can include 19 storage means, such as a ROM, for example a CD ROM or a micro-electronic circuit ROM, or magnetic storage means, for example a floppy disk or a hard disk. □’Ewan pp^a , crnxan ;w:n ατα inia^a pnow pnszn irn nr -jaoa ,ρνιη pxan paoana ma™ na^maa np’ioa .zrtwan rwaa mp’pan ρ-ίΛ axnria ......r. Geo&lt;su«r □innn Pi? »*···»· · ·* sw J«w\ _ ♦ ^~... 1$ Apr2)1211:1T:254G300 .(mcna nannn) cras&amp;amp;’an nwa

<img img-format="tif" img-content="drawing" file="IL197865AD00024.tif" id="idf0004" />