GB2497153A

GB2497153A - Method for determining a measurable target variable and a corresponding system

Info

Publication number: GB2497153A
Application number: GB1215204.7A
Authority: GB
Inventors: Heikki Nieminen; Erik Lindman
Original assignee: Suunto Oy
Current assignee: Suunto Oy
Priority date: 2011-09-16
Filing date: 2012-08-28
Publication date: 2013-06-05
Anticipated expiration: 2032-08-28
Also published as: DE102012216194A1; GB201215204D0; DE102012216194B4; FI20115910A0; GB2497153B; FI123761B; FI20115910A

Abstract

The invention relates to a method for determining a target variable, such as speed or altitude, to be measured, in a mobile device (such as a wrist-mounted computer or mobile telephone), and a corresponding mobile system. In the method, a first physical variable is measured with the aid of a first sensor and a second physical variable with the aid of a second sensor. The second physical variable is typically different to the first physical variable, or at least is measured using a different technique. With the aid of the measurements, the value of the target variable is calculated with the aid of the measurement of the first and second physical variables, in such a way that an estimate for the target variable is determined with the aid of at least the measurement of the first physical variable (step 32), at least a first error estimate is determined (33a, 33b), which depicts the accuracy of the measurement of the first physical variable, and the estimate of the target variable is filtered (step 35) at a strength that depends on both the first error estimate and the measurement of the said second physical variable.

Description

Method for Determining a Measurable Target Variable and a Corresponding System

Field of the invention

The invention relates to sensor-fusion technology in mobile devices, i.e. more specifically the processing of information provided by many sensors. The invention particularly relates to improving the accuracy of a target variable measured with the aid of a mobile device or system. The sensors can be, for example, a GPS sensor or pressure sensor, as well as an acceleration sensor, and with their aid the speed or altitude respectively of the target variable to be measured. The mobile device can be, for example, a wristop computer, a mobile telephone, other portable device or a sensor unit, or some functional combination of these.

Prior art

GPS speed on the wrist or elsewhere on the body contains a great deal of noise, but has a very small bias error, i.e. a systematic error. I.e., the accuracy of the GPS speed shown on the frequency level is very close to DC and diminishes rapidly as the frequency increases, as shown in Figure 1 with the aid of curve I. The measurement accuracy can be improved using traditional signal filtering methods, or with the aid of GPS-Doppler measurement.

Nevertheless, at a typical measurement frequency of 1 Hz, and with the person walking, the noise in a purely GPS-based speed measurement can be ii the order of as much as 20- 3⁄4, compared to the signal.

On the other hand, the speed estimated from an acceleration sensor on the wrist or elsewhere on the body contains less noise, so that its accuracy will remain good when the frequency increases, up to a certain limit. However, there can be even a large bias error in speed measured in this way. This means that the best accuracy of speed estimated from acceleration is poorer than when using CJPS. The accuracy in the frequency plane of speed measured with the aid of an acceleration sensor is typically according to curve 2 of Figure However, in practice the aim would be to obtain speed accuracy on the frequency level according to curve 3 of Figure 1, i.e. accurate measurement over a wide range. The aim is also to create measurement that reacts sufficiently well to changes in the state of motion, but poorly to error sources relating to the measurement event itself.

Similar problems also relate to the liPS-based determining of altitude and vertical speed.

Further, a somewhat similar problem also relates to the determining of altitude information with the aid of a pressure sensor, though in this case the errors are caused by slow (low-frequency) variations in atmospheric pressure.

Summary of the invention

The invention is intended to provide a new type of method for determining, in a mobile device, a target variable being measured, and a corresponding system. The invention's objective is particularly to improve the accuracy of vertical and/or horizontal speed and/or altitude calculated with the aid of positioning-sensor information, or of altitude or vertical speed calculated with the aid of pressure information, in varying movement and ambient conditions.

In the invention, sensor fusion is exploited in a new way, i.e. more specifically the calculation of a desired variable exploits the information provided by at least two different sensors measuring the same or a differcnt physical variable.

The method and system according to the invention are defined with greater precision in the independent Claims.

In the method according to one embodiment, a first and a second physical variable are measured with the aid of a first and second sensor respectively. The value of the target variable is calculated with the aid of the first and second measurement, in such a way that -an estimate of the target variable is determined with the aid of the measurement of at least a first physical variable, -at least a first error estimate is determined, which depicts the error of the measurement of the first physical variable, -an error estimate for the computation model is determined, which estimates how vell the computation model corresponds to the real situation, with the aid of the measurement of at least a second physical variable, and -the estimate oIthe target variable is filtered at a strength that depends on both the error estimate of the measurement and the error estimates of the said computation model.

The second physical variable can be a different variable to the first physical variable, or at least measured using a different technique. The difference in technique can be, for instance, a different placing of the sensor, or a different measurement modeL by means of which the measurement is converted into the target variable. Thus, the first and the second sensor are typically based or a different operating principle, even when they are measuring the same physical variable. For example, horizontal speed (speed of progression) can be measured with the aid of a satellite-positioning sensor and with the aid of an acceleration sensor. Correspondingly, altitude or rate of climb (vertical speed) can be measured with the aid of a satellite-positioning sensor, an acceleration sensor, and a pressure sensor. The first and second sensors are preferably based on measurement techniques that have error profiles that differ essentially from each other, as functions of the measuring frequency.

However, some estimate of the target variable or its changc must be abLe to be derived from the data provided by both sensors, either directly or through a mathematical model and/or initial data.

According to one central embodiment, in the method speed is measured with the aid of a satcllite-positioning sensor (such as a GPS sensor) and acceleration with the aid of an acceleration sensor. The final speed value to be given to the user is calculated with the aid of the speed and acceleration measurements, in such a way that -an estimate is determined for speed, with the aid of a satellite-positioning measurement and/or an acceleration measurement, -a first error estimate is determined, which depicts the error in the satellite-positioning-based speed measurement, -a second error estimate is determined, which depicts the error in the acceleration-based speed measurement, -a first error estimate for the computation model is determined with the aid of a satellite-positioning measurement, a second error estimate for the computation model is determined with the aid of an acceleration measurement, and -the speed estimate is filtered using a strength that depends on the error value of the said measurements and the said error values of the computation model.

According to a preferred embodiment a speed estimate is determined, at least when the preset quality conditions of the measurement are met, with the aid of both satellite-positioning measurement and acceleration measurement, weighting them in a desired manner. It is also possible to determine error estimates, which depict the magnitude of the errors of the measurements of speed. These error estimates can be further used to determine the weightings of the calculation of the speed estimate. Further, it is possible to determine a second error estimate, which depicts the magnitude of the error of the computation model. The error of the computation model and the error of the speed estimate can be further used to determine the filtering strength of the speed estimate.

The speed can be either the horizontal or the vertical speed, or the sum of these velocities.

The method according to the invention can be performed either entirely or partly in a wristop computer. If it is performed only partly in a wristop computer, some part of it can be performed in a remote sensor, which can be located in a separate device unit, or belong to some second device unit, such as a mobile telephone. The part in question can be the measurement of the first and'or second physical variable, i.e. for example, in the ease of the speed measurement described above, the measurement of speed and/or acceleration, and/or of the calculation.

An embodiment is particularly preferred, in which the acceleration measurement is performed with the aid of an acceleration sensor in a wristop device, because acceleration measurement from the wrist is very reliable, due to the natural movement of the hand.

The method according to the invention can aLso be perfoniied entirely or partly in a mobile telephonee. If it is performed only partly in a mobile telephone, some part of it can be performed in a remote sensor, which can be located in a separate device unit, or belong to some other device unit, such as a wristop computer. The part in question can be the measurement and/or calculation of the first and/or second physical variable, i.e. for example, in the case of the speed measurement described above, the measurement and/or calculation of speed and/or acceleration.

Generalizing, it can be stated that the method according to the invention can also be performed entirely or partly in a portable device with a display, which is arranged to display the calculated value of the target variable to the user.

On the other hand, the method according to the invention can also be performed entirely or partly in a portable device without a display, such as in a satellite-positioning module, which is linked wirelessly to a portable device with a display, such as with a wristop computer or mobile telephone, and/or the stored data can be read later, for example, to a computer. The advantage of this embodiment is the reduction in power consumption in the portable device with a display.

The system according to the invention comprises corresponding device units, the wireless communications means that may be required between them, and is arranged to implement the method according to the invention. Various examples of alternatives and their advantages will be described later in greater detail.

Considerable advantages are gained with the aid of the invention. When the target variable changes rapidly, this can be detected and, in turn, the filtering level, which is used when calculating the value of the target variable, can be altered. Correspondingly, if the values of the target variable given by the sensors differ considerably from each other, it can he concluded that there must be some explainable error source in one of the measurements.

Such a situation is, for example, when a GPS sensor is taken under a large bridge, where there is no UPS signal. If the acceleration sensor still shows that the device is moving, the speed calculated on the basis of the acceleration sensor can be given greater weight in the final determining of the speed.

According to one preferred embodiment, the said estimate for the target variable is calculated with the aid of the measurement of a first and second physical variable, and thrther a second error estimate is determined, which depicts the accuracy of the measurement of the second physical variable. Finally, the value of the target variable is calculated by filtering the estimate of the target variable at a strength that depends on both the first and the second error estimates. Thus, the accuracy of both the first and the second sensor can be taken into account, before that final result is displayed to the user.

According to one embodiment, in the calculation of the target variable the strength of the filtering is increased when the accuracy of the measurement of the first and/or second physical variable weakens, and vice versa. Thus, variations in the target variable due to a measurement error are not transmitted detrimentally to the device's user, but when the measurement error is small the time resolution of the measurement will nevertheless remain good.

According to one preferred embodiment, during the measurement the rate of change of the target variable is detected, either on the basis of an estimate of it or of its final value, or directly from the measurement data of the first or second sensor, if it is noticed that the rate of change of the target variable exceeds a preset limit, or is increasing, the strength of the filtering is reduced in the calculation of the target variable. The method then reacts faster to variations in conditions and the user can be provided with information that is more real-1 5 time.

As stated above, one significant practical application of the invention presents a solution, in which thc target variable is speed and the first sensor is a satellite-positioning sensor, such as a UPS receiver, in that case, the first physical variable is speed or absolute/relative position. If position is measured, the speed can be calculated on the basis of position and time information. On the other hand, if the UPS speed is measured, for example, with the aid of the Doppler effect, the speed is obtained directly from the GPS data. A combination of these ways of measurement is also possible.

As a second possible application of the invention, a solution is referred to, in which the target variable is altitude or vertical speed ("rate of climb"), the first sensor is an atmospheric-pressure sensor, and the first physical variabic is correspondingly atmospheric pressure. Altitude and vertical speed, or at least estimates for them, can be calculated with the aid of atmospheric pressure, if the atmospheric pressure at sea-level or some other reference level is known.

In all of the aforementioned applications, one sensor is preferably an acceleration sensor and the one physical variable is acceleration. Acceleration measurement shows the device's state of motion and on the basis of this it is then possible to calculate an estimate of its speed. If desired, an error estimate depicting measurement error can also be determined for acceleration measurement, and can be exploited, along with or instead of only the state of motion, for filtering temporally at the desired strength according to the invention an estimate of the target variable calculated on the basis of the first sensor.

As one possible application of the invention, it is also possible to refer to a solution, in which, with the aid of satellite positioning, a first physical variable, which is for example altitude or vertical speed, is measured. An atmospheric-pressure sensor measures a second physical variable, which is atmospheric pressure. In that case, an estimate depicting the vertical state of motion can be determined using atmospheric-pressure measurement, which can be used to control the filtering strength of the physical variable measured using GPS.

In the following, the practical implementation and advantages of the invention are described in greater detail, with reference to the accompanying drawings.

Brief descriptiun of the drawings Figure 1. GPS speed (1), speed (2) estimated from an acceleration sensor on the wrist or elsewhere on the body, and the accuracy of the combined speed (3) as a function of frequency.

Figure 2. Object model of a sensor-fUsion system according to one embodiment, which combines data from several speed sources, and then filters the combined speed information adaptively.

Figure 3. Flow diagram of the method according to the invention, according to one embodiment.

Figure 4. Example of speed measurement data (the blue line on which there arc squares is the speed calculated from the wrist using an acceleration sensor and the red line on which there arc circles is the speed obtained from GPS measurement), the speed corrected using a traditional filter (broken line), and the speed calculated using the method according to the invention (solid thick line).

Figures 5a -Sf show measurement-system implementations according to different embodiments of the invention

Detailed description of embodiments

The basic principle of the invention is first examined with the aid of the object model shown in Figure 2. In it, three different ways of determining speed are shown byway of example, speed determining 14 made on the wrist using acceleration-sensor measurement 15, GPS-based speed determining 16, and speed determining 18 made using a shoe sensor (typically based on the acceleration of the foot). The velocities determined are combined centrally with thc aid of data processing 22. In the example of Figure 2, information on the person's stale of motion obtained from the acceleration measurement 15 is exploited in the filtering 20 of the combined speed.

Figure 2 also illustrates how the system can be used for the auto-calibration of the measurements, with the aid of stages 28. 24, if a measurement is, or measurements are avaiLable, the error estimate of which is less than the error estimate of the measurement to be calibrated. For example, a combined variable can be used, which does not, however, contain information on the measurement to be calibrated, and by using this combined variable the measurement model is calibrated, in order to obtain more accurate measurement values in the future.

The necessary calculation is performed in the data-processing unit 11.

The combining of the velocities can be done in many ways. One way is to form for each separate speed measurement i (corresponding speed v1) a relative number (R,), which depicts how much error there is in the measurement of the speed. The greater the number, the greater the error estimate of the measurement, and the smaller the number, the smaller the error estimate of the measurement. The number for GPS can be formed, for example, with the aid of the GPS's HDOP number (horizontal dilution of precision) and the number of satellites seen. In acceleration and shoe determining, it is possible to use a relative or absolute pre-determined error estimate, or an error estimate determined dynamically during the perfonnance.

The combined speed Vco,nhined is then obtained using the equation I Z&-s = Z The combined measurement error estimate Rflbi,,ed of the speed is obtained from the equation

-_____________________

tc:,bmed -V in which R,0zkS( is the smallest of the error estimates R. The filtering of the combined speed can be done, for example, with the aid of a Kalman filter (Introduction to random signals and applied kal,nan filtering, 3 edition, R. Grover and P Hwang, John Wiley & Sons, 1997). In a Kalman filter, a linear model of the system to be modelled is built, which takes into account the measurement error and the error of the system model. In the case of the example, the Kalman filter consists of only one state, which is the filtered speed VJpfte,-s that is desired as the end result. Because the Kalnian fluter filters the combined speed, the measurement error is obtained from the aforementioned equation, by means of which the measurement-error estimates of the various velocities are I 5 combined to form a single number The system-mode! error Wsystem can be obtained by combining the system-error estimates calculated from the various measurements. For example, using the following equation = + (v9) + ffootpod (v!QQQd) in which the functionsf depict the system-error estimates of each measurement. The calculation of the system-error estimate of a measurement depends on the filtering model used. In the case of the example, the Kalman filter consists of only a single state, which is speed. The system-error estimate should then be depicted as changes detected in speed.

After this, the equations of the Kalman filter are used to create an adaptive filter, which reduces the filtering when the measurements are accurate and, in turn, increases the filtering when the error of a measurement increases. Modelling of the error of the system model permits the filtering to be reduced if rapid changes are detected in the system. If, on the other hand, the frequency band is narrow, i.e. changes do not take place in the speed, filtering can be increased.

Figure 3 shows a flow diagram of the method according to the invention, according to one embodiment. The determining of the speed starts in stage 30. After this, the measurement of the first and second physical variables starts in stages 31 a and 3 lb. Once sufficient data has been collected, the combined speed is calculated in stage 32, for example using the equation shown above. The system-model error is calculated in stage 34. After this, the combined speed is filtered in stage 35, in order to obtain a speed estimate with less noise.

In this, the error estimate of the first measurement, determined in stage 33a, is exploited, along with either the second measurement or the error estimate determined for it in stage 33b. Once the filtered speed has been calculated, the result is typically stored in the device's memory and/or notified to the user in stage 36.

Figure 4 shows the speed (squares) calculated from the wrist using an acceleration sensor and the speed (circles) obtained from GPS measurement. Both measurements are in the correct order of magnitude, but contain relatively much noise.

There is considerably less noise in the speed (broken line) combined and filtered in a traditional manner (median filtering), but still contains relatively sharp variations.

Particular attention should be paid to the slowness of the change in the estimate seen when speed changes, i.e. too high a speed estimate, which is due to the use of a filtering function that does not adapt to the situation, which always filters data over a constant time.

The speed that has been combined and filtered according to the invention (the solid thick line) is noticeably even at the start of the movement and reacts rapidly to a genuine change in the speed. This is due to the fact that the filter takes into consideration a steep change in speed from the GPS and/or acceleration data and the filtering is diminished. Thus, despite the fact that noise is filtered effectively, the speed calculatcd in the manner of the invention reacts to genuine change in speed more rapidly than the median filtered speed.

The principle described above can be applied not only to the measurement of speed, but also to altitude measurement. In that case, with the aid of acceleration measurement and/or pressure measurement it will be possible to detect the person's state of motion and, with the aid of pressure measurement or GPS measurement, to adjust the temporal filtering of the estimated altitude or vertical speed, in such a way that, at points of change in motion or particularly the state of motion, the filtering will be less than when stationary, so that the system will react more rapidly to real changes in altitude.

According to one embodiment, the system comprises a possibility to select from at least S two alternatives the type of sport being performed and/or the location of the sensor, through suitable interface elements. For example, in the case of an acceleration sensor, both the type of sport (e.g., running/walking) and the location of the acceleration sensor (eg., sole of foot/thigh/wrist/shoulder) will affect the signal's strength, quality, and specific features. Thus, several different signal-processing algorithms can be programmed into the device, from which the most suitable for the conditions is selected according to the type of sport or the location of the sensor. In certain situations, the selection of the sport and thus also the changing of the algorithm can also take place with the aid of detection, for example by utilizing a defined step frequency.

More generally, it can be stated that the system can comprise a condition parameter, which can he given different values and which effects, in turn, which manner of calculation of the target variable will be used.

Figure Sa shows an example of a measuring system utilizing the method according to the invention, which comprises a wristop computer SOa, which can receive satellite-positioning information from a satellite-positioning system 51a, as well as acceleration information, for exampk from an acceleration sensor 52a installed in a shoe. On the basis of the measurement data, the wristop computer is arranged to calculate the filtered speed according to the invention.

Figure Sb shows a system corresponding to that of Figure Sa, but in it the acceleration sensor 52b is integrated directly into a wristop computer SOb. The satellite-positioning system 41b is linked wirelessly 50h to the wristop computer. In the case of altitude or vertical-speed measurement, the sensor 52b can also be a pressure sensor.

Figure Sc shows a variation of the system of Figure 5a, in which instead of a wristop computer a mobile telephone or tablet device SOc is used as the terminal device, which receives information from a satellite-positioning system SIc and an acceleration sensor 52c'.

Figure Sd shows a system, in which both a mobile telephone SOd and a wristop computer 53d are used. The mobile telephone SOd can receive and store satellite-positioning information from a system 51d and forward position or speed information to the wristop computer 53d. The acceleration sensor 52d' can be connected directly to the wristop S computer 53d or the mobile telephone 40d.

Figure Se shows a system, in which a wristop computer 53e and a separate GPS measuring device SOe (a GPD pod') are used, which communicate wirelessly with the wristop computer 53e. The GPS measuring device SOe can thus forward position or speed information to the wristop computer 53e. A separate acceleration sensor 52e' can be linked directly to the wristop computer 53e or the GPS measuring device SOc.

Figures 5c -Se also show variations, in which the acceleration sensor 52c", 52d", 52d", 52e", or 52e" is located not in a shoe, but in a mobile telephone, a GPS measuring device, or a wristop device.

Figure Sf ftirther shows a particularly interesting application in the form of a system corresponding to that of Figure 5; which uses a wristop computer 53f and a separate satellite-positioning module, such as a GPS measurement device SOf (a GPS pod'), which communicates wirelessly with the wristop computer 53f and also contains an acceleration sensor 52f. Thus, a separate acceleration-sensor device is not required; instead acceleration information too can be transmitted wirelessly from the GPS measurement device SOf to the wristop computer 53f. In addition to, or instead of the acceleration sensor, there can be a pressure sensor in the GPS measurement device, so that it will also be suitable for altitude measurement according to the method.

Some of the embodiments olFigures Sc -Sf have the advantage that GPS measurement takes place with the aid of its own power supply, and the operating time of the wristop device or mobile telephone thus increases. In these cases, the wristop device can also be designed to be dry-cell-battery operated instead of rechargeable-battery operated.

Especially advantageous is an embodiment, in which at least one of the measurement, preferably both measurements, and the calculation of the target variable are performed outside the wristop device or mobile telephone. Such a situation is, for example, in the system according to Figure 5f in which the GE'S measurement device 50f also contains the means for performing the necessary calculation. The wristop device or mobile tetephone then acts essentially as only a display, so that its power consumption is very small compared to a situation in which one or all of these operations would be performed in it.

This is important, if it is remembered that exercise can last for a long time and a wristop device or mobile telephone will typically have other power-consuming applications too while a sport is performed (for example, heart-rate measurement or music functionalities), and operations reLating to the measurement of speed and position typically consume a relatively great deal of power.

Instead of a satellite-positioning system, it is also possible to use a terrestrial wireless positioning system, for example a base-station-based positioning system.

If the satellite-positioning sensor and/or the acceleration sensor is located in a different device unit to the actual terminal device, for example according to any of the solutions described above, the error estimate depicting the accuracy of the positioning and/or the acceleration measurement, typically a so-called quality factor or factors, can be transmitted wirciessly from the sensor unit to the tcrmimil device, in the terminal device, such a quality factor or some variable derived from it can be used directly as an error estimate.

Depending on the variable being measured, the quality factor can depend on, for example, the type of sensor, the measurement location, and/or the values of the sensor.

Particularly in the case of satellite-positioning, the quality factor depends greatly on the values of the data provided by the satellite-positioning sensor. In the case of the GPS standard, if the GPS gives only ordinary NMEA (National Marine Electronics Association) information, the quality factor can be calculated from the following values contained in an NMEA message: 1. number of satellites in solution 2. horizontal dilution of precision, HDOP If available, it is also possible to utilize the value according to the SIRF IV standard: 3. estimated horizontal position error, EHPE, on the basis of which a better quality factor will be obtained.

Thus, in order to calculate the quality factor, either these numbers separately, or alternatively the quality factor calculated on their basis should be transmitted to the terminal device, In the case of a shoe acceleration sensor, a cycling acceleration sensor, or a wrist acceleration sensor, typically only the type of sensor is of significance, so that on the other hand it will be sufficient to transmit only information on the type of sensor as the applied value, instead of a quality factor.

The values required to calculate the quality factor, or the quality factor itself can be transmitted according to a suitable radio protocol, such as the ANT or Bluetooth protocols.

Combinations of the solutions described above and variations other than those described in detail are also possible.

Claims

<claim-text>Claims: I. Method for determining a measurab]e target variable, such as speed or altitude, in which method -a first physical variable is measured with the aid of a first sensor, -a second physical variable is measured with the aid of a second sensor, -the value of the target variable is calculated with the aid of the measurement of the first and second physical variable, characterized in that the calculation of the value of the target variable comprises the following stages: -an estimate for the target variable is determined with the aid of the measurement of at least the first physical variable, at least a first error estimate is determined, which depicts the error in the measurement of the first physical variable, and -the estimate of the target variable is filtered at a strength that depends on both the first error estimate and the measurement of the said second physical variable.</claim-text> <claim-text>2. Method according to Claim 1, characterized in that the said estimate for the target variable is calculated with the aid of the measurement of the first and second physical variable.</claim-text> <claim-text>3. Method according to Claim 1 or 2, characterized in that -a second error estimate is determined, which depicts the error of the measurement of the second physical variable, and -the estimate of the target variable is filtered at a strength that depends on both the first and the second error estimates.</claim-text> <claim-text>4. Method according to any of the above Claims, characterized in that, in the calculation of the target variable, the strength of the filtering is increased when the error estimate of the first and/or second physical variable increases, and vice versa.</claim-text> <claim-text>5. Method according to any of the above Claims, characterized in that -the rate of change of the target variable is detected, -the strength of the filtering in the calculation of the target variable is altered, depending on what the rate of change of the first and/or second variable is.</claim-text> <claim-text>6. Method according to any of the above Claims, characterized in that the estimate of the target variable is filtered using a filter adapting time, the length of which filtering period is arranged to alter on the basis of the changes detected in the measurement of the first and/or second physical variable.</claim-text> <claim-text>7. Method according to any of the above Claims, characterized in that -the target variable is speed, -the first sensor is a satellite-positioning sensor, and -the first physical variable is speed, or absolute or relative location.</claim-text> <claim-text>8. Method according to any of Claims 1 -6, characterized in that -the target variable is altitude, -the first sensor is an atmospheric-pressure sensor, and -the first physical variable is atmospheric pressure.</claim-text> <claim-text>9. Method according to any of Claims I -6, characterized in that -the target variable is altitude, -the first sensor is a satellite-positioning sensor, and -the first physical variable is altitude.</claim-text> <claim-text>10, Method according to any of the above Claims, characterized in that the second sensor is an acceleration sensor and the second physical variable is acceleration.</claim-text> <claim-text>11. Method according to any of the above Claims, characterized in that the second sensor is a pressure sensor and the second physical variable is atmospheric pressure.</claim-text> <claim-text>12. Method according to any of the above Claims, characterized in that it is performed entirely or partly in a wristop computer.</claim-text> <claim-text>13. Method according to any of the above Claims, characterized in that it is performed entirely or partly in a mobile telephone.</claim-text> <claim-text>14. Method according to any of the above Claims, characterized in that it is performed entirely or partly in a satellite-positioning module or other device without a display.</claim-text> <claim-text>15. Method according to any of the above Claims, characterized in that it is performed entirely or partly in a portable satellite-positioning device, an outdoor-recreation computer, a yachting computer, or other device with a display.</claim-text> <claim-text>16. Method according to any of the above Claims, characterized in that the estimate of the target variable is filtered with the aid of a Kalman filter.</claim-text> <claim-text>17. Method according to any of the above Claims, characterized in that the first and the second sensors are based on different measurement techniques, which have essentially different error profiles as a function of the measurement frequency.</claim-text> <claim-text>18. Method according to any of the above Claims, characterized in that the filtering of the target variable is implemented in such a way that -an error estimate is determined for the computation model, which depicts how well the computation model used corresponds to the real situation, with the measurement of at least the measurement of the second physical variable, and -the estimate of the target variable is filtered at a strength that depends on both the error estimate of the measurement and the error estimate of the said computation model.</claim-text> <claim-text>19. Method according to any oIthe above Claims, characterized in that at least two of the following are located in separate device units, which are arranged to communicate wirelessly with each other a first sensor, a second sensor, means for calculating the value of the target variable, and a device unit, which contains at least one of the said sensors and is arranged to transmit an error estimate of the measure made with the aid of the sensor in question, or the necessary data for calculating this, wirelessly to the device unit that comprises means for calculation the value of the target variable.</claim-text> <claim-text>20. Method according to any of the above Claims, characterized in that it comprises a circumstance parameter, which can have different values and which value in turn affects the manner of calculation used for the target variable.</claim-text> <claim-text>21. Method according to Claim 20, characterized in that the circumstance parameter is the type of sport, 22. System for determining a target variable to be measured, such as speed or altitude, which system comprises -a first sensor for measuring a first physical variable, -a second sensor for measuring a second physical variable, -means for calculating the value of the target variable with the aid of the measurement of the first and second physical variable, characterized in that the means for calculating the value of the target value comprise -means for determining an estimate for the target variabLe with the aid of the measurement of at least the first physical variable, -means for determining at least a first error estimate, which first error estimate depicts the accuracy of the measurement of the first physical variable, and -means for filtering the estimate of the target variable at a strength that depends on both the first error estimate and the measurement of the said second physical variable.23. System according to Claim 22, characterized in that it is arranged to perform the method according to any of Claims 1 -18.24. System according to Claim 22 or 23, characterized in that at least two of the following are located in separate device units, which are arranged to communicate wirelessly with each other: the first sensor, the second sensor, the means for calculating the value of the target variable.25. System according to Claim 24, characterized in that the second device unit is a wristop device, or other portablc device with a display.26. System according to Claim 24 or 25, characterized in that the second device unit is a satellite-positioning module or outdoor-recreation computer, a yachting computer, or other device with a display.27. System according to any of Claims 22 -26, characterized in that either or both sensors, and optionally also the means for calculating the value of the target variable, are located in a device unit without a display.28. System according to Claim 27, characterized in that it comprises a device unit with a display, in which case the said device unit without a display is arranged to be in a wireless connection with the device unit with a display, in order to display the value of the target variable.29. System according to any of Claims 24 -28, characterized in that the device unit that contains at least one of the said sensors is arranged to transmit wirelessly an error estimate of the measurement made with the aid of the sensor in question, or the necessary data for calculating this.30. System according to any of Claims 22 -28, characterized in that the first and the second sensor, as well as the means for calculating the value of the target variable, are located in the same portable device unit.31. System according to Claim 30, characterized in that the said device unit is a wristop device, a mobile telephone, a satellite-positioning device with or wilhout a display, an outdoor-recreation computer, a yachting computer.</claim-text>