US20040111212A1

US20040111212A1 - Method for determining a track of a geographical trajectory

Info

Publication number: US20040111212A1
Application number: US10/330,447
Authority: US
Inventors: Wim Broeck
Original assignee: Individual
Current assignee: Individual
Priority date: 2002-11-22
Filing date: 2002-12-30
Publication date: 2004-06-10
Also published as: WO2004048892A1; JP4553191B2; WO2004048893A1; JP2006507493A; AU2003302430A1; EP2270429B1; EP1563255B1; EP1563255A1; EP2270429A2; ATE549601T1; EP2270429A3; HK1152747A1; JP5144711B2; US7623961B2; JP2010210634A; US20070203639A1; HK1083235A1; AU2002357415A1

Abstract

A method for determining a track of a geographical trajectory which has a predetermined start location and a predetermined end location, said method comprising the steps of: moving a data collecting device from the start location to the end location, the device comprising a set of sensors providing for sample wise collection physical measurement data during the movement of the device and enabling the determination of a heading profile, a pitch profile, a distance profile and a roll profile from said measurement data; determining said heading profile, said pitch profile, said distance profile and said roll profile based on the collected measurement data; and determining said track based on said heading profile, said pitch profile, said distance profile and said roll profile. All said physical measurement data are stored during the movement of the device from the start location to the end location. After retrieval of the stored measurement data and before determining said profiles, the method further comprises the steps of: determining at least first, respectively second trajectory parameters based on measurement data of a first, respectively second sensor of said set and mapping said first trajectory parameters on said second trajectory parameters; determining an error compensation for said measurement data of said first and said second sensors based on said mapping; and correcting said measurement data of said first and second sensors by means of said error compensation.

Description

The present invention relates to a method for determining a track of a geographical trajectory.

The track which is obtainable by the method of the invention can for example be used for including the location or path of a whole trajectory in a global coordinate system, for determining the length of a trajectory, for obtaining information about specified sections of a trajectory, for determining the geographical location of a specified event or other purposes.

A method for determining a track is for example known from FR-A-2610100, which relates to a method and device for determining deformations and the path of a pipe. The method known from FR-A-2610100 comprises the steps of moving a data collecting vehicle, which comprises sensors such as accelerometers and gyroscopes, through the pipe, collecting and storing samples during the movement and afterwards transferring the collected samples to a computer for determining the length and horizontal projection profile of the pipe. A known problem of sensors such as accelerometers and gyroscopes is that for measurements over a long period of time, the measurements become inaccurate due to a drift of the sensors, i.e. a build-up of errors. There is no teaching in FR-A-2610100 of how such drifts may be compensated or, in other words, how the accuracy of the collected data can be improved.

EP-A-534338 and U.S. Pat. No. 2002/0005297 relate to methods and devices for steering or navigating a boring machine. In these methods, a sensing device comprising sensors such as accelerometers and gyroscopes is attached to the head of the boring machine, which communicates in real time with a steering unit. The steering unit sends steering signals to the head in response of signals from the sensing device. Due to the drifts of the sensors, it is necessary to reset the sensing device at given time intervals to reestablish a proper heading or orientation of the head. In U.S. Pat. No. 2002/0005297, this resetting is performed by detecting the location of the sensing device by means of a tracker unit. In EP-A-534338, the resetting is performed by detaching the sensing device from the head and moving it to a reference location with known coordinates. There is no teaching in EP-A-534338 nor US-A-2002/0005297 of how accuracy of the sensor signals can be improved without having to reset the sensing device.

A method for determining geographical data of a trajectory is further known from NL-C-1017128, which describes a method for measuring a borehole. The borehole extends between a first and a second location, from which DGPS (Differential Global Positioning System) coordinates are measured. A sensing unit comprising an optical gyroscope is moved from the first to the second location for measuring the borehole. The optical gyroscope performs measurements on the first location, a plurality of intermediate locations and at the second location. A drift which occurs in the calculated track is compensated by determining the difference between the measured DGPS coordinates of the second location and gyroscope measurement of the second location and proportionally correcting each of the measurements on the intermediate locations. Such a proportional correction of measurements is however only possible without adversely affecting the accuracy of the data if the trajectory or borehole is substantially straight or has a substantially continuous bend. In case of an arbitrary trajectory, such proportional correction is insufficient for obtaining data of sufficient accuracy.

There is thus a need for a method for determining geographical data of a trajectory, by means of which the accuracy of the measurement data can be improved irrespective of the shape of the trajectory.

It is therefore an aim of the present invention to provide a method for determining geographical data of a trajectory by means of which the accuracy of data obtained from a data collecting device which is moved along the trajectory can be improved.

This aim is achieved according to the invention in that the method comprises the steps of the characterising part of

claim

1.

In the following, the term “trajectory” is used to refer to the physical path which is travelled by the data collecting device during its measurement and the term “track” is used to refer to the calculated path which is obtained from the measurements of the data collecting device.

According to the invention, the data collecting device which is used for measuring data relating to the trajectory is provided with a set of sensors, enabling the sample wise collection of physical measurement data during the movement of the device and enabling the determination of a heading profile, a pitch profile, a distance profile and a roll profile from the collected measurement data. During the movement of the device from the start location to the end location of the trajectory, the measurement data are stored, preferably in a memory of the device, for evaluation later on. Before the evaluation of the collected measurement data for determining the profiles and the track from the profiles, an error compensation is applied to measurement data of at least a first and a second sensor of the device. In other words, in the method of the invention an error compensation is already applied on the level of sensor data, before any of this data is used for calculating profiles for roll, pitch, distance or heading. As a result, the accuracy of the track which is finally obtained from the compensated measurement data can be improved.

The error compensation which is applied on the level of the measurement data according to the invention is determined by correlating the measurement data of at least the first and second sensors with each other. This is performed by determining trajectory parameters by using each time the data of each sensor and mapping the trajectory parameters obtained from the data of the first sensor on the trajectory parameters obtained from the data of the second sensor. This means that measurement data of the first and second sensors are converted to trajectory parameters for enabling a comparison between the two, so that deviations in the measurement data of the first sensor can be evaluated by means of the trajectory parameters of the second sensor and vice versa.

For determining the trajectory parameters of the sensors, measurement data of previous measurements can be taken into account. These previous measurements may for example comprise previous runs of the device on the same trajectory, previous runs on a different trajectory or runs of the device on a predetermined trajectory of which the path is known. In case of first use of the data collecting device or in case no trajectory parameters are available yet, a raw track is calculated from the uncompensated measurement data and trajectory parameters are obtained by correlating the measurement data of the sensors with each other and the raw track. These trajectory parameters can then be fine-tuned by means of measurement data from subsequent runs, so that a self-learning or self-improving system can be obtained.

An example of measurement data level error compensation which can be used in the method of the invention is correlating the measurement data of a first sensor for measuring the gravitational force in a vertical plane of the device and a second sensor for measuring the angle variation around the longitudinal direction of the device. In this case, a roll position of the data collecting device is obtainable from both the first and the second sensor, more particularly by integration over the measurement data of the second sensor and directly from the measurement data of the first sensor. The integrated measurement data of the first sensor constitute first trajectory parameters for the roll position and the measurement data of the second sensor constitute second trajectory parameters for the roll position. In previous measurements, it has been determined how the first and second trajectory parameters are to be interpreted in relation to each other for detecting deviations for both sensors. These deviations in the trajectory parameters are then converted back to data deviations, which form the error compensation to be applied on the measurement data of each sensor. This error compensation is then applied to the measurement data of both sensors, before the roll profile of the data collecting device is determined on the basis of the corrected measurement data of both sensors.

Due to the error compensation on the level of measurement data, the accuracy of the pitch, heading, roll and distance profiles which are determined on the basis of the corrected measurement data can be enhanced. Consequently, by using the method of the invention for evaluating the measurement data, the data collecting device can be allowed to move independently along the trajectory, without needing additional tracking devices for pinpointing the position of the data collecting device at intermediate locations of the trajectory. Furthermore, the need for resetting the sensors at a given time during the movement of the device along the trajectory can be obviated by the method of the invention, so that control electronics for resetting sensors can be omitted in the data collecting device. Due to the possibility of moving the data collecting device independently along the trajectory, the method of the invention can be applied for obtaining geographical data of a variety of trajectories, such as for example utility ducts, onshore and offshore pipelines or the like, a track for car racing or the like, vertical or horizontal boreholes or other trajectories.

The method of the invention preferably comprises the initial step of pre-calibrating the data collecting device. This pre-calibration comprises the steps of moving the data collecting device along a predetermined trajectory in a controlled environment and mapping the measurement data collected during the movement along the predetermined trajectory on predetermined measurement data for this trajectory. From the predetermined trajectory, the track and as a result the data which should be measured by the data collecting device is known, so that by mapping the measured data and the predetermined data, an initial error compensation can be determined. This initial error compensation can comprise the above mentioned error compensation on the basis of the trajectory parameters or can form an addition to this error compensation.

Preferably, in the method of the invention at least one of the heading, pitch, roll and distance profiles is determined by fusing the measurement data of a group of sensors. This group is chosen in such a way among the set of sensors of the device that a first portion of the measurement data of said group is complementary to a second portion of the measurement data of said group. For example, a sensor group for determining the roll profile may comprise the above mentioned first sensor for measuring the gravitational force in a vertical plane of the device and a second sensor for measuring the angle variation around the longitudinal direction of the device. The measurement data of the first sensor is complementary to those of the second sensor in that the roll position is obtainable by integration over the second sensor and directly from the first sensor. This shows that, according to the invention, the measurement data of two sensors may be first correlated for error compensation and later fused for determining one of the profiles. However, the sensor groups may also comprise other sensors than those used for determining the error compensation on the level of measurement data.

The step of fusing the measurement data of the sensor group for determining one of the profiles can be performed by selecting between the complementary portions of the measurement data, for example on the basis of performance characteristics of the corresponding sensors, or by combining the complementary portions, for example by weighed average. The manner in which the measurement data is to be fused can be determined on the basis of previous measurement results, for example from the pre-calibration of the device on the predetermined trajectory. Due to the complementary portions, the profile which is obtained from the fused measurement data of the sensor group can be more accurate than a profile which would be obtained from the measurement data of a single sensor. As a result, the accuracy of the pitch, heading, roll and distance profiles may be enhanced, which may in turn lead to a higher overall accuracy of the track which is determined from these profiles.

The method of the invention preferably further comprises the step of resting the data collecting device on the start location, the end location and/or an intermediate location of the trajectory for a predetermined amount of time. This means that for example a warm-up period is applied before moving the device along the trajectory, so that the sensors are allowed to stabilise before the measurement is conducted. This may further enhance the accuracy of the measurement data and the track obtained with the method of the invention.

The method of the invention preferably further comprises the step of compensating measurement data of at least one sensor for systematic error build-up. This systematic error build-up is derived from a difference between the data measured during the resting period at the start and end locations. As the data collecting device lies still during the resting periods, the data remains substantially constant during these periods, so that a systematic error can be determined for the sensor. When this systematic error at the end location differs from that at the start location, this difference is taken into account by compensating the collected data of the sensor proportionally from start to end.

The method of the invention preferably further comprises the step of compensating the data collected by means of at least one sensor for temperature variations, which are measured during the movement by means of a temperature sensor within the data collecting device. This enables removing temperature dependent measurement errors, which may further enhance the accuracy of the measurement data and the track obtained with the method of the invention. The temperature dependency of sensors can be determined by the pre-calibration.

The method of the invention preferably further comprises the steps of determining coordinates, such as for example GPS coordinates, for at least one location of the trajectory, such as for example the start, end and/or an intermediate location, and compensating the heading, pitch and/or roll profiles for these coordinates. This may further enhance the accuracy of the profiles and the track obtained with the method of the invention.

The method of the invention preferably further comprises the steps of measuring heading, pitch and/or roll of the data collecting device at the start and end locations and compensating the calculated heading and pitch profiles for deviations from the measured heading and pitch at start and end. This may further enhance the accuracy of the profiles and the track obtained with the method of the invention.

The accuracy of the obtained data may be further enhanced according to the invention by including the steps of compensating the calculated track for trajectory constraints and/or by calculating the track twice per movement of the measurement device along the trajectory, namely forwards from start to end and backwards from end to start and then combining the two tracks into an average track.

In the following, the invention will be further elucidated by means of the following description and the appended figures. [0024]
FIG. 1 shows a schematic representation of the method for obtaining geographical data of a trajectory according the invention. [0025]
FIGS. 2[0026] a and 2 b show a flow chart of a preferred method for measuring a trajectory by means of a data collecting device FIGS. 3a and 3 b show a flow chart of a preferred algorithm for evaluating the data obtained with the method of FIGS. 2a and 2 b and building a track from these data.
FIG. 4 shows a flow chart of a sub-algorithm of the algorithm of FIGS. 3[0027] a and 3 b for compensating systematic build-up.
FIG. 5 shows a flow chart of a sub-algorithm of the algorithm of FIGS. 3[0028] a and 3 b for compensating sensor data for temperature variations.
FIG. 6 shows a flow chart of a sub-algorithm of the algorithm of FIGS. 3[0029] a and 3 b for compensating sensor data based on cross dependencies.
FIG. 7 shows a flow chart of a sub-algorithm of the algorithm of FIGS. 3[0030] a and 3 b for calculating an initial roll profile.
FIG. 8 shows a flow chart of a sub-algorithm of the algorithm of FIGS. 3[0031] a and 3 b for compensating the roll profile for start and end boundary conditions.
FIG. 9 shows a flow chart of a sub-algorithm of the algorithm of FIGS. 3[0032] a and 3 b for compensating an initial distance profile for waypoint boundary conditions.
FIG. 10 shows a flow chart of a sub-algorithm of the algorithm of FIGS. 3[0033] a and 3 b for calculating initial heading and pitch profiles.
FIG. 11 shows a flow chart of a sub-algorithm of the algorithm of FIGS. 3[0034] a and 3 b for implementing trajectory constraints.
In the scheme shown in FIG. 1, it is shown how the measurement data, i.e. the actual samples of the sensors included in the data collecting device, are evaluated for obtaining the final track with the method of the invention. More particularly, the scheme shown in FIG. 1 shows that the method of the invention involves compensation on three levels, namely on the level of measurement data, on the level of profiles determined from the measurement data and on the level of the track determined from the profiles. Due to this compensation on three different levels, an accuracy of for example 0.05% of distance can be obtained. [0035]
The compensation on the level of measurement data comprises error compensation which is determined from a systematic error analysis. More particularly, this error compensation comprises a correction on the basis of mapping trajectory parameters, a correction for systematic error build-up and a correction for temperature variations. [0036]
The compensation on the level of pitch, heading, roll and distance profiles comprises the fusing of measurement data of selected sensor groups A, B, C, . . . n and a compensation for boundary conditions associated with each of the groups. For example, one sensor group may comprise a differentation-based sensor and an integration-based sensor or, in other words one sensor which delivers more reliable samples in a stable portion of the trajectory (e.g. a long straight) and another sensor which delivers more reliable samples in an unstable portion (e.g. a bend). As a result, more data is obtained than strictly needed or, in other words, the measurement data of each group comprises complementary portions. This complementary data is then processed intelligently with mathematics and/or algorithms for obtaining a profile of higher accuracy. [0037]
The compensation on track level comprises a correction for determined coordinates of specified locations of the trajectory and possibly also a correction for trajectory constraints. [0038]
The data collecting device which is used for measuring the trajectory comprises a number of sensors for sample wise collecting different physical measurement values during the movement, such as for example heading, roll, acceleration, speed, temperature, gravity or other, so that data on the trajectory as well as data on the measurement conditions is obtained. Examples of sensors which are incorporated in the data collecting device are gyroscopes, accelerometers, magnetometers and a thermometer. During the movement of the device through the duct, the samples of the various sensors are accumulated in a memory of the data collecting device for evaluation later on. Alternatively, the device may also be equipped with wired or wireless means for communicating the samples to an external storage device. [0039]
In the data collecting device, the sensors are preferably associated with electronics for reducing offset errors and increasing repeatability characteristics. The latter means that it is desirable that, if a sensor shows a given offset in its samples, this offset should remain between narrow boundaries for a high number of samples. [0040]
A preferred embodiment of the data collecting device, which is extremely suitable for measuring a duct, comprises the following sensors and parts: [0041]
a 2-axis mechanical rotating gyro for the measurement of the heading and pitch angles (Wy, Wz), with built-in-thermometer [0042]
a vibrating structure gyro for the dynamic measurement of the roll angle (Wx), [0043]
at least four accelerometers for the measurement of xyz-displacements of the device, one (Ax) directed along the longitudinal or x-axis of the device and the three others (Ayz000, Ayz120 and Ayz240) mounted in star configuration in the yz-plane, substantially perpendicular to the x-axis, [0044]
a sensor for the measurement of high frequency beacons (high frequency coils), for example placed at the entry, exit and optionally at intermediate locations of the duct, [0045]
magnetometers for measuring the earth magnetism along the axes of the device (Mx, My and Mz), [0046]
two thermometers for temperature measurement (T), one (T1) attached to the mechanical rotating gyro and one (T2) attached to the displacement accelerometers, [0047]
an odometer or other device for measuring the distance travelled with respect to the starting point, [0048]
a memory for storing the measured samples, [0049]
a rechargeable battery for powering the device, [0050]
the device is preferably able to operate independently in ducts of DN40 (32,6 mm internal), or narrower or wider, [0051]
The preferred embodiment of the method of the invention described here, of which the trajectory measurement steps are shown in FIGS. 2[0052] a and 2 b and the data evaluation steps are shown in FIGS. 3a and 3 b, further comprises the initial step of pre-calibrating the data collecting device for obtaining information on how the sensors of the device react to different circumstances and different movements. This pre-calibration may for example comprise a series of laboratory tests such as for example moving the device along a known trajectory, subjecting the device to given accelerations, testing reactions of the device to temperature or changes in temperature, or other tests. From the pre-calibration results, it can be determined how the dependency of sensor data of a given sensor on different circumstances/movements shows up in the sensor data of another sensor. An overview of dependencies of sensor data on various circumstances/movements and the activators of the circumstances/movements are given in table 1. From this pre-calibration, an error compensation can be determined for at least some of the sensors. This error compensation comprises a correction for a cross dependency between sensors and a correction for temperature variations and possibly other corrections. The information which is obtained from the pre-calibration, i.e. the performance characteristics of sensors under different circumstances and in relation to the movements of the data collecting device during the measurement of a trajectory, is used for implementing compensation algorithms in the evaluation method of FIGS. 3a and 3 b.
The cross dependency of two sensors is determined by converting the measurement data of the one and the other sensor, collected during the pre-calibration tests, into a comparable format, which is herein referred to as trajectory parameters. The error compensation for cross dependency of the sensors is determined by mapping the trajectory parameters of the sensors on each other, so that deviations in the measurement data of the one sensor can be shown by means of the other sensor and vice versa. From this mapping, an error compensation is determined for both sensors, which is later on applied in the evaluation of measurement data for building the track of an unknown trajectory (see FIG. 6). [0053]
In the following, the different steps of the preferred method of FIGS. 2[0054] a and 2 b for measuring a trajectory by means of the data collecting device are described in detail.
First, the data collecting device is switched on and placed on the start location of the trajectory to be measured. The device is allowed to rest for a warming up period of for example 30 seconds at the starting point, so that the sensors of the device are allowed to stabilise before measurement of the trajectory is started with. The samples measured during this warming up period are used for systematic error build-up compensation (see FIG. 4). [0055]
Next, the coordinates, heading and pitch of the start location of the trajectory are measured. This can either be performed by means of the data collecting device, if the appropriate sensors are provided in the device, or by means of external measurement devices. The accuracy of these measurements will correspondingly impact the accuracy of the calculated track. [0056]
The data collecting device is then moved along the trajectory, i.e. through the duct to be measured in this example. The data collecting device is preferably moved through the duct by air propulsion, driven by a compressor, at a speed which is held as stable as possible. The movement of the device can be achieved by means which are part of the device, such as for example a battery operated electric motor or other. The movement can also be achieved by external means, such as for example gas propulsion, liquid floating, cable pulling or other. The external moving means are preferred for mapping longer ducts. [0057]
At the end location of the trajectory, the data collecting device is again allowed to rest for a period of for example 30 seconds, so that further samples are obtained by means of which compensation for systematic error build-up can be performed (see FIG. 4). The coordinates, heading and pitch of the end location are measured as well. Then, measurement of the data collecting device is stopped. [0058]
If applicable, coordinates may also be measured at waypoints or intermediate locations of the trajectory. At these waypoints, high frequency coils or other position markers may be placed so that passage of the device can be detected by means of a sensor. [0059]
Next, the accumulated samples are downloaded from the memory of the device to an on-site evaluation system, which may for example be formed by a computer with appropriate software. The sensor data, i.e. the accumulated samples, is validated for consistency, temperature range, measurement range and saturation, and it is established whether an additional run of the data collecting device along the trajectory is needed for enhancing the accuracy. Finally, the data file comprising the sensor data and the data from heading, pitch and coordinate measurements at start, end and intermediate points is transmitted to a central processing unit. [0060]
For improving the accuracy of the measurement by means of the method shown in FIGS. 2[0061] a and 2 b, it is preferred that the following measures are taken:
the data collecting device is allowed to stabilise its temperature, which is verified by means of uploading 5 minutes of data; [0062]
some daily calibration tests are performed at the starting location. [0063]
In the following, the different steps of the preferred method of FIGS. 3[0064] a and 3 b for evaluating the data file in the central processing unit are described in detail.
First the data file, which comprises the sensor data or samples and the data from heading, pitch and coordinate measurements, is imported and checked a second time for validity. Before any of the sensor data is combined or fused for obtaining a profile for heading, pitch, roll and/or distance, a number of compensation steps are applied to the sensor data. More particularly, the sensor data is compensated for systematic error build-up (as shown in FIG. 4), for temperature variations (as shown in FIG. 5) and for cross dependencies (as shown in FIG. 6). [0065]
The compensation on the level of sensor data for systematic error build-up is shown in the sub-algorithm of FIG. 4. This compensation is based on a difference between the value of the sensor in consideration during the stable period at the start location with respect to its value at the end location. This difference or built-up systematic error is compensated by spreading it out proportionally over the entire sensor data during the run along the trajectory. [0066]
The compensation on the level of sensor data for temperature variations is shown in the sub-algorithm of FIG. 5. The samples of the sensor in consideration are compensated by correlating them with the data of the corresponding temperature sensor and with the pre-calibration results. [0067]
The compensation on the level of sensor data for cross dependencies is shown in the sub-algorithm of FIG. 6. The samples of the sensor in consideration are compensated by correlating them with the samples of sensors which are cross dependent with this sensor and with the pre-calibration results. [0068]
Next, an initial roll profile is calculated by fusing the sensor data of a first sensor group in the sub-algorithm of FIG. 7. The first sensor group comprises the Wx and Ayz sensors as basis for calculating the roll profile and the Wy, Wz, Ax and distance sensors for making a selection between the Wx and Ayz sensors. The Ayz sensor data is first corrected on the basis of its triangular configuration, which implies that the sum of the three accelerometers is necessarily 0. The decision of whether to use the Wx or the Ayz sensor for the respective location of the trajectory is made using the equation: [0069]
VxSqrt(Wy ² ×Wz ²)<trigger value
wherein Vx (speed) is obtained from the odometer or other speed measuring device. The trigger value is obtained from pre-calibration results. [0070]
Next, the sensor data of the first sensor group is compensated by comparing the initial roll profile with start and end boundary conditions for the roll profile, by means of the sub-algorithm of FIG. 8. The start and end roll positions are calculated with the Ayz triangle sensor data. A first correction is applied forwards over the entire roll profile, i.e. from start to end, in such a way that a first roll profile is obtained which meets the end roll position of the Ayz sensor. A second correction is applied backwards over the entire roll profile, i.e. from start to end, in such a way that a second roll profile is obtained which meets the start roll position of the Ayz sensor. For both corrections, more weight is given to corrections on measurements in the curves of the trajectory, since the roll position accuracy is more affected in curves than in straights. Preferably, the correction factor curves/straights equals 6/1. Finally, the first and second roll profiles are combined by means of weighed average. [0071]
Next, an initial distance profile is calculated by fusing the sensor data of a second sensor group, which is formed by the Ax sensor and the distance sensor. The initial distance profile is preferably calculated by double integration of the Ax samples and subsequently combining the twice integrated Ax with the distance samples by means of weighed average. [0072]
The sensor data of the second sensor group is then compensated by comparing the initial distance profile with start, end and waypoint boundary conditions for the distance profile, by means of the sub-algorithm of FIG. 9. This compensation is performed by calculating the distance between the start, waypoints and end and correcting the sensor data in such a way that the distance profile meets the waypoints. [0073]
Next, initial pitch and heading profiles are calculated by fusing the sensor data of a third sensor group in the sub-algorithm of FIG. 10. The third sensor group comprises the Wy and Wz sensor (2-axis mechanical rotating gyro) and the My and Mz sensors (magnetometers) as basis for calculating the pitch and heading profiles and the Ayz (triangle accelerometer), Ax (x-axis accelerometer) and distance sensors for making a selection between the WyWz and MyMz sensor data for calculating the pitch and heading profiles. The Ayz samples are first corrected through triangular configuration, as mentioned above with respect to the sub-algorithm of FIG. 7. The selection is based on the equation: [0074]
Ayz variations<trigger value,
with the trigger value being derived from pre-calibration results. [0075]
Next, the sensor data of the third sensor group is compensated by comparing the initial pitch and heading profiles with start and end boundary conditions, namely the measured pitch and heading at the start and end locations of the trajectory. [0076]
After these steps of calculating initial profiles and compensating sensor data by comparison of the initial profiles with boundary conditions, the roll, distance, pitch and heading profiles are recalculated with the compensated sensor data. These new profiles are then used for calculating a first track. This first track is then tested for compliance with the desired accuracy by comparing the calculated coordinates of the end of the calculated track with the measured coordinates of the end location of the trajectory, which can be summarised in the condition: [0077]
Calculated XYZ−Measured XYZ<limit
A suitable limit is for example 10 cm on all axes, but this limit may also be chosen wider or narrower. [0078]
If the first track is within the limit, the track is recalculated by fusing the final roll, distance, pitch and heading profiles from start to end into a forwards track, fusing the final profiles from end to start into a backwards track and combining the forwards and backwards tracks into by weighed average into a final track. The backwards calculated track may lead to a different result since the measurements of the sensors have a two-way influence on each other. This forwards and backwards recalculation leads to a further improvement in the accuracy of the resulting final track, so that the final track is well within the desired accuracy limits. [0079]
On the other hand, if the first track is outside the limit, some further compensation algorithms are applied for compensating the roll, distance, pitch and heading profiles for trajectory constraints and/or geographical information. These further compensation algorithms may comprise one or more of the following: [0080]
an algorithm for correcting the track for deviations of the calculated start, end and/or waypoint coordinates to the measured coordinates of the start location, the end location and/or intermediate locations of the trajectory and applying a proportional correction to track points in between the measured coordinates; [0081]
an algorithm for correcting the track to align with measured coordinates of objects outside the trajectory, such as for example a river or other objects; [0082]
one or more algorithms for correcting the track to physical constraints of the trajectory, such as for example a maximum or standard curvature (e.g. 30°, 45°, . . . ) of bent portions of the trajectory, a minimum length of straight portions of the trajectory, or in the case of a drilling pipe, the distance between welds of pipe portions having a standard length, or other physical constraints. [0083]

An example of an algorithm for correcting the track to the physical constraint of a maximum curvature of a bent portion of the trajectory is shown in FIG. 11. After recalculation of the pitch, heading and roll profiles, bends in the track are detected by zooming in on angle variations of for example 10° in a time period of 5 seconds. For the bends, the condition is applied whether the bending radius is below or above the maximum bending radius of the physical trajectory of for example 0.8 m. If this condition is fulfilled, the bend is skipped and the next bend is considered. If this condition is not fulfilled, the bend of the track is corrected by either adjusting the roll profile over the bend or by limiting the bending radius of the bend until the physical requirement is fulfilled.

TABLE 1


sensor dependencies

SENSOR	DEPENDENCY	ACTIVATOR

Wx	Temperature	Temperature.
	Earth rotation	Earth rotation as a function of heading
		and pitch.
	Direct acceleration	From acceleration of data collecting
		device in the x-axis and centrifugal
		force from Wy and Wz.
	Cross acceleration in all yz directions	From centrifugal force on data collecting
		device in heading and pitch
		directions, and from acceleration of the
		Wy and Wz rate.
	Misalignment with the x-	Misalignment.
	axis of the data collecting
	device
Wy	Temperature	Temperature.
	Roll	From gravitation field.
	Pitch	From gravitation field.
	Earth rotation	Earth rotation as a function of heading
		an pitch.
	Direct acceleration in the	From centrifugal force on the data
	y axis	collecting device in heading and pitch
		directions, and from centrifugal force
		due to Wx
	Cross acceleration in all	From centrifugal force on the data
	zx directions	collecting device in heading and pitch
		directions, and from acceleration in the
		x-axis.
	Misalignment with the	Misalignment.
	Ayz000 direction
	Non-orthogonality with the	Misalignment.
	x-axis
Wz	Temperature	Temperature.
	Roll	From gravitation field.
	Pitch	From gravitation field.
	Earth rotation	Earth rotation as a function of heading
		an pitch.
	Direct acceleration in the	From centrifugal force on the data
	z axis	collecting device in heading and pitch
		directions, and from centrifugal force
		due to Wx
	Cross acceleration in all	From centrifugal force on the data
	yx directions	collecting device in heading and pitch
		directions, and from acceleration in the
		x-axis.
	Non-orthogonality with the	Misalignment.
	y-axis
	Non-orthogonality with the	Misalignment.
	x-axis
Ax	Temperature	Temperature.
	Roll	From gravitation field.
	Pitch	From gravitation field.
	Direct acceleration in the	From centrifugal force due to Wy and
	x-axis	Wz.
	Cross acceleration in all	From centrifugal force on data
	yz directions	collecting device in heading and pitch
		directions, and from acceleration of the
		Wy and Wz rate.
	Misalignment with with the	Misalignment
	x-axis
Ayz000	Temperature	Temperature.
Ayz120	Direct acceleration in the	From Wx rotation, and from centrifugal
Ayz240	000, 120 or 240 axis	force in heading- and pitch direction.
	Cross acceleration in all	From centrifugal force on the data
	directions: perpendicular	collecting device in heading and pitch
	to 000, 120 or 240 axis	directions, and from acceleration in the
		x-axis.
	Non-120° -between axes	Misalignment.
	Non-orthogonality with the	Misalignment.
	x-axis

Claims

1. A method for determining a track of a geographical trajectory which has a predetermined start location and a predetermined end location, said method comprising the steps of:

a) moving a data collecting device from the start location to the end location, the device comprising a set of sensors providing for sample wise collection physical measurement data during the movement of the device and enabling the determination of a heading profile, a pitch profile, a distance profile and a roll profile from said measurement data,

b) determining said heading profile, said pitch profile, said distance profile and said roll profile based on the collected measurement data,

c) determining said track based on said heading profile, said pitch profile, said distance profile and said roll profile,

characterised in that all said physical measurement data are stored during the movement of the device from the start location to the end location and that, after retrieval of the stored measurement data and before determining said profiles, the method further comprises the steps of:

d) determining at least first, respectively second trajectory parameters based on measurement data of a first, respectively second sensor of said set and mapping said first trajectory parameters on said second trajectory parameters,

e) determining an error compensation for said measurement data of said first and said second sensors based on said mapping,

f) correcting said measurement data of said first and second sensors by means of said error compensation.

2. The method of claim 1, characterised in that the method further comprises the initial steps of pre-calibrating the data collecting device for determining an initial error compensation to be applied in step f), by moving the data collecting device along a predetermined trajectory in a controlled environment and mapping the measurement data collected during the movement along the predetermined trajectory on predetermined measurement data for the predetermined trajectory.

3. The method of claim 1 or 2, characterised in that at least one of said heading, pitch, roll and distance profiles is determined by fusing the measurement data collected by means of a group of sensors, said group being chosen in such a way among said set of sensors that a first portion of the measurement data of said group is complementary to a second portion of the measurement data of said group.

4. The method of claim 3, characterised in that said complementary first and second portions of the measurement data are fused by means of selection based on performance characteristics of the sensors included in said group and/or by means of combination.

5. The method of any one of the claims 1-4, characterised in that the method further comprises the step of resting the data collecting device on the start location, the end location and/or one or more predetermined intermediate locations of the trajectory for a predetermined amount of time.

6. The method of claim 5, characterised in that the method further comprises the step of compensating the measurement data of at least one sensor for systematic error build-up, based on a difference between the measurement data of said sensor collected during the resting period at the start location and the measurement data of said sensor collected during the resting period at the end location.

7. The method of any one of the claims 1-6, characterised in that the data collecting device comprises at least one temperature sensor and that the method further comprises the step of compensating the measurement data of at least one other sensor for temperature variations.

8. The method of any one of the claims 1-7, characterised in that the method further comprises the steps of determining coordinates on at least one location of the trajectory and compensating the heading profile and/or the pitch profile and/or the roll profile and/or the distance profile for the coordinates of the at least one location.

9. The method of claim 8, characterised in that coordinates are determined on the start location, the end location and/or one or more predetermined intermediate locations.

10. The method of any one of the claims 1-9, characterised in that the method further comprises the steps of measuring heading, pitch and/or roll of the data collecting device at the start location and/or the end location of the trajectory and compensating the heading profile and/or the pitch profile and/or the roll profile and/or the distance profile for deviations from the measured heading and/or pitch at the start and end locations.

11. The method of any one of the previous claims, characterised in that the method further comprises the step of compensating said track determined in step c) for trajectory constraints and/or geographical information.

12. The method of claim 11, characterised in that the trajectory constraints and/or geographical information comprise at least one of the following: coordinates of the start location, the end location and/or one or more intermediate locations of the trajectory, coordinates of objects outside the trajectory, a maximum curvature of bent portions of the trajectory, a minimum length of straight portions of the trajectory.

13. The method of any one of the claims 1-12, characterised in that step c) comprises determining a forwards track based on said heading profile, said pitch profile, said distance profile and said roll profile from start to end, determining a backwards track based on said heading profile, said pitch profile, said distance profile and said roll profile from end to start, and combining the forwards and backwards tracks into an average track.

14. The method of any one of the claims 1-13, characterised in that steps a) to f) are repeated at least once, a track being determined from the measurement data collected during each move of the data collecting device, and that the method further comprises the step of combining said tracks into a final track.