DE102012210466A1 - Method for operating navigation device mounted in e.g. motor car, involves determining velocity signal related to speed of vehicle along longitudinal axis based on acceleration signals and/or link signal for a specific time - Google Patents

Abstract

The method involves generating a position signal using satellite-based measurement principle. The observer parameters of an observer are determined based on the position signal and three acceleration signals respectively output from acceleration sensors. The training tuples are generated with valid available position signal. A velocity signal related to speed of vehicle (1) along longitudinal axis (Y) is determined based on acceleration signals and/or a link signal for a specific time. The distance traveled by the vehicle is determined based on acceleration signals. An independent claim is included for navigation device.

Description

The invention relates to a method for operating a navigation device and a navigation device.

Navigation devices are regularly arranged in motor vehicles. Navigation devices may be formed, for example, in a mobile terminal, such as a smartphone. In these cases, they are regularly fixed in the vehicle by means of a holder. In connection with such a navigation device, a position determination unit is used which generates a position signal by means of a satellite-based measurement principle. In this way, then a route guidance along a predetermined route. However, in large cities with tall buildings, for example, the position signal can not be validly provided. This is also the case, for example, when traveling in tunnels or parking garages. However, it is desirable that a precise position determination is possible even in such places and so a comfortable route guidance can be done.

If the navigation device is firmly integrated in the vehicle, for example in the dashboard, access to further vehicle sensors is regularly possible and thus access to inertial sensor data from the vehicle electrical system. In this way, measurement signals can be supplied which are representative of an odometry and / or a steering angle and / or an acceleration. In this way, in such integrated navigation devices, even in the absence of the position signal over long distances, an accurate position determination can be carried out.

If the navigation device is embodied in a mobile terminal, it is preferably fastened in the vehicle in a holding device in the front region of the vehicle. The holding device is attached, for example, to a windshield. Without a valid position signal, the position of the navigation device in the cases where no valid position signal is available, can only be followed inaccurately.

The object on which the invention is based is to provide a method for operating a navigation device and a navigation device, which contribute to a comfortable operation of the navigation device.

The object is solved by the features of the independent claims. Advantageous embodiments are characterized in the subclaims.

The invention is characterized by a method for operating a navigation device and by the navigation device designed to carry out the method. The navigation device is arranged as intended in a reference device. The reference device may for example be a vehicle, in particular a motor vehicle. The navigation device has three acceleration sensors which detect accelerations in mutually orthogonal directions and generate respective first, second and third acceleration signals. Furthermore, a position determination unit is provided which generates a position signal by means of a satellite-based measurement principle. The acceleration sensors and the position determination unit or even a subset of these can be arranged in the navigation device or in principle external to it, for example in a holding device.

A learning of observer parameters of an observer is dependent on training tuples of the first to third acceleration signals and the position signal. The training tuples are generated when the position signal is valid and in particular temporally correlated with each other.

A speed signal for a speed of the reference device along its longitudinal axis is determined by the observer depending on at least one of the first to third acceleration signals. Alternatively or additionally, a path signal for a distance traveled within a predetermined time along the longitudinal axis of the reference device is determined by means of the observer as a function of at least one of the first to third acceleration signals.

In this way, the inertial sensor system, which is often present anyway in the navigation apparatus or its associated with the acceleration sensors, can be used simply to determine the current position of the reference device, using the path signal and / or the speed signal. In this connection, further corresponding gyrometer signals of a gyrometer can be used, so as to reconstruct a movement trajectory of the reference device. Thus, in particular, independently of the presence of a favorable position signal, the speed or traveled distance of the reference device can be determined precisely.

Rotational movements, such as turning operations, of the reference device may occur in This connection can be reconstructed particularly simply and precisely by means of the gyroscope signals.

By learning the observer parameters by means of the training tuples, the path signals and the speed signals can be precisely determined in spite of corresponding disturbances of the acceleration signals by accelerations of the holding device, which moves regularly relative relatively swinging to the reference device.

The position signal thus serves as a reference during the learning of the observer parameters and thus contributes to the adjustment of the observer, that is to say in particular to his observer parameters with regard to a precise determination of the path signal or the speed signal.

According to an advantageous embodiment, the training tuples comprise at least one gyroscope signal which is generated by a gyroscope assigned to the navigation device. In this way, a quality of the observer can be increased and thus the path signal or the speed signal can be determined particularly precisely.

According to a further advantageous embodiment, the training tuples comprise a compass signal that is generated by a compass associated with the navigation device. In this way, a quality of the observer can be increased and thus the path signal or the speed signal can be determined particularly precisely. In this context, it is advantageous if the compass is a three-axis magnetometer.

According to a further advantageous embodiment, the training tuples comprise a barometer signal which is generated by a barometer assigned to the navigation device. In this way, the observer parameters can be learned in a particularly favorable manner and then enable a particularly precise determination of the path signal or of the speed signal. In this context, the barometric signal or else the compass signal can also be used to determine the speed signal and / or the path signal.

According to a further advantageous embodiment, a position signal is generated as a function of the first to third acceleration signals, which is representative of a spatial orientation of the navigation device with respect to the reference device. Further, the learning of observer parameters of the observer is performed depending on the position signal. The speed signal or the path signal is determined as a function of the position signal. In this way, it can be easily taken into account in which position the navigation device is arranged in the reference device. Thus, for example, the holding device may have a flexible element and thus be different depending on the setting of this flexible element, the position of the flexible element. By means of the position signal, the respective acceleration signals can then be assigned to the respective axes of the reference device, that is to say a vertical axis, a transverse axis and a longitudinal axis of the reference device. This can also be done, for example, by means of a corresponding transformation depending on the position of the navigation device.

According to a further advantageous embodiment, at least one of the first to third acceleration signals is filtered as part of the learning of the observer parameters. In this way, the observer can be learned with his observer parameters with a particularly high quality.

In this context, it is advantageous if, as part of the learning of the observer parameters, at least one of the first to third acceleration signals is filtered by means of a sliding averaging. Furthermore, it is advantageous in an optional embodiment if, as part of the learning of the observer parameters, at least one of the first to third acceleration signals is filtered by means of frequency filtering. In this context, for example, a Fourier transform can also be used.

Furthermore, according to an optional embodiment, it is advantageous if, during the learning of the observer parameters, at least one of the first to third acceleration signals is filtered by means of an IIR filter. According to a further optional embodiment, an FIR filter may be provided in addition to or as an alternative to the IIR filter.

According to a further advantageous embodiment, at least one of the first to third acceleration signals is integrated or doubly integrated within the scope of learning the observer parameter.

According to a further advantageous embodiment, the observer comprises a neural network and / or a support vector machine and / or a decision tree and / or a random forest, which is also referred to as random forrest, and / or a Bayesian approach.

According to a further advantageous embodiment, a position velocity signal for the speed of the reference device along its longitudinal axis is determined depending on the position signal. Further, alternatively or additionally determines a position-distance signal for the distance traveled within the predetermined time along the longitudinal direction of the reference device as a function of the position signal. Depending on a predetermined deviation of the position-velocity signal from the velocity signal or depending on a predetermined deviation of the position-range signal from the path signal, learned observer parameters are initialized in a predetermined manner.

In this way, the knowledge is used that such a changed arrangement of the navigation device can be detected in the reference device and so caused by the predetermined initialization a new learning process can be started and then as soon as possible again the observer is ready for precise determination of the path signal or the speed signal.

Embodiments of the invention are explained in more detail below with reference to the schematic drawings. Show it:

1 a reference device having a navigation device disposed therein,

2 the navigation device,

3 a first flowchart of a first program, which is processed in the navigation device and

4 a flowchart of a second program, which is executed in the navigation device.

Elements of the same construction or function are identified across the figures with the same reference numerals.

A reference device 1 ( 1 ) is designed, for example, as a vehicle, in particular as a motor vehicle. In the reference device 1 is a navigation device 3 by means of a holding device 5 arranged. The holding device 5 For example, it is attached to the windshield, this can be done for example by means of a suction cup. In addition, the holding device 5 be designed and arranged arbitrarily different, so for example be fixed to the dashboard. The navigation device 3 is thus in particular a mobile navigation device, such as a mobile device, such as a smartphone. A longitudinal axis of the reference device 1 is denoted by Y. A transverse axis of the reference device 1 is denoted by X and a vertical axis of the reference device 1 is denoted by Z.

The navigation device 3 comprises a computing unit RE, a data and program memory DPS and interfaces, such as first to third interfaces SST1, SST2, SST3. The arithmetic unit RE and / or the data and program memory DPS can be formed in one unit or even distributed over a plurality of units. The arithmetic unit RE can comprise, for example, a microprocessor and / or a microcontroller.

The first interface SST1 is configured to communicate with an external server OBS. For this purpose, it may include, for example, a mobile radio interface.

The navigation device 3 further comprises the second interface SST2 to an input unit EE, which may be, for example, a keyboard, a joystick or a touch screen. In this way, input of a start position and / or a destination position by a user, who may be a vehicle driver, for example, is possible.

The start position can for example also be provided automatically by means of a position determination unit GPS based on a current position at which the navigation device NAV is located. The position determination unit GPS is designed to generate a position signal POS_S by means of a satellite-based measurement principle. This can be done for example GPS-based.

The navigation device 3 further comprises the third interface SST3 to an optical output unit GD, which is preferably a screen. In addition, in principle, a further interface to an acoustic output unit may be present. Further, the navigation device 3 also first to third acceleration sensors 7 . 9 . 11 associated with detecting in each orthogonal directions accelerations and generate respective first, second and third acceleration signals B1_S, B2_S, B3_S. Further, the navigation device 3 a gyrometer 12 which generates gyro signals GYR_S representative of spin. In the case of three gyrometer signals GYR_S, they are each representative of one rotational acceleration about one each of three axes. Further, the navigation device 3 a compass 13 assigned, which is preferably a three-axis magnetometer. There is also a barometer 15 the navigation device 3 which generates a barometric signal BARO_S. The compass 13 generates a compass signal KOMP_S. The position determination unit GPS, the acceleration sensors 7 . 9 . 11 , the compass 13 and the barometer 15 or a subset of these may be in unit with the navigation device 3 However, they can be arranged also externally to this, for example in the holding device 5 be arranged.

To operate the navigation device 3 are stored in their data and program memory DPS preferred programs that during operation of the navigation device 3 can be processed.

A first program is described below with reference to the flowchart of 3 explained in more detail.

The program is started in a step S1 in which variables can be initialized if necessary.

In a step S3, it is determined whether the position signal POS_S is validly available. This is not the case, for example, if corresponding satellite signals can not be received, as may be the case for example in street canyons, tunnels or car parks or the like.

If the condition of step S3 is not fulfilled, the processing, if appropriate after a predetermined waiting period, is continued again in step S3. On the other hand, if the condition of the step S3 is satisfied, the processing is continued in a step S5.

In step S5, training tuples TUP are generated from first to third acceleration signals B1_S, B2_S, B3_S which correlate with one another in time to the position-correlated POS_S signal and the time-correlated compass signal KOMP_S and the barometric signal BARO_S and the gyroscope signals GYR_S. Thus, a plurality of training tuples are generated in a plurality of passes of step S5. In this case, the multiple passage through step S5 is indicated by the dashed line back to the step S3. The processing of step S7 can be carried out after the generation of a predetermined number of training tuples, but it can also take place virtually parallel to the generation of the tuples.

In step S7, observer parameters B_PARAM of an observer are learned depending on the training tuples TUP. This takes place in the context of a so-called machine learning. In this context, the observer may for example comprise a neural network and / or comprise a support vector machine and / or comprise a decision tree and / or comprise a random forest and / or a Bayesian approach.

In addition, depending on the first to third acceleration signals B1_S, B2_S, B3_S, a position signal LAG_S can be generated which is representative of a spatial orientation of the navigation device 3 with respect to the reference device 1 , The learning of the observation parameters B_PARAM is then also carried out as a function of the position signal LAG_S in the sense that, depending on the position signal LAG_S, an assignment under possibly carrying out a predetermined transformation of the first to third acceleration signals B1_S, B2_S, B3_S to the respective axis, ie in particular the longitudinal axis Y of the reference device 1 , he follows. Furthermore, as part of learning the observer parameter B_PARAM, at least one of the first to third acceleration signals B1_S, B2_S, B3_S can be filtered, for example by means of a moving averaging and / or frequency filtering and / or an IIR filter. In addition, as part of learning the observer parameter B_PARAM, at least one of the first to third acceleration signals B1_S, B2_S, B3_S can be integrated or doubly integrated.

The program is then ended in a step S9. For example, the program is restarted after a predetermined period of time or even after the fulfillment of another condition, which includes, for example, that an initialization flag INIT is set.

A second program is described below with reference to the flowchart of 4 explained in more detail. The second program is started in a step S11 in which variables are initialized if necessary.

In a step S13, a speed signal VS becomes a speed of the reference device 1 along its longitudinal axis Y by means of the observer BEOB determined depending on at least one of the first to third acceleration signals B1_S, B2_S, B3_S, further also optionally the compass signal KOMP_S and / or the barometer BARO_S and / or the Gyrometersignale GYR_S are used for this purpose. In addition, the position signal LAG_S can be used accordingly. In a step S15, the route signal S_S is then determined by means of an analogous procedure to that of the step S13 for the speed signal V_S by means of the observer BEOB in the step S15.

The speed signal V_S and / or the route signal S_S can then be used to determine a trajectory of the navigation device 3 , So a course of the position of the navigation device 3 and so also to determine observer position signals. These can then be used for route guidance, for example.

In subsequent optional steps, a position-speed signal PV_S is first determined in step S17 as a function of the position signal POS_S. In this context, the waveform of the position signal POS_S can be evaluated accordingly, so be differentiated, for example. Depending on the design of the position signal POS_S, the position signal POS_S can also be assigned directly to the position velocity signal PV_S.

In a step S21, the position range signal PS_S is determined as a function of the position signal POS_S.

In a step S23, an initialization flag is detected. This can be done particularly simply by evaluating the first to third acceleration signals B1_S, B2_S, B3_S in the sense of checking whether the arrangement of the navigation device 3 in the reference device 1 was changed.

Alternatively or additionally, in step S23 the initialization flag can be determined depending on the position speed signal PV_S and the speed signal V_S and / or the position-path signal PS_S and the path signal S_S, in particular for a validly available position signal POS_S. In this case, the initialization flag INIT is set, for example, if a predetermined deviation of the position-path signal PS_S to the path signal S_S occurs and / or a predetermined deviation of the position-velocity signal PV_S appears to the speed signal V_S. This is an indication that the arrangement of the navigation device 3 in the reference device 1 has been changed, such as the holding device 5 has been fixed elsewhere, or the navigation device 3 from the holding device 5 was taken out and was used again in another position.

If the initialization flag INIT is set, the observer parameters B_PARAM are initialized in a predetermined manner. In this way, they can, for example, be provided with a neutral value again or reset accordingly to enable a re-learning efficiently.

The processing is then terminated in a step S25. The second program is preferably executed cyclically in order to provide the speed signal V_S and / or the route signal S_S as frequently as possible.

LIST OF REFERENCE NUMBERS

1

reference device

Y

longitudinal axis

X

transverse axis

Z

vertical axis

3

holder

5

navigation device

GPS

Position Determination Entity

RE

computer unit

DPS

Data and program memory

SST1-3

first to third interface

DG

optical output unit

EE

input unit

OBS

external server

7, 9, 11

first, second, third acceleration sensor

13

compass

15

barometer

POS_S

position signal

B_PARAM

observers parameters

OBS

observer

B1_S, B2_S, B3_S

first, second and third acceleration signals

TUP

Trainingstupel

V_S

speed signal

S_S

track signal

KOMP_S

compass signal

BARO_S

barometer signal

LAG_S

position signal

PV_S

Position speed signal

PS_S

Position signal path

INIT

initialization flag

S1-S25

step

Claims (14)

Method for operating a navigation device ( 3 ) when used in a reference device ( 1 ), wherein the navigation device ( 3 ) three acceleration sensors ( 7 . 9 . 11 ) are associated, which detect in respective mutually orthogonal directions accelerations and respective first, second and third Generate acceleration signals (B1_S, B2_S, B3_S) associated with a position determining unit (GPS) which generates a position signal (POS_S) by means of a satellite-based measurement principle, in which observer parameter learning (B_PARAM) of an observer (BEOB) is dependent on training tuples (TUP ) of the first to third acceleration signals (B1_S, B2_S, B3_S) and the position signal (POS_S), wherein the training tuples (TUP) are generated when the position signal (POS_S) is validly available, - in which a speed signal (V_S) for a Speed of the reference device ( 1 ) is determined along its longitudinal axis (Y) by means of the observer (BEOB) as a function of at least one of the first to third acceleration signals (B1_S, B2_S, B3_S) and / or a path signal (S_S) for one within a predetermined time along the longitudinal axis (Y ) of the reference device ( 1 ) distance is determined by the observer (BEOB) depending on at least one of the first to third acceleration signals (B1_S, B2_S, B3_S). Method according to Claim 1, in which the training tuples (TUP) comprise at least one gyroscope signal (GYRO_S) transmitted by one of the navigation devices ( 3 ) associated gyrometer ( 12 ) is produced. Method according to Claim 1 or 2, in which the training tuples (TUP) comprise a compass signal (KOMP_S) which is provided by one of the navigation devices ( 3 ) associated compass ( 13 ) is produced. Method according to Claim 3, in which the compass ( 3 ) is a three-axis magnetometer. Method according to one of the preceding claims, in which the training tuples (TUP) comprise a barometric signal (BARO_S) transmitted by one of the navigation devices ( 3 ) associated barometer ( 15 ) is produced. Method according to one of the preceding claims, in which - depending on the first to third acceleration signals (B1_S, B2_S, B3_S) a position signal (LAG_S) is generated which is representative of a spatial orientation of the navigation device ( 5 ) with respect to the reference device ( 1 ), - observer observer parameter learning (BEOB) is performed as a function of the position signal (LAG_S), and - the speed signal (V_S) or the path signal (S_S) is determined as a function of the position signal (LAG_S). Method according to one of the preceding claims, in which at least one of the first to third acceleration signals (B1_S, B2_S, B3_S) is filtered as part of the learning of the observer parameter (B_PARAM). Method according to Claim 7, in which, as part of the learning of the observer parameter (B_PARAM), at least one of the first to third acceleration signals (B1_S, B2_S, B3_S) is filtered by means of a sliding averaging. Method according to one of claims 7 or 8, wherein in the course of learning the observer parameter (B_PARAM) at least one of the first to third acceleration signals (B1_S, B2_S, B3_S) is filtered by means of frequency filtering. Method according to one of claims 7 to 9, wherein in the course of learning the observer parameter (B_PARAM) at least one of the first to third acceleration signals (B1_S, B2_S, B3_S) is filtered by means of an IIR filter. Method according to one of the preceding claims, in which as part of the learning of the observer parameter (B_PARAM) at least one of the first to third acceleration signals (B1_S, B2_S, B3_S) is integrated or doubly integrated. Method according to one of the preceding claims, in which the observer comprises a neural network and / or a support vector machine and / or a decision tree and / or a random forest and / or a Bayesian approach. Method according to one of the preceding claims, in which a speed of the reference device (PV_S) ( 1 ) is determined along its longitudinal axis (Y) in dependence on the position signal (POS_S) and / or a position range signal (PS_S) for the within the predetermined time along the longitudinal axis (Y) of the reference device ( 1 ) traveled distance is determined depending on the position signal (POS_S) and depending on a predetermined deviation of the position speed signal (PV_S) to the speed signal (V_S) or depending on a predetermined deviation of the position range signal (PS_S) to the path signal (S_S) learned observer parameters (B_PARAM ) can be initialized. Navigation device ( 3 ), which is adapted to perform a method according to any one of claims 1 to 12.

Images