DE102018009114A1 - Method for determining the position of a mobile part movable on a travel surface and installation with mobile part for carrying out the method - Google Patents

Abstract

Method for determining the position of a movable on a moving surface handset and system with handset for performing the method, the handset has at least one laser scanner, in particular for detecting the relative to the handset relative positions of arranged on the travel surface objects, and the handset odometrisches Has system, in particular for determining a traversed by the handset distance.

Description

The invention relates to a method for determining the position of a movable on a travel surface handset and system with handset for performing the method.

It is known to perform laser scans with laser scanners and thus to realize security functions, for example when approaching people.

To determine the position of the GPS system is generally known and in vehicles, the use of odometric systems, which, however, have a lower accuracy than GPS systems.

From US 2012/0 303 176 A1, a method for precise localization for an industrial vehicle is known as the closest prior art.

From US 2007/0 262 884 A1 a method for controlling the density of visual landmarks in a visual simultaneous location system is known.

The invention is therefore based on the object to develop the position determination in a system with handset.

According to the invention the object is achieved in the method according to claim 1 and in the system according to the features indicated in claim 14.

Important features of the invention in the method are that the method is provided for determining the position of a movable on a travel surface, in particular on a travel surface of a system, movable handset,
wherein the mobile part has at least one laser scanner, in particular for detecting the positions of objects arranged on the travel surface relative to the mobile part,
wherein the mobile part has an odometric system, in particular for determining a distance traveled by the mobile part,
in which
In a first method step, the handset is moved along a trajectory on the traversing surface, in particular manually guided or track-guided, and the positions of objects are determined by means of the laser scanner and the odometric system, the positions being assigned to a card and stored in a memory become,
wherein in a second method step the map together with the positions of the associated objects is displayed with a display means, in particular a screen, and regions are identified as static or semidynamic, the unmarked regions being considered as dynamic regions,
wherein in a third method step, the mobile part is moved on the traversing surface, in particular wherein the handset on the traversing surface, in particular for fulfilling a driving order, independently driven navigated, wherein by means of the laser scanner, the positions of objects relative to the handset are detected and to determine the new current Position of the handset can be used
wherein only the objects detected in regions identified as static are used to determine the new current position of the handset and the objects detected in regions identified as semidynamic are used,
wherein for determining the new current position of the mobile part of each virtual position from a set of virtual positions is assigned a respective evaluation parameter,
wherein the particular absolute positions of the detected objects are determined for each of the virtual positions by adding up the positions of the detected objects relative to the mobile part by means of laser scanners to the respective virtual position and then determining the evaluation parameter,
wherein one of the virtual positions of the set is an odometrically determined position,
wherein as a new current position that virtual position is determined, which is assigned the smallest evaluation parameter.

It is therefore important that to determine the respectively new current position of the handset, a table is created, which assigns a plurality of possible, in particular virtual, positions in each case an evaluation parameter,
wherein one of the possible positions of the set is an odometrically determined position, in particular the currently odometrically determined position,
wherein as the new current position of the mobile part that one of the possible positions is determined, which is assigned an extremal evaluation parameter, in particular the smallest one of these evaluation parameters, in particular to the table,
wherein for each of the possible positions the particular absolute positions of the detected objects are determined by the laser scanner relative to the mobile part related positions of the detected objects are added to the respective possible position in particular vectorially and then the respectively assigned evaluation parameter is determined.

The advantage here is that a fast and accurate position determination is possible. Because the odometric determination is more inaccurate than that Determination by comparison of the detected objects with the objects stored in the map. Thus, the odometrically determined position is the starting position for comparing the detection with the map. Subsequent comparisons are then executable starting from positions adjacent to the starting position.

It is also advantageous that the card also has positions of non-static objects. Thus, therefore, a position determination is carried out using objects whose existence is not foreseeably certain. In the map, however, the positions of such objects are taken into account.

In an advantageous embodiment, in the first method step, the positions of the objects are determined by adding the position relative to the handset by means of the laser scanner relative to the position determined by the odometer of the mobile part. The advantage here is that when commissioning a first detection is performed, so that the card can be created. Thus, then the regions are markable below.

In an advantageous embodiment, the handset on a controller, an energy storage and controllable by the controller, can be supplied from the energy storage drive so that the handset moves on the travel surface, in particular so that the handset is moved autonomously on the travel surface. The advantage here is that by means of the card navigation is possible in an autonomous manner, as by means of the odometric system approximate position data can be determined and the precise determination of the position by evaluating the laser scanning is possible.

In an advantageous embodiment, the sum of the squares of the distances of the detected objects to the respective objects corresponding to the map is determined to check the correspondence with the map. The advantage here is that an easily determinable weighting factor is used.

In an advantageous embodiment, the time profile of the steering angle of the mobile part and the time profile of the speed of the mobile part is detected by means of the odometric system,
in particular, wherein the rotational speed of a drive wheel is detected and from the respective current rotational speed, taking into account the diameter of the drive wheel, the speed is determined and wherein the steering angle of the mobile part is detected. The advantage here is that an odometric determination of the position is possible, the anyway existing steering unit and the anyway existing drive wheel can be used and only a Winkelstellungs- and / or speed detection is necessary.

In an advantageous embodiment, the evaluation parameter is the sum of the squares of the distances between the respective position of a respective detected object and the position of a nearest object stored in the map, if these objects are arranged in a region characterized as static or semidynamic. The advantage here is that semidynamic objects can also be used to determine the position if these objects are present. Instead of squares, the amounts are usable, although then less computational effort is necessary, but the extremum is less pronounced.

In an advantageous embodiment, it is checked in a fourth method step whether, among the objects detected in a laser scan, one of the objects stored in the static region map has not been detected, and depending on the result of this check, a warning or error message is displayed or further reported. The advantage here is that in the map objects are drawn in static regions. If one of the static objects is not detected during a laser scan in a direction covered by the laser scanner, an error is reported. For a static object can not be detected again only if it were outside the detection range of the laser scan or if another object is arranged between the static object and the laser scanner, for example a dynamic object. Otherwise, the static object must therefore be detected. If it was not detected, although it would have to be located in the detection area according to the map.

In an advantageous embodiment, the card is designed as a grid of area elements, in particular a regular grid of area elements,
in particular so that the value range of the coordinates of the positions is discrete. The advantage here is that a low resolution in the detection of the objects is necessary. For example, a resolution in the detection of the objects, which is between 10% and 30% of the maximum diameter of the surface elements is sufficient.

In an advantageous embodiment, the card and the positions assigned to the card are stored in a volatile memory,
whereby the respective current map and the positions assigned to the map are stored in a non-volatile memory in a time-recurring manner,
wherein after failure of the power supply of the navigation device with volatile memory stored on the non-volatile memory last information about the card and the map associated positions are stored in a non-volatile memory. The advantage here is that at Power failure the recorded data remain as possible and are recoverable from the non-volatile memory, that are loadable in the volatile memory.

In an advantageous embodiment, in the second method step, the regions are identified only as semidynamic, the unmarked regions being considered dynamic regions,
wherein in the third method step, the regions occupied by objects, which are detected unchanged after a predetermined number of laser scans and / or after a predetermined period of time, are identified as static. The advantage here is that semidynamic regions are present and thus objects are used when they are detected, but disappear despite the deposit of their position in a map again. The map also has positions of non-static objects.

Important features of the system with mobile part for carrying out an aforementioned method are that the mobile part has at least one laser scanner, which is connected to a navigation device of the mobile part, which is connected to a control of the mobile part,
wherein the navigation device is connected to a volatile memory of the mobile part and is connected to a non-volatile memory of the mobile part.

The advantage here is that the navigation device is connectable to a controller which controls the handset and thus by means of the position determination a navigation, ie moving the handset along a trajectory, in particular target trajectory, to a predetermined target position out allows.

Further advantages emerge from the subclaims. The invention is not limited to the combination of features of the claims. For the skilled person, further meaningful combination options of claims and / or individual claim features and / or features of the description and / or figures, in particular from the task and / or the task posing by comparison with the prior art arise.

• In der 1 ist der Informationsfluss von den Laserscannern 11 des Mobilteils 32 bis zur Steuerung 14 des Mobilteils schematisch skizziert.
• In der 2 wird die zeitliche Abfolge bei der Positionsbestimmung näher dargestellt.
• In der 3 ist ein beispielhafter Verfahrbereich 33 mit dynamischen und statischen Objekten (31, 34) dargestellt.
• In der 4 wird der Verfahrbereich 33 in Draufsicht dargestellt.
• In der 5 ist eine Anlage gezeigt, innerhalb derer die Mobilteile 32 entlang einer Bahnkurve fahren.

The invention will now be explained in more detail with reference to schematic illustrations:

• In the 1 is the flow of information from the laser scanners 11 of the handset 32 to the controller 14 of the handset sketched schematically.
• In the 2 the time sequence in the position determination is shown in detail.
• In the 3 is an exemplary travel range 33 with dynamic and static objects ( 31 . 34 ).
• In the 4 becomes the travel range 33 shown in plan view.
• In the 5 is shown a facility within which the handsets 32 drive along a trajectory.

As in 1 shown points the handset 32 on its outer circumference two laser scanners 11 on, which are located diametrically opposite, in particular so on the handset to each other maximum distance. Thus, the detection of a first, especially in the direction of travel, half space by the first laser scanner 11 and the detection of a second, in particular in the direction of travel rear, half-space by the second laser scanner 12 allows.

The ones from the laser scanners 11 captured information is through communication channels 12 to a navigation device 13 which manages in a memory a map of the travel range of the handset and which also information of a odometrical detection system of the handset 32 be supplied.

Cyclic repetitive laser scans are performed. The laser scanners capture this 11 their sensitive area and transmit the measured data belonging to the objects detected during the laser scan, in particular in each case the respective distance and direction, to the navigation device 13 , This navigation device 13 determines the current position of the handset 32 by comparing, for possible positions, the data obtained during laser scanning with the positions of objects present in the card. In this case, the first of the possible positions is a position determined by odometry and from this, more distant positions are used. Once the most suitable position is found, this position will be used as the current position.

The navigation device 13 is with the controller 14 connected to the data exchange.

As in 2 represented, in a first process step, the laser scanning is performed, this step as a laser scan. recording 20 in 2 is marked. Thereafter, in a second method step, based on the odometrically acquired data since the last determined current position, the distance traveled and the direction are determined and thus relative to the last determined current position a new position.

This new position is then in a third step as a starting position for a Search method used, which compares the detected laser scan with the detected objects on the card. In this case, starting from the starting position, further positions are used to carry out the comparison until the appropriate position has been found.

As in 3 shown, there are different varieties of objects. These are static, so immovably arranged, objects 34 and dynamic objects 31 seen.

As in 4 In a laser scan, a plurality of light beams are emitted in different directions, and the distance to an object is determined for each of these directions. Thus, the result of the laser scan can be represented as a function or table which associates each of the directions with a respective distance.

As in 5 shown, the local traversing is arranged within a building or a plant.

Walls ( 50 . 52 ) of the building act as static objects ( 50 . 52 ).

A majority of handsets 32 moves within the travel range, In 5 are two handsets 32 shown, with a first with A and the second with B is marked. If the handset 32 A performs a laser scan, in the detection range, the handset 32 B penetrates, acts the handset 32 B as a dynamic object.

For dynamic objects can be detected at one position on one laser scanner - but can not be detected at this position by all laser scanners.

In addition, parked pallets function as semidynamic objects which can be detected several times by laser scanning, but not in all laser scans.

The method according to the invention comprises a first method step, in which a first movement of the mobile part 32 is performed, wherein the handset is manually steered or otherwise guided along a trajectory. Such another embodiment is for example a tracking along a laid on the ground primary conductor.

The laser scans performed during this journey are evaluated by storing the detected objects in a map and displaying them graphically.

In a second method step, regions, ie spatial regions, are entered into the map, in particular marked in the map, which are provided for static objects, and regions are entered, in particular marked in the map, which are provided for semidynamic objects. The remaining regions of the travel range, including the map, are intended for dynamic objects.

In this input, especially during this drawing, an orientation on the detected and graphically displayed in the map objects is helpful.

For example, walls or pillars of the building are included in respective static regions. For example, a palette is classified in a semidynamic region. Another handset or person is located in the rest of the dynamic region.

In a subsequent process step, the handset independently carries out trips on the traveling surface and thereby orients itself according to the map. In order to determine the respective current position, laser scans are furthermore carried out in succession and the objects detected thereby, ie their distance and direction relative to the mobile part, are determined.

For this purpose, a new position, which is used as the starting position for the search method, is preferably estimated as the starting position for the search method from the last determined current position by means of the odometrically acquired measured values, from which then further positions are used. For each of the positions, the result of the respective laser scan is then compared with the information contained in the map. If a sufficient match is not detected, a next position is used and the laser scan related thereto, that is, the measurement data is evaluated relative to that next position and again compared to the information contained in the map.

To determine the current position of the handset 32 Thus, both objects in the static regions and objects in the semidynamic regions are used.

If, during a subsequent journey, an object could no longer be detected in a semidynamic region that was previously present there, this object is no longer used to determine the position but only the objects currently detected based on the objects recorded in the map.

Thereafter, the map is updated by deleting the currently no longer detected object in the map.

Objects in dynamic regions are not used for position determination.

The handset 32 in particular the navigation device has a computer which is connected to a volatile and to a non-volatile memory. Although the card is stored in the volatile memory and then updated there, however, a backup copy of the current card is copied cyclically recurring in the non-volatile memory.

The card is preferably deposited as a grid, wherein the traversing area is divided into grid boxes whose dimensions are for example 10 cm by 10 cm. Each of these grid boxes is given the feature dynamic region, semidynamic region or static region. Thus, the determination of the objects must be carried out only with a correspondingly low resolution.

In further exemplary embodiments according to the invention, only the regions which are not dynamic are marked in the first method step, ie during the first drive. It is therefore initially not distinguished by static and semi-dynamic regions.

The objects detected during the first trip are entered in the map.

In the subsequent journeys it is then observed whether one of the objects entered in the map is missing. If this is not the case even after a predetermined number of trips, this object is marked as a static object.

Thus, the semidynamic region may have static objects over time.

LIST OF REFERENCE NUMBERS

11

laser scanner

12

communication channel

13

navigation device

14

control

15

communication channel

20

Laser scanning acquisition

21

odometry

22

location

23

Time Delay

31

dynamic object, especially human

32

Handset, in particular driverless transport system

33

traversing

34

static object

41

abscissa

42

laser beam

43

ordinate

50

static object

51

semidynamic object

52

static object

Claims (15)

Method for determining the position, in particular the respectively new current position, of a mobile part movable on a travel surface, in particular on a travel surface of a system, wherein the mobile part has at least one laser scanner, in particular for detecting the relatively relative positions of the mobile device on the travel surface arranged objects, wherein the mobile part has an odometric system, in particular for determining a distance covered by the mobile part, characterized in that in a first method step, the mobile part is moved along a trajectory on the traversing surface, in particular manually guided or track guided, and by means the laser scanner and the odometric system, the positions of objects are determined, wherein the positions of a map are assigned and stored in a particular volatile memory, wherein in a second method step, the card together with the positions of the associated objects with a display means, in particular screen, is displayed and regions are characterized as static or as dynamic, in particular by means of an input device, wherein the unmarked regions are regarded as semidynamic regions, in particular treated, wherein in a third process step Handset is moved on the traversing surface, in particular where the handset on the traversing surface, in particular for fulfilling a driving order, self-propelled driven navigated, wherein by means of the laser scanner, the positions of objects are detected relative to the handset and used to determine the new current position of the handset wherein for determination of the new current position of the handset, only the objects detected in regions identified as static and the objects detected in regions identified as semidynamic are used, wherein to determine the new a In the current position of the handset, a table is created which a plurality of possible, in particular virtual, positions assigns a respective evaluation parameter, wherein one of the possible positions of the set is an odometrically determined position, in particular the odometrically currently determined position, as the new current position of the handset that of the possible positions is determined, which in particular is assigned to the table related, extreme evaluation parameters, in particular the smallest of these evaluation parameters, wherein for each of the possible positions, the particular absolute positions of the detected objects are determined by the relative to the handset using laser scanner positions of the detected objects to the respective possible Position are added in particular vectorially and then the respectively assigned evaluation parameter is determined. Method according to Claim 1 , characterized in that in the third method step the presence of the changes in semi-dynamically marked regions is checked, and depending on the result of the check, the objects in these regions are deleted or newly added. Method according to at least one of the preceding claims, characterized in that in the third method step the statically and dynamically marked regions are not updated, in particular where walls are marked as static regions and traffic routes, in particular roads, are characterized as dynamic regions. Method according to at least one of the preceding claims, characterized in that in the first method step, the positions of the objects are determined by the laser scanner relative to the handset related position is added to the odometrically determined position of the handset. Method according to at least one of the preceding claims, characterized in that the mobile part has a controller, an energy storage and controllable by the controller, can be supplied from the energy storage drive, so that the handset moves on the travel surface, in particular so that the handset autonomously on the Moving surface is moved. Method according to at least one of the preceding claims, characterized in that the sum of the squares of the distances of the detected objects to the respectively corresponding objects in the map is determined for checking the correspondence with the map. Method according to at least one of the preceding claims, characterized in that by means of the odometric system, the time profile of the steering angle of the handset and the time course of the speed of the handset is detected, in particular wherein the rotational speed of a drive wheel is detected and from the respective current speed under consideration the diameter of the drive wheel, in particular and the kinematic model of the handset, the speed is determined and wherein the steering angle of the handset is detected. Method according to at least one of the preceding claims, characterized in that the evaluation parameter is the sum of the squares of the distances between the respective position of a respective detected object and the position stored in the map of an object closest thereto, in particular if these objects are in one as static or arranged semidynamically marked region, or that the evaluation parameter is the sum of the amounts of the distances between the respective position of a respective detected object and the stored in the map position of a nearby object, in particular if these objects arranged in a region marked as static or semidynamic are. Method according to at least one of the preceding claims, characterized in that it is checked in a fourth step, whether among the objects detected in a laser scan objects of the stored in the map in a static region objects was not detected, and depending on the result of this check a warning or error message is displayed or forwarded. Method according to at least one of the preceding claims, characterized in that the card is designed as a grid of surface elements, in particular a regular grid of surface elements, in particular so that the range of values of the coordinates of the positions is discrete, in particular wherein the surface elements are labeled as cells. Method according to at least one of the preceding claims, characterized in that the card and the positions assigned to the card are stored in a volatile memory, wherein the respective current map and the positions assigned to the map are stored in a non-volatile memory in a time-recurring manner, wherein after failure of the power supply of the navigation device with volatile memory the information last stored on the non-volatile memory is assigned via the map and the map associated with the map Positions are stored in a non-volatile memory. Method according to at least one of the preceding claims, characterized in that in the second method step, the regions are characterized only as dynamic, wherein the non-marked regions are considered as semidynamic regions, and / or that in the third process step, the regions occupied by objects, which after a predetermined Number of laser scans and / or after a predetermined period of time are detected unchanged, in particular automatically be marked as static. Method according to at least one of the preceding claims, characterized in that in the third method step, the temporal rate of change of occupancy of a surface element is calculated, the calculated rate of change is compared with a threshold, and depending on the result of the test, the surface element is automatically marked as static or semidynamic. System with handset for carrying out a method according to at least one of the preceding claims, characterized in that the mobile part has at least one laser scanner, which is connected to a navigation device of the mobile part, which is connected to a controller of the mobile part, wherein the navigation device with a volatile memory is connected to the handset and connected to a non-volatile memory of the handset. Plant after Claim 14 , characterized in that the handset has at least one distance measuring device based on ultrasound or radar signals or imaging sensors instead of a laser scanner, and / or that the odometry values are not measured but estimated, in particular by extrapolation of previously determined localization values.

Images