Method for cleaning a room using a self-propelled cleaning device

Abstract

Method for cleaning a room (R) by means of an automatically movable cleaning device (1), wherein the cleaning device (1) has a map-like representation (K) of the room (R), at least with regard to the boundary walls (W), characterized in that the room (R) is detected by means of an algorithm in accordance with the majority of the walls (W') running in one direction, whereby a main orientation (H, H') of the room (R) results due to the longitudinal orientation of the walls (W'), and that the cleaning device (1) then travels over the area of ​​the room by moving back and forth, oriented along the main orientation (H, H') or displaced parallel thereto.

Description

[0001] The invention relates to a method for cleaning a room by means of an automatically movable cleaning device, wherein the cleaning device has a map-like representation of the room, at least with regard to the boundary walls.

[0002] Methods of the type in question are known. The cleaning device used to clean the room floor moves automatically according to a preferably pre-programmed driving and, if applicable, behavior strategy. In this context, it is also known that the cleaning device has a map or map-like representation of the room to be cleaned, possibly several maps for a corresponding number of rooms. This map is preferably stored in a non-volatile memory and is thus available for each cleaning process. Alternatively, methods are known in which the cleaning device first performs such a mapping before beginning to clean the room floor, in order to move around the room after creating the map. In the map-like representation of the room, the boundary walls are particularly preferably marked.For map-like detection of the room, it is also known to have the cleaning device move, in particular along the boundary walls, as part of a learning run, more preferably as a result of remote control of the cleaning device by the user. In another known embodiment, the cleaning device has its own means for detecting the room, in particular for detecting the room boundaries, more preferably in the form of an all-round scanner that represents obstacle detection. In this regard, reference is made to DE 10 2008 014 912 A1. The content of this patent application is hereby fully incorporated into the disclosure of the present invention, also for the purpose of incorporating features of this patent application into claims of the present invention. From this patent application, an obstacle detection system based on an optical triangulation system is known.The system's optical elements are mounted on a rotating platform, allowing a 360° scan. The resulting distance measurements allow the cleaning device to create a room map. The cleaning device preferably remains stationary; however, it can also move around the room floor to create an appropriately adjusted map, particularly for projections and recesses in the room's floor plan.

[0003] Furthermore, DE 10 2004 004 505 A1 discloses a cleaning device that detects environmental features along the outer contour of a floor surface and creates an environmental map from them. The floor surface is divided into subsegments, which are then processed sequentially.

[0004] In view of the prior art described above, a technical problem of the invention is seen in further improving a method of the type in question, in particular with regard to the driving strategy to be selected.

[0005] This problem is initially and essentially solved by the subject matter of claim 1, which involves the room being recorded by means of an algorithm according to the majority of walls running in one direction, whereby a main orientation of the room results due to the longitudinal orientation of the wall and the cleaning device then travels back and forth across the room area oriented along the main orientation or displaced parallel to it. As a result of this embodiment, a favorable travel strategy of the cleaning device for traveling over and cleaning the floor of the room can be derived in particular from the floor plan of the room. After a corresponding analysis of the room or the map-like representation of the room, the cleaning device travels along quasi-logical travel paths resulting from this. For this purpose, the majority of walls running in one direction are recorded by means of a control-side algorithm of the cleaning device.the longest boundary line running in one direction in the map-like representation is recorded, which wall or boundary line in the map serves as a reference for the travel strategy to be created spontaneously for this room. If, in particular, the map-like representation of the room does not allow the detection of several walls running in the same direction and correspondingly parallel - for example, in the case of a room floor plan in the form of an irregular polygon with no parallel edges or, for example, in the case of a map-like representation of a partial area of ​​the room created as a result of a learning drive - a rectangle with the longest possible dimensions is virtually projected into the free area of ​​the map-like representation or within the surrounding boundary of the map-like representation, after which the longitudinal orientation of the rectangle or the long side of the rectangle is used for the subsequent travel strategy calculation.This preferably also applies if obstacles are detected within the room area, which can then also have boundaries running parallel to one another, such as projections or room dividers. If most of the walls running in one direction are detected, the primary orientation of the room, more particularly the primary longitudinal orientation of the room, is determined using a line algorithm. In one embodiment, the cleaning device then travels across the room area oriented towards the main orientation of the room, and more preferably in a parallel orientation thereto, i.e. in a parallel orientation to most of the walls or to the long side of the rectangle. In an alternative embodiment, the cleaning device travels perpendicular to the main orientation of the room with reference to a floor plan of the room area, accordingly preferably crossing the room area in the width direction thereof.Furthermore, the room area is preferably completely covered and preferably cleaned by moving the cleaning device back and forth, wherein the respective travel lines of the cleaning device preferably run parallel to one another, more preferably offset parallel to the main orientation of the room, alternatively in a perpendicular orientation thereto. The back and forth travel of the cleaning device is achieved by turning it 180° when it reaches a room boundary that crosses the travel line. To create a correspondingly adapted travel strategy or to calculate at least the first travel line to be followed by the cleaning robot, further algorithms are preferably used, such as in particular cell decomposition algorithms or area coverage algorithms, which receive the main orientation of the room detected in the first step as input parameters. In this case, it is more preferably sufficient to only use a first, e.g.to calculate a travel path of the cleaning device running parallel to the main orientation of the room, more preferably parallel to most of the walls running in one direction, after which, when the automatic cleaning of the room surface begins and when a room boundary crossing the travel path is reached, according to a behavior strategy preferably provided in the cleaning device, the device automatically turns by 180° and then uses a travel path offset parallel to the previous travel path in the opposite direction of travel. In this way, a furrow-turn-like course of the entire travel path is preferably achieved.

[0006] In this context, it is further preferred that the cleaning device, for example after switching on an assigned base station, orients itself preferably once within the apartment environment and in doing so establishes an orientation baseline for all rooms. Alternatively, the cleaning device works with individual maps for each room and with individual orientation baselines, which are mathematically linked to one another by path maps to create an overall map of the apartment environment in such a way that the differences between the various orientation baselines are known and taken into account. In order to specify the deviation of the specified room orientation from the orientation baseline of the cleaning device, the angle between the calculated room orientation and the orientation baseline stored on the map (of the room in question) is preferably determined and, if necessary, saved.

[0007] Furthermore, the algorithm can also be used to determine the longest wall running in one direction to determine the room orientation.

[0008] The proposed method advantageously achieves energy-efficient driving behavior of the cleaning device. A clear basic orientation of at least one main direction of travel is predetermined based on appropriate calculations. The number of changes of direction when traveling across the entire room area is preferably reduced to a minimum. Accordingly, energy is saved compared to a driving strategy that appears random to the user, with the result that the cleaning device can travel and clean for longer with the same battery charge. Furthermore, an "intelligent" driving strategy also results for the user or observer. This is particularly evident in approximately rectangular floor plans, for example in an elongated hallway.Here, it is also possible for the algorithm to take into account the ratio of a longest wall running in one direction to a longest wall running perpendicular to it. If the ratio exceeds a maximum value preferably stored in the control system (e.g. 3:1, 5:1 or 10:1), a travel line running parallel to the main orientation of the room is preferably selected using the subsequent algorithms, so that in the exemplary case of an elongated corridor, the cleaning device moves along the longitudinal direction of the corridor and requires fewer turns than with a travel line selected perpendicular to the main orientation of the room. In the event that no plurality of walls running in one direction can be determined or that the rectangle to be projected into the room results in a square, the algorithm preferably does not determine a main orientation of the room, which mayHowever, the error value passed on is preferably implemented by subsequent algorithms in such a way that, for example, the wall closest to the device is defined as the wall along which the main alignment is assumed or the travel line of the cleaning device is calculated.

[0009] Further features of the invention are explained below, also in the description of the figures, often in their preferred association with the subject matter of claim 1 or with features of further claims. However, they may also be important in association with only individual features of claim 1 or the respective further claim, or each may be important independently.

[0010] For example, in a further development, the cleaning device carries out a wall-following movement before and / or after traveling across the room area. The room boundaries (preferably walls) determined or stored in this way based on the map-like representation serve to extract the start and stop processes of the wall-following movement, so that more preferably the starting point of the wall-following movement of the cleaning device is also the stopping point of the wall-following movement after the room boundary has been completely traveled. If the wall-following movement takes place before traveling across the room area, the stopping point of the wall-following movement is, in a preferred embodiment, also the starting point for traveling across the room area. If the wall-following movement only takes place after traveling across the room area, then in a preferred embodiment the stopping point of the room area departure defines the starting point of the wall-following movement.Such a wall-following movement is also possible both before and after traveling across the room surface according to the previously described driving strategy.

[0011] In a further embodiment, it is preferably provided that a map-based subdivision of the room is carried out into different areas, wherein more preferably each area is cleaned according to the method described above. The map-based subdivision of a room into different areas is preferably carried out, for example, in rooms composed of different floor plans, more particularly in floor plans in which, for example, two or more rooms flow into one another without further boundaries. The map-based subdivision is more preferably carried out before the main orientation of the room is determined, more preferably during a mapping process as described above. For this purpose, artificial area boundaries are more preferably predetermined - possibly to be created by the user - so that sub-areas of the overall room are present within the map-like representation of the overall room stored in the controller.Alternatively, it is also possible to analyze the entire room, which will later consist of several different areas, using an algorithm and to calculate a strategically advantageous division of the areas, preferably resulting in the smallest possible number of different areas. Each of the specified or determined areas of the room is analyzed and cleaned according to the previously described procedure, so that, for example, after one area has been processed and the cleaning device has been transferred to the other area, the main orientation of the room—in this case, the area—can be changed in order to preferably process the area in this next area in an energy-efficient manner.

[0012] The invention is explained in more detail below with reference to the accompanying drawings, which merely represent exemplary embodiments. They show: Fig. 1 in perspective view a cleaning device with obstacle detection for all-round detection; Fig. 2 a map-like representation of a room, for determining a main orientation of the room in a first embodiment; Fig. 3 one of the Fig. 2 corresponding representation, the movement strategy of the cleaning device within the map according to Fig. 2 concerning the performing space; Fig. 4 one of the Fig. 2 corresponding map-like representation of another room, into which map a rectangle is projected to capture a main orientation of the room (not an embodiment of the invention); Fig. 5 one of the Fig. 3 corresponding representation, but the map or the room according to Fig. 4 concerning; Fig. 6 another map-like representation of a room, divided into two areas; Fig. 7 another of the Fig. 3 corresponding representation, but the movement strategy of the cleaning device for cleaning the divided areas according to Fig. 6 concerning.

[0013] It is presented and described first with reference to Fig. 1 a self-propelled cleaning device 1, preferably in the form of a vacuum and / or sweeping device, further in the form of a movable household floor cleaning device. This has a chassis which, on the underside facing the floor to be cleaned, carries electric motor-driven travel wheels and a brush which protrudes beyond the lower edge of the chassis floor and is also electric motor-driven. The chassis is covered by a device hood 2, whereby the device 1 has a circular floor plan. With regard to the design of the cleaning device 1 as a vacuum and / or sweeping device, reference is made to DE 102 42 257 A1. The content of this patent application is hereby incorporated in its entirety into the disclosure of the present invention, also for the purpose of incorporating features of this patent application into claims of the present invention.

[0014] Furthermore, although not shown, the cleaning device 1 can have a suction port in addition to or as an alternative to the brush. In this case, an electrically operated suction fan motor is also arranged in the device 1.

[0015] The electrical supply of the individual electrical components of the device 1, such as the electric motor, the travel wheels, the electric drive of the brush, if applicable, the suction fan and, in addition, the further electronics provided in the device 1, preferably for controlling the same, is provided via a rechargeable battery (not shown).

[0016] The cleaning device 1 is further equipped with an obstacle detection system 3 in the form of a laser scanner, which is centrally located on the device hood 2 and can rotate 360° about a vertical axis x. The rotating obstacle detection system 3 enables a map-like recording of the room R in which the cleaning device 1 is located. By scanning the room boundary walls W all around, a map-like representation K of the respective room R is created and preferably stored. The scanning of the room boundaries is preferably carried out, as described above, using a triangulation system, more preferably using a distance measurement.

[0017] The cleaning device 1 further comprises control and regulation electronics, further preferably for controlling and regulating the cleaning device 1 during its movement and further preferably during the cleaning of the room floor carried out during this process. In order to calculate a particularly effective driving strategy, and also a strategy of the cleaning device 1 that appears sensible to the user or observer, the provided control and regulation electronics first perform an analysis of the main orientation H of the room R relative to an orientation baseline O within the map K before the cleaning of the room floor begins.

[0018] For this purpose, in a first step, a predetermined algorithm is used to determine the majority of walls W' running in one direction and thus parallel to one another. This is done with the aid of the map-like representation K, in which the boundary walls W and thus also the majority of walls W' running in one direction are represented by boundary lines of the map area representing the room area. The longitudinal orientation of these majority of walls W', according to appropriate analysis, results in the main orientation H of the room R, which is directed parallel to it and which, if necessary, encloses an angle α with the map-side orientation line O.

[0019] In the event that a plurality of walls running in one direction cannot be clearly determined (e.g. in the case of an elongated rectangular floor plan with two parallel walls and two transverse walls running perpendicular to them and to each other), the orientation of the longest boundary wall W is preferably included in the determination of the main orientation H of the room R in the course of a further analysis.

[0020] In another (non-inventive) case - as in Fig. 4 schematically shown - in which no plurality of walls running in the same direction or a longest wall can be clearly identified, in an alternatively proposed solution a virtual rectangle A is projected into the free area of ​​the room R delimited by the boundary walls W or into the free inner area of ​​the map-like representation K, which rectangle A has the longest possible dimensions with reference to the free inner area of ​​the room R or the map K. The orientation of the rectangle's long side C defines the main orientation H of the room R, preferably relative to the orientation baseline O.

[0021] The main orientation H is accordingly parallel to the detected plurality of walls W' running in one direction or to the rectangular long side C of a rectangle A projected into the space R.

[0022] After a corresponding calculation of the main orientation H of the room R, the cleaning process preferably begins automatically as the cleaning device 1 moves across the room floor.

[0023] In this case, it is further preferred that the cleaning device 1 first performs a wall-following movement, during which the cleaning device 1 moves along the boundary walls W of the room R, preferably oriented according to the available map-like representation K. In this case, a starting point 4 of the wall-following movement is also the stopping point thereof. Upon reaching the starting point 4 after the wall-following movement, the cleaning device 1 switches to a mode for traveling and preferentially cleaning the room surface, in which mode further algorithms receive the previously determined main orientation H as an input parameter. The cleaning device 1 accordingly moves preferably parallel (alternatively at an angle of 90°) to the determined main orientation H, after which the room surface is processed in furrows (cf. Fig. 3 and Fig. 5).

[0024] In addition, a map-based subdivision of the space R into different areas B and B' is possible, for example in the case of a space R as shown in Fig. 6, which essentially consists of two room sections with different floor plan designs, for example, due to the arrangement of an ancillary room not separated from a main room by a boundary wall. The corresponding subdivision of the map-like representation K is preferably carried out immediately after the creation of the map K, or alternatively immediately before a corresponding analysis of the room orientation.

[0025] In this case, the control of the cleaning device 1 determines an adapted driving strategy for each area B or B', thereby determining the respective main orientation H or H' of the individual areas B and B'.

[0026] This further results in the Fig. 6 and Fig. 7, a map K composed of two areas B and B' has a main orientation H' of the area B', which main orientation H' is oriented perpendicular to the main orientation H of the area B.

[0027] Both areas B and B' are calculated according to the Fig. 3 and Fig. 4, preferably initially by performing a wall-following travel, wherein a virtual boundary line G between the areas B and B' is used to perform a transfer travel of the wall-following travel. The subsequent, preferably automatic, travel across the room area for cleaning the same occurs as a result of the cleaning device 1 traveling along travel lines running parallel to the respective main orientation H or H', which more preferably run in a furrow-turning manner. List of reference symbols 1 cleaning device 2 device hood 3 Obstacle detection 4 Starting point x axis of rotation A rectangle B area B' area C Rectangle long side G boundary line H Main orientation H' main orientation K Card O Orientation baseline R Room W boundary wall W' longest wall α angle

Claims

[1] Method for cleaning a room (R) by means of an automatically movable cleaning device (1), wherein the cleaning device (1) has a map-like representation (K) of the room (R), at least with regard to the boundary walls (W), characterized by that by means of an algorithm the room (R) is detected according to the majority of walls (W') running in one direction, whereby a main orientation (H, H') of the room (R) results due to the longitudinal orientation of the walls (W') and that the cleaning device (1) then travels over the room area oriented along the main orientation (H, H') or displaced parallel thereto by moving back and forth. [2] Method according to claim 1, characterized by that the cleaning device (1) carries out a wall-following movement before and / or after moving across the room area. [3] Method according to one of the preceding claims, characterized bythat a map-like subdivision of the room (R) into different areas (B, B') is made, each area (B, B') being cleaned according to this method.

Images