WO2020148138A1

WO2020148138A1 - Robotic tool

Info

Publication number: WO2020148138A1
Application number: PCT/EP2020/050356
Authority: WO
Inventors: Marcus LILJEDAHL; Jonas RANGSJÖ
Original assignee: Husqvarna Ab
Priority date: 2019-01-15
Filing date: 2020-01-09
Publication date: 2020-07-23
Also published as: EP3912005A1; SE542918C2; SE1950041A1

Abstract

The present disclosure relates to a self-propelled robotic tool (1) comprising a work implement (3), a driving arrangement (5, 7, 9, 11) for moving the robotic work tool, a control device (13) for controlling the driving arrangement, and an acquisition device (23, 25), configured to record a signal (27; 31) transmitted in bursts by a boundary wire loop (19), wherein the control device is configured to determine whether the robotic work tool is located within an area (17) defined by the boundary wire loop based on the recorded signal. There is provided a digital signal processor (39) which is configured to provide an output corresponding to an average of a plurality of recorded transmitted signal bursts, wherein the control device (13) is configured to determine whether the robotic tool (1) is located within said area (17) at least partly based on said output.

Description

ROBOTIC TOOL

Technical field

The present disclosure relates to a self-propelled robotic tool comprising a work implement, a driving arrangement for moving the robotic work tool, a control device for controlling the driving arrangement, and an acquisition device, configured to record a signal transmitted in bursts by a boundary wire loop, wherein the control device is configured to determine whether the robotic work tool is located within an area defined by the boundary wire loop based on the recorded signal.

Further, a corresponding method for controlling a self-propelled robotic tool is considered.

Background

Such a self-propelled robotic tool is disclosed for instance in EP-2741160-A1 , where a recorded signal is correlated with a reference signal, or the like, in order to determine whether the robotic work tool is located within the area in question.

One general problem associated with tools of this type is how to increase the reliability in determining whether the robotic tool remains located within the area defined by the boundary wire loop.

Summary

One object of the present disclosure is therefore to provide a self-propelled robotic tool which determines with an increased reliability whether the tool is located within a work area.

This object is achieved by means of a self-propelled robotic tool as defined in claim 1. More particularly, in a self-propelled robotic tool of the initially mentioned kind, a digital signal processor is provided which is configured to provide an output corresponding to an average of a plurality of recorded transmitted signal bursts. The control device is configured to determine whether the robotic tool is located within said area at least partly based on said output. This allows the determining of whether the robotic tool remains located within the area in question in a more reliable manner in particular to deal with noisy conditions. The output may correspond to an average of a sequence of bursts repeated with an estimated period time (P).

In this average burst, the order of flanks or other pulse characteristic may then be determined and compared to a reference. Alternatively, the determined average may be correlated with a reference pattern.

Typically, the average may be made up from between 2 and 20 of the most recently received pulses.

The estimated period time may be determined using a recursive algorithm.

Typically, the self-propelled robotic tool may be a robotic lawn mower.

It is possible to prepare in parallel two or more averages, based on different pluralities of recorded transmitted signal bursts.

The present disclosure further considers a corresponding method comprising the steps carried out in a self-propelled robotic tool as defined above.

Brief description of the drawings

Fig 1 illustrates schematically a self-propelled robotic tool operating in a work area.

Fig 2 schematically illustrates system units of the self-propelled robotic tool in fig 1.

Fig 3 illustrates an example of burst with certain signatures or patterns.

Fig 4 shows an example of a signal with burst having specific flank orders.

Fig 5 illustrates digital signal processing using a number of successive bursts.

Fig 6 illustrates schematically units in a system carrying out a detection method and fig 7 illustrates an example flowchart of such a detection method.

Detailed description

The present disclosure relates generally to self-propelled robotic work tools, of the kind disclosed for instance in EP-2741160-A1 , where a robotic mower for cutting lawns is shown as an example. Such a mower is shown schematically in figs 1 and 2 where a mower 1 has a cutting implement 3, e.g. a rotating knife, and wheels 5, 7 typically driven by electric motors 9, 11. There may also be provided non-driven, e.g. swiveling wheels, and various other wheel configurations are possible. The steering of the robotic work tool is carried out by controlling the driving motors 9,

11 by means of a control unit 13, of. fig 2. A battery arrangement 15 powers the motors 9, 11 as well as the control unit 13.

By controlling the motors 9, 11 , the control unit 13 moves the robotic work tool 1 over a work area 17, typically a lawn. A boundary wire 19, which may typically be buried a few centimeters deep in the ground, defines the work area 17, the boundary wire 19 forming a loop, which is typically closed at a transmission unit 21 , transmitting a signal on the boundary wire 19. The ends of the boundary wire 19 are typically connected to the transmission unit 21 which may be integrated with a charging station that can be used to intermittently charge the battery arrangement 15.

The transmission unit 21 thus feeds an electric signal to the boundary wire 19 that operates as a form of antenna, radiating an electromagnetic signal which the robotic work tool 1 can pick up using one or more coils. Based on the signal, the robotic work tool can be controlled, typically making sure that the robotic work tool remains within the work area 17. In the aforementioned document, the transmitted signal comprises bursts of signals with predetermined patterns and separated by pauses. Within the bursts, the signal switches between high and low states, where the duration of the high states is different from the duration of the low states within a burst.

The electromagnetic signals transmitted by the boundary wire 19 are thus picked up by two or more coils 23, 25 connected to the control unit 13 in the robotic work tool 1. The picked-up signal may be more or less equivalent to the first order derivative of the current fed to the boundary wire 19, and therefore also the signals fed to the control unit 13 have a distinguishable pattern. If all of the two or more coils 23, 25 are located inside the work area 17 defined by the boundary wire 19, the patterns will be similar, and any disturbances will also affect both signals similarly. When the robotic tool 1 is about to move outside the work area, however, one of the coils 23 will be located outside and another one 25 inside the work area 17. At this point, the signals acquired by the first and second coils 23, 25 will become diametrically different.

It has also been suggested to use a transmission unit 21 , transmitting a signal 27 with bursts having a predetermined pattern 29 or signature as illustrated in fig 3.

When receiving a signal, using one single coil, the self-propelled robotic tool may correlate the received signal with an internally stored template. As long as the correlation between the received signal’s pattern and the template is high, the self- propelled robotic tool is considered to remain within the work area 17. When the self- propelled robotic tool moves outside this area however, the phase of the received signal reverses and the correlation drops, which can be determined by the control unit 13.

Another example of a signal transmitted via a boundary wire is described in EP- 1512053-A1. In that example, a signal 31 with bursts comprising a set of pulses with positive and negative flanks (voltage) having a predetermined order is transmitted, as illustrated in fig 4. It may be determined, on the receiving side, for instance that a burst beginning with a positive pulse with a certain length is received, followed by a negative pulse of a certain length. This may be sufficient to determine whether or not the self-propelled robotic tool remains within the work area. Bursts are repeated with a period time P or offset time. The order of flanks is one example of a pulse characteristic, other examples include lengths of pulses in a burst and/or delays therebetween.

In both the initially described schemes however, noise and interference may make the detection difficult. This has been addressed by using a high enough burst amplitude such that correct detecting takes place. That however may lead to high power consumption and, more importantly, to excessive interference disturbing other systems nearby.

In the present disclosure, correct detection is instead achieved by using a digital signal processor which provides an output based on an average of a plurality of recorded transmitted signal bursts. The control device is configured to determine whether the robotic tool is located within said area at least partly based on said output.

An example of this is illustrated in fig 5 where a received signal 33 corresponding to the transmitted signal of fig 4 is shown. As illustrated, the received signal may be very noisy, and it may be difficult to assess e.g. the order of flanks, positive-negative- positive based on an individual burst 35. However, by using a digital signal processor it is possible to superpose a plurality of such individual burst to form a sum corre sponding to an average burst 37 thereof. In this signal e.g. the order of flanks can be detected more easily. By a burst is here generally meant a varying signal that lasts for a certain amount of time.

As mentioned, the averaging may typically be made up from between 2 and 20 of the most recently received bursts. However, it is also possible to provide parallel averages based on different numbers of recent bursts, such as for instance 3 and 15 pulses. A small number of pulses will give an average that changes more quickly, while a greater number is more stable. Which one to use may be determined by noise conditions or other quality parameters. It is also possible to make decisions based on comparison of averages based on different numbers of bursts.

Fig 6 illustrates schematically units in a system carrying out a detection method and fig 7 illustrates an example flowchart of such a detection method. The method may be carried out by a digital signal processor 39, DSP, in the control unit 13. By a DSP is here meant circuitry carrying out digital operations on data although other hardware configurations are conceivable. Further, a buffer 41 and a timing unit 43 may be provided, which are shown as individual units although they may be integrated for instance in the DSP 39.

With reference to fig 7, the method may begin with clearing 45 the buffer 41 or memory. The system waits 47 an estimated period time P_est, and thereafter samples 49 data input from the coil 23. The sampled data is added 51 to the buffer 41. By adding is here meant that the received signal is added to corresponding data from previous pulses such that the value at a time t, v_t is added to the corresponding value of previous pulses v_t-p+v_t-2_P+... etc. The sum in the buffer thus corresponds to an average of pulses received since the last clear event, even if not divided by the number of pulses. Also, other and different FIR- and MR filter configurations are possible.

Detection 53 is then done on the data in the buffer, whether by testing the order of flanks or other pulse characteristic, or carrying out a correlation procedure. If the detection is considered successful 55, the process may go on to report the condition (inside/outside work area) to other functions in the control unit 13. If, however, the accumulated data is insufficient to carry out detection, the process in the DSP 39 may instead wait another period P, sample a new burst and add to the buffer 49, 51 , and again attempt to detect 53 the condition. Once a condition has been reported 57, it may be useful to update the estimated period time. This may be done by adjusting the previously estimated period P_est which was used during the last accumulation as a linear combination with the period time P_de_t that can be determined from the last accumulated bursts, e.g. such that the new estimated period time Pestnew can be defined as: P_estnew= 0.9^*P_est+0.1 ^*Pdet .

This provides a recursive algorithm that adapts the detection to the transmitted signal and may be further developed using Kalman filters and the like.

The present disclosure is not limited to the above-described examples and may be varied and altered in different ways within the scope of the appended claims.

For instance, even if the transmitted signals have been shown as short bursts interleaved with long pauses, the burst may be as long as the period time such as the signal appears to be transmitted continuously.

Claims

1. A self-propelled robotic tool (1 ) comprising a work implement (3), a driving arrangement (5, 7, 9, 11 ) for moving the robotic work tool, a control device (13) for controlling the driving arrangement, and an acquisition device (23, 25), configured to record a signal (27; 31 ) transmitted in bursts by a boundary wire loop (19), wherein the control device is configured to determine whether the robotic work tool is located within an area (17) defined by the boundary wire loop based on the recorded signal, characterized by a digital signal processor (39) which is configured to provide an output corresponding to an average of a plurality of recorded

transmitted signal bursts, wherein the control device (13) is configured to determine whether the robotic tool (1 ) is located within said area (17) at least partly based on said output.

2. Self-propelled robotic tool according to claim 1 , wherein the output corresponds to an average of a sequence of pulses repeated with an estimated period time (P).

3. Self-propelled robotic tool according to claim 2, wherein a pulse characteristic in the determined average is compared to pulse characteristic reference.

4. Self-propelled robotic tool according to claim 2, wherein the determined average is correlated with a reference pattern.

5. Self-propelled robotic tool according to any of claims 2 to 4, wherein the average is made up from between 2 and 20 of the most recently received pulses.

6. Self-propelled robotic tool according to any of claims 2 to 5, wherein the estimated period time is determined using a recursive algorithm.

7. Self-propelled robotic tool according to any of the preceding claims, wherein the self-propelled robotic tool is a lawn mower.

8. Self-propelled robotic tool according to any of the preceding claims, wherein in parallel two or more averages, based on different pluralities of recorded transmitted signal bursts, are prepared.

9. A method for controlling a self-propelled robotic tool comprising a work implement, a driving arrangement for moving the robotic work tool, a control device for controlling the driving arrangement, and an acquisition device, configured to record a signal transmitted in bursts by a boundary wire loop, wherein the control device determines whether the robotic work tool is located within an area defined by the boundary wire loop based on the recorded signal characterized by a providing an output corresponding to an average of a plurality of recorded transmitted signal bursts, and determining whether the robotic tool is located within said area at least partly based on said output.