DE102021209036A1 - Method and system for operating a work machine in forestry - Google Patents

Abstract

A method for operating a working machine (100) in forestry, with the steps: detecting a point cloud associated with a stand of trees using a laser scanner (10), detecting a tree (2) in the detected point cloud, deriving a tree-specific parameter of the detected tree (2 ), Controlling a working device (20) of the working machine (100) to carry out a work step with the working machine (100) for performing a forestry work task depending on the derived tree-specific parameter. In addition, a forestry work system (200), which is set up to carry out the method.

Description

technical field

The present invention relates to a method for operating a working machine and to a working system for carrying out such a method.

State of the art

In forestry, a large number of work tasks have to be carried out in the tree population. Performing a work task may require the use of a work machine operated by a forest worker. The occupational safety of the forest worker can depend on a visual inspection of the work environment. While carrying out the work task, the forest worker may be exposed to environmental hazards which may not be perceptible during such a visual inspection by the forest worker.

Presentation of the invention

In one aspect, the present invention relates to a method for operating a working machine in forestry. The work machine can be a self-propelled work machine or a vehicle. The working machine can have a working device for performing a forestry work task. The self-propelled work machine can also be an agricultural machine or a construction machine.

As a step, the method has a detection of a point cloud assigned to a stand of trees with a laser scanner. The laser scanner can be designed as a two-dimensional or three-dimensional scanning sensor. The laser scanner can be a lidar sensor. The tree population can be selectively recorded or scanned with the laser scanner. The point cloud can be associated with the stand of trees as a discretization thereof, with the point cloud being able to discretize the stand of trees. Furthermore, the tree population can be recorded contact-free with the laser scanner. A point cloud, which can have a large number of measuring points related to the tree population, can thus be recorded with the laser scanner. The point cloud can be a two-dimensional or three-dimensional point cloud.

As a further step, the method includes recognizing a tree in the detected point cloud. The detected point cloud can have a large number of point coordinates. The tree can be recognized in the detected point cloud based on a segmentation for segmenting a geometric model of the tree in the point cloud or the point coordinates.

The method can therefore have a segmentation of the geometric model of the tree in the point cloud as a further step. The geometric model can be, for example, a cylindrical or frusto-conical geometry. The tree can be recognized in the captured point cloud based on a Kl-based extraction of the tree from the point cloud or the point coordinates. As a further step, the method can therefore include recognizing the tree in the point cloud based on an AI approach.

As an alternative or in addition to the segmentation, the tree can be recognized in the detected point cloud based on a feature extraction for extracting a geometric feature of the tree in the point cloud or the point coordinates. The method can therefore include, as a further step, extracting the geometric feature of the tree in the point cloud. As an alternative or in addition to the segmentation, the tree can be recognized in the detected point cloud based on a classification for classifying the tree in the point cloud or the point coordinates. As a further step, the method can therefore have a classification of the tree in the point cloud.

As a further step, the method includes deriving a tree-specific parameter of the identified tree. The tree-specific parameter can be a spatial parameter which can define a spatial position of the tree. The spatial location may include a spatial location of the tree. Alternatively or additionally, the spatial position can have a spatial orientation of the tree. The tree specific parameter can be derived from the segmented geometric model of the tree. Alternatively or additionally, the tree-specific parameter can be derived from the extracted feature of the tree. Alternatively or additionally, the tree-specific parameter can be derived based on the classification step.

As a further step, the method includes controlling a working device of the working machine to carry out a working step with the working machine for performing a forestry work task as a function of the derived tree-specific parameter. At the forestry work task it can be a forestry work task to be performed in a forest. The working device can be arranged on the working machine. The working device can be a device for driving the working machine to perform the forestry work task. The working device can therefore be controlled to drive or intervene in a drive of the working machine. The work step can therefore include driving or moving the work machine.

As an alternative or in addition to the device for driving the working machine, the working device can be a tool for carrying out the forestry work task. The work step can therefore include operating or controlling the work device or the tool to carry out the forestry work task. The tool can be a saw. As a further step, the method can include selecting the working device of the working machine to be controlled for carrying out the working step with the working machine for performing the forestry work task as a function of the derived tree-specific parameter.

According to one embodiment, the step of capturing the point cloud can be carried out repeatedly or multiple times. The steps of recognizing the tree and deriving the tree-specific parameters can also be carried out repeatedly or multiple times based on the repeatedly recorded point clouds. The method can therefore have a further step of comparing the repeatedly derived tree-specific parameter. The controlling step can then be carried out as a function of a comparison result based on the comparing step. The result of the comparison can show agreement or discrepancy among the repeatedly derived tree-specific parameters.

With the invention, a forestry work task can be carried out automatically based on the sensory detection of the tree population. In addition, with the invention, the forestry work task can be carried out more efficiently through automated control of the working machine, with occupational safety also being able to be increased through automated detection of the tree population in the working environment of the working machine. With the invention, you can therefore reduce the amount of work and wear and tear on the machine when performing the forestry work task.

According to a further embodiment of the method, in the step of capturing the point cloud, the point cloud can be captured using a laser scanner arranged on the work machine. The laser scanner can therefore be a scanning environment detection sensor that can be carried on the work machine. The laser scanner can have a detection field that deviates from the visual field of view of an operator of the working machine. In this way, areas in the working environment of the working machine that are shaded for the operator can be captured by the laser scanner. By detecting shaded areas, work safety can be guaranteed more reliably. As an alternative to the arrangement on the working machine, the laser scanner can also be arranged at a distance from or separately from the working machine. The detection of areas in the working environment of the work machine that cannot be seen by the operator can thus be further improved and work safety can be maintained in a task-specific manner.

According to a further embodiment of the method, a tree trunk can be recognized in the detected point cloud in the step of recognizing the tree. For example, the tree trunk can be detected based on segmenting a cylinder in the point cloud. The tree trunk can be a tree trunk of a standing or grown tree or a tree trunk of a felled tree. According to this embodiment, a parameter of the identified tree trunk can be derived in the step of deriving the tree-specific parameter. The tree specific parameter can be derived from the segmented cylinder. The identified tree trunk can be a tree trunk at or on which the forestry work task is to be performed. The identified tree trunk can also be a tree trunk that represents an obstacle or a hazard for the work machine or the operator when performing the forestry work task. Work tasks on a tree trunk can therefore also be automated and carried out more safely using the method.

According to a further embodiment of the method, a spatial orientation of the identified tree trunk can be determined in the step of deriving the tree-specific parameter. Thus, in the step of deriving the tree-specific parameter, a skewed position or inclination of the identified tree trunk can be determined. A fallen or at least partially broken tree trunk can also be determined in the step of deriving the tree-specific parameter. Alternatively or in addition to the determined spatial orientation, a trunk diameter or a trunk circumference of the recognized tree trunk can be determined in the step of deriving the tree-specific parameter. Thus, in the step of deriving the tree-specific parameter, an age or a mass of the identified tree trunk can be derived from the determined trunk diameter. As a further alternative or in addition to the determined spatial orientation, a trunk length of the identified tree trunk can be determined in the step of deriving the tree-specific parameter. The age or mass of the detected log can be derived from the log length of the detected log.

Thus, for example, tree-specific parameters relevant for felling a tree can be determined automatically with the method and the working device can be controlled based on this for felling the tree. The tree-specific parameter can also be a foliage of the tree, which can also be derived from at least one of the other parameters described. For example, the foliage can be derived from a point density in the point cloud. The work machine can thus also be controlled to carry out the work task depending on the foliage of the identified tree.

According to a further embodiment of the method, this can have, as a further step, a classification of the recognized tree based on the tree-specific parameter into at least one forestry tree class. The tree class can be, for example, a stability tree class, a vitality class, a quality class or a spatial distribution class. For example, thinning of the tree population can be carried out in a more targeted manner and is therefore more environmentally friendly.

According to the previous embodiment, the step of controlling the working device can be performed based on a classification based on the step of classifying. The classification can be, for example, a classification into stability classes, a classification into vitality classes, a classification into quality classes or a classification into spatial distribution classes. The forestry work task can therefore be carried out in a targeted manner directed towards a tree class, with trees of at least one further tree class being excluded from the work task and thus being preserved.

According to a further embodiment of the method, this can have, as a further step, the definition of a work route for carrying out the work step based on the step of deriving or the derived tree-specific parameter. According to this embodiment, the step of controlling the working device can be performed based on the set working route. The work task or the work step can therefore also include navigating or moving the work machine in the tree population. The work task or work step can also be navigating or moving the work machine in the tree population.

As a further step, the method can therefore include navigating the work machine based on the step of deriving or the derived tree-specific parameter. With the method, trees that are not to be processed can thus also be taken into account when moving or navigating the work machine in the tree population. When a tree is felled, trees that are to be preserved and that are in the vicinity of the tree can therefore be spared. Trees that may pose a hazard when performing the forestry work task or work step can also be recognized and taken into account in the control step.

According to a further embodiment of the method, the work route for carrying out the work step can be defined in such a way that protected areas in the work environment of the work machine are not violated. Based on the derivation step or the derived tree-specific parameters, the protection areas in which the trees to be protected are located can thus be determined in a further step. The stock of trees to be protected can be determined based on the derivation step or the derived tree-specific parameter. The work machine can thus be controlled in such a way that a stand of trees to be protected is not damaged when the work step or the work task is carried out.

According to a further embodiment of the method, this can include, as a further step, determining a position of the work machine in the tree population in order to define the work route. The step of determining the position can be carried out with a positioning sensor, for example a GNSS antenna. The positioning sensor can be arranged on the work machine. The working machine can thus be set up for automating navigation in the stock of trees.

According to a further embodiment of the method, the forestry work task to be performed can include felling the identified tree. The felling may be a future felling. The work step or the work task can therefore also be a work step that prepares the felling or a work task that prepares the felling. Alternatively or in addition to felling, the forestry task or work step to be performed may include documenting the derived tree-specific parameter. According to this embodiment, the step of controlling the working device for felling the identified tree can be carried out depending on the derived tree-specific parameter. Alternatively or additionally, the step of controlling the working device can be performed to document the derived tree-specific parameter.

The documentation can be carried out in a digital forest information system or a digital geographic information system, wherein the work machine or a control device for controlling the work machine or for executing the method can be connected to the forest information system or the digital geographic information system. A degree of digitization of forestry work tasks can thus be increased with the invention and the forestry work task can thus be carried out more efficiently and sustainably.

According to a further embodiment of the method, this can include, as a further step, detecting at least one climatic element acting on the tree population. The climate element can be, for example, a temperature, an air pressure, an air humidity, a wind force, a precipitation value, a dryness value, solar radiation or environmental radiation. The climate element can be derived based on the captured point cloud. According to this embodiment, the step of controlling the working device can be performed based on the detected climate element.

As an alternative or in addition to detecting the climate element, the method can include detecting at least one climate factor as a further step. The climatic element can depend on the recorded at least one climatic factor or can be determined based on this. The climate factor can be, for example, a geographic location, a topographical elevation, an aspect, a distance from a body of water, a groundwater level, a type of soil or a ground cover. The step of controlling the working device can be performed based on the detected climate factor. The step of controlling the working device can also be performed based on the climate element determined based on the detected climate factor.

As an alternative or in addition to the detection of the climatic element, the method can, as a further step, detect at least one anthropogenic factor which, for example, can describe a geographical location of a building or infrastructure. The step of controlling the work device can therefore also be performed based on the detected anthropogenic factor.

According to a further embodiment of the method, this can include, as a further step, carrying out an environmental damage analysis for the work step for performing the forestry work task. The environmental damage analysis can be performed based on the derived tree specific parameter. The environmental damage analysis can show damage to the tree population or the environment, for example the forest floor, occurring when performing the work task, which damage can occur to the tree population or the environment independently of the tree population involved in the work task. According to this embodiment, the step of controlling the working device can be carried out based on the environmental damage analysis that has been carried out. Setting the work route can be done based on the environmental damage analysis. Environmental damage can be reduced or minimized in this way. Further, the work route can be optimized based on the environmental damage analysis.

In a further aspect, the present invention relates to a forestry work system. The forestry work system can be set up to carry out the method according to the preceding aspect. The forestry work system can be set up to automatically carry out the method according to the previous aspect.

The forestry work system has a laser scanner for capturing a point cloud of a stand of trees. The forestry work system also has an evaluation unit for detecting a tree in the detected point cloud and for deriving a tree-specific parameter of the detected tree. The forestry work system has a tax unit for controlling a working device of the working machine to carry out a working step of the working machine for performing a forestry work task depending on the derived tree-specific parameter. The control unit can have the evaluation unit.

Features and embodiments of one aspect of the present invention may form corresponding features and embodiments of another aspect of the present invention.

character list

• 1 Fig. 12 shows schematically a forestry work system according to an embodiment of the invention.
• 2 shows a flowchart for performing a method for operating a work machine in forestry according to an embodiment of the invention.

Detailed Description of Embodiments

1 shows a forestry work system 200, which is used to carry out the in 2 shown method steps is set up.

The forestry work system 200 has a laser scanner 10 for capturing a point cloud, which at least partially discretely depicts a stock of trees. The stock of trees has a large number of trees 2 . The laser scanner 10 is arranged on a work machine 100 , the laser scanner 10 being arranged above a housing of the work machine 100 according to one embodiment. This results in an advantageous detection range of the laser scanner 10 in which shadowing, for example by the housing of the work machine 100 itself, is minimized.

The forestry work system 200 also has an evaluation unit 30 which is set up to recognize a tree 2 in the recorded point cloud of the tree population. The evaluation unit 30 is set up to derive a tree-specific parameter of the identified tree 2 . The evaluation unit 30 is arranged on the work machine 100 and is connected to the laser scanner 10 for reading in the detected point cloud.

The forestry working system 200 also has a control unit 40 which is set up to control a working device 20 of the working machine 100 . The work device 20 is controlled to perform a work step of the work machine 100 . The work step can be carried out with the work device 20 for performing a forestry work task depending on the derived tree-specific parameter. The working device 20 is arranged on the working machine 100 and connected to the control unit 40 . The control unit 40 has the evaluation unit 30 or is connected to it.

According to one exemplary embodiment, evaluation unit 30 is set up to determine an orientation of a tree trunk as the tree-specific parameter of identified tree 2 . According to this exemplary embodiment, the evaluation unit 30 is also set up to compare the determined alignment of the tree trunk with a predetermined threshold value for the alignment of the tree trunk and, based on this, to output a work signal for felling the tree 2 . According to this exemplary embodiment, the control unit 40 is set up to control a work tool based on the work signal in order to cut down the tree 2 with the work tool in the stand of trees or to exempt it from being felled in the stand of trees. Trees 2 whose tree trunks have an orientation that exceeds the predetermined threshold value are felled, and trees 2 whose tree trunks have an orientation that does not exceed the predetermined threshold value are exempt from being felled. According to this exemplary embodiment, trees 2 are selectively felled in an automated manner.

2 FIG. 1 schematically shows a flow chart for carrying out a method for operating the work machine 100 in forestry. In a first step S1, a point cloud acquisition is carried out. In the first step S1, the point cloud associated with the tree population is recorded with the laser scanner 10. In a second step S2, tree identification is carried out. In the second step S2, a tree 2 of the tree population is recognized by the evaluation unit 30. A parameter derivation is carried out in a third step S3. In the third step S3, the tree-specific parameter of the recognized tree 2 is derived. In a fourth step S4, control is performed. In the fourth step S4, the working device 20 is controlled by the control unit 40 in order to carry out a working step in a fifth step S5. The work step is carried out in order to perform a forestry task with the work device 20 .

Reference List

2

Tree

10

laser scanner

20

work facility

30

evaluation unit

40

control unit

100

working machine

200

working system

S1

point cloud acquisition

S2

tree detection

S3

parameter derivation

S4

steering

S5

work step

Claims (10)

Method for operating a working machine (100) in forestry, with the steps: detecting (S1) a point cloud associated with a stand of trees using a laser scanner (10), detecting (S2) a tree (2) in the detected point cloud, deriving (S3) a tree-specific parameter of the identified tree (2), controlling (S4) a working device (20) of the working machine (100) to carry out a work step (S5) with the working machine (100) for performing a forestry work task depending on the derived tree-specific parameter. procedure after claim 1 , wherein in the step of capturing (S1) the point cloud, the point cloud is captured with a laser scanner (10) arranged on the working machine (100). procedure after claim 1 or 2 , wherein in the step of recognizing (S2) the tree (2) a tree trunk is recognized in the detected point cloud, and wherein in the step of deriving (S3) the tree-specific parameter a parameter of the recognized tree trunk is derived. procedure after claim 3 , wherein in the step of deriving (S3) the tree-specific parameter at least one of a spatial orientation, a trunk diameter and a trunk length of the identified tree trunk is determined. Method according to one of the preceding claims, with the further step of classifying the identified tree (2) based on the tree-specific parameter in at least one forestry tree class, wherein the step of controlling (S4) the working device (20) based on a based on the step of Classification based classification is performed. Method according to one of the preceding claims, with the further step of determining a work route for carrying out the work step based on the step of deriving (S3), wherein the step of controlling (S4) the working device (20) is carried out based on the determined work route. Method according to one of the preceding claims, wherein the forestry work item to be performed comprises felling the identified tree (2), and wherein the step of controlling (S4) the working device (20) to fell the identified tree (2) depending on the derived tree-specific parameter is performed. Method according to one of the preceding claims, with the further step of detecting a climatic element acting on the stock of trees, the step of controlling (S4) the working device (20) being carried out on the basis of the detected climatic element. Method according to one of the preceding claims, with the further step of performing an environmental damage analysis for the working step of performing the forestry work task, wherein the step of controlling (S4) the working device (20) is performed based on the performed environmental damage analysis. Forestry work system (200), which is set up to carry out a method according to one of the preceding claims, with a laser scanner (10) for detecting a cloud of points assigned to a stand of trees, an evaluation unit (30) for recognizing a tree (2) in the cloud of points recorded and for deriving a tree-specific parameter of the recognized tree (2), a control unit (40) for controlling a working device (20) of the working machine (100) for carrying out a work step of the working machine (100) for performing a forestry work task depending on the derived tree-specific parameter.

Images