A farm operation path planning method and its application

Abstract

The present invention belongs to the category of vehicle path planning technology, mainly involving the field of agricultural vehicle path planning, and in particular, to a farm operation path planning method and application. The present invention performs spatial analysis based on the pre-mapped plot boundaries, target strip obstacle boundaries, and non-target obstacle boundaries. According to the strip-shaped geometric characteristics of the target strip obstacle, its long side is extracted as the boundary to be operated, and the outer areas of the two short sides are used as turning areas. Starting from the starting point, a global path is generated in the specified direction that traverses the target strip obstacle boundary to be operated and avoids non-target obstacles, until the last target strip obstacle in the specified direction is traversed. The present invention makes full use of the strip-shaped geometric characteristics of the target strip obstacle, reduces the redundancy rate and omission rate of the operation path, improves the quality and efficiency of unmanned operations on the farm, and at the same time, the use of a pure geometric algorithm also greatly improves the efficiency of the algorithm, and can quickly generate a global operation path that traverses each specified target strip obstacle.

Description

Technical Field

[0001] The present invention belongs to vehicle path planning technology, mainly relates to the field of agricultural vehicle path planning, and in particular to a farm operation path planning method and application. Background Art

[0002] Farm operation path planning is a key component of unmanned operations, especially on farms. In an unmanned farm operation system, a global operation path must be generated based on farm information. This serves as input to the control layer and guides the operation of the tracked vehicles. Generating a well-formed and comprehensive operation path for the task in complex farm environments is a crucial prerequisite for ensuring the quality and efficiency of unmanned farm operations.

[0003] Currently, most existing farm operation path planning methods focus on aerial 3D path planning for drones, while path planning for ground vehicles is relatively limited. For farm ground vehicle path planning, there are two main approaches: local obstacle avoidance path planning based on environmental perception, and global path planning based on raster analysis. Local planning methods based on environmental perception primarily use lidar or cameras to acquire local environmental information around the vehicle. These methods then extract traversable areas through point cloud analysis or image segmentation, generating a local path within this traversable area. While these methods do not require prior mapping of the farm environment, they do require the vehicle to be equipped with additional sensors and more powerful computing units, resulting in high costs. Furthermore, these local planning methods fail to consider global plot boundaries and the distribution of target obstacles, leading to the generation of redundant paths and compromised farm operation quality and efficiency.

[0004] Different from the local planning methods mentioned above, the global farm operation path planning method based on grid analysis requires mapping the farm in advance to obtain the boundaries of the plots and the boundaries of the target strip obstacles, and then converting them into a grid map. Based on the grid map, search algorithms such as A* and ant colony are used to obtain the optimal obstacle avoidance operation path between the given starting point and the end point. Although this type of method takes into account the global information of the farm and does not produce redundant paths, it is prone to omissions because it simplifies the operation path into an obstacle avoidance path between two points. As a result, some operation targets (such as fruit trees) that are not on the optimal obstacle avoidance path are omitted. In addition, the use of heuristic search algorithms also makes this method less efficient. Summary of the Invention

[0005] In order to solve the problems that existing farm operation path planning methods are prone to redundancy or omissions, and have high hardware requirements leading to increased costs, the present invention provides a geometric algorithm based on spatial analysis to generate a global farm operation path, specifically a farm operation path planning method and application.

[0006] The present invention provides the following technical solutions:

[0007] A farm operation path planning method comprises the following steps:

[0008] S100: Farm information acquisition: The farm information includes the land boundary, the target strip obstacle boundary and the non-target obstacle boundary; wherein, bd orchard is the plot boundary, is the boundary of n target strip obstacles, are m non-target obstacle boundaries;

[0009] S200: generating a job strip: specifically including:

[0010] S201: Target strip obstacle boundary segmentation: for Divide it into four parts Represent the two long boundaries and two short boundaries of the target strip obstacle respectively. The specific segmentation method is to calculate in sequence The slope k of the line segment formed by each two adjacent points in j , then the slope change will have the first four jumps, these jump points are the dividing points between the long side and the short side of the long target strip obstacle polygon. At these four jump points, The closed loop is broken into four broken line segments to complete the segmentation;

[0011] S202: Operation baseline generation: First obtain the first target obstacle strip The outer long side The minimum distance between the path of the given working vehicle and the target strip obstacle during normal operation is d min , then Outward deviation d min Get the first job baseline For the middle area with target strip obstacles on both sides, get each two adjacent groups of target strip obstacles The adjacent long side and And calculate the medians of these two sets of long sides The specific method is: 1) Get as well as The two outermost endpoints that are longer along the extension line of the long side 2) There are k1+1 long sides, If there are k2-k1 long sides, the left and right pairing will generate max(k1+1,k2-k1) non-intersecting midlines, where the outer endpoints of the two outermost midlines are 3) Connect all the midlines in order to get Finally get the last target strip obstacle The outer long side Offset it to the outside d min Get the last job baseline

[0012] S203: Operation baseline obstacle avoidance processing: first set the non-target obstacle boundary d min Expand outward and recycle to determine the operation baseline After the expansion Do the polygons intersect? If they do intersect but are not tangent, calculate the intersection point of the two polygons and get The shorter side between the two intersection points, and Fusion obtains the final obstacle avoidance operation strip

[0013] S300: Working strip connection: Connect each strip based on the reciprocating law Generate a turn-around route.

[0014] Preferably, in the above S300, the turning path is any one of a straight line connection path, an arc turning path or a light bulb turning path.

[0015] Preferably, in the above S300, e) for adjacent working strips and Target strip obstacles sandwiched between strips Both ends of are connectable regions, and the initial connection region is determined according to the specified clockwise / counterclockwise direction combined with the strip sorting;

[0016] f) Determine whether there is any inconsistency in the initial connection area and If there is a collision line, calculate the line and and The straight line formed by the two intersection points is the connecting line

[0017] g) If there is no collision-free straight line in the initial connection area, change the initial specified direction from clockwise / counterclockwise to counterclockwise / clockwise, select another connectable area, and repeat step f);

[0018] h) If there is no collision-free line between the two connectable areas, the path planning is completed.

[0019] Preferably, in S100, the farm information is preprocessed, and the preprocessing step specifically includes:

[0020] a) Check the input bd orchard 、 as well as Check whether the starting point and end point of each boundary array are the same. If they are not the same, add the starting point coordinates to the end of the boundary array to form a closed loop;

[0021] b) When bd orchard 、 as well as After all the boundaries in are closed, determine whether each closed boundary has self-intersection. If so, remove the small closed loops formed by self-intersection, and the remaining large closed loop coordinate set is used as the final boundary array;

[0022] c) Loop check bd orchard and The geometric relationship of The polygon is completely in bd orchard If the polygon is inside, it is normal; if with bd orchard If there is an intersection, then according to Cutting bd orchard , remove with bd orchard Overlay part to form a new plot boundary and delete the

[0023] d) Loop check bd orchard and The geometric relationship between bd orchard The polygon contains The polygon formed is normal; if with bd orchard If there is intersection, cut The polygons formed only retain those in bd orchard The part inside the polygon.

[0024] An application of a farm operation path planning method, running a farm operation path planning method as described above, for farm operation path planning.

[0025] Preferably, the farm operation is a farm operation.

[0026] An automatic driving agricultural machine, comprising:

[0027] Memory for storing computer programs;

[0028] A processor is used to implement a farm operation path planning method as described in any of the above items when executing the computer program.

[0029] A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements a farm operation path planning method as described in any one of the above items.

[0030] Compared with the existing technology, the beneficial effects of the present invention are as follows: the present invention performs spatial analysis based on the pre-mapped plot boundaries, target strip obstacle boundaries, and non-target obstacle boundaries. Based on the strip-shaped geometric characteristics of the target strip obstacle, its long side is extracted as the boundary to be operated, and the outer areas of the two short sides are used as turning areas. Starting from the starting point, a global path is generated in the specified direction that traverses the target strip obstacle boundary to be operated and avoids non-target obstacles, until the last target strip obstacle in the specified direction is traversed. The present invention fully utilizes the strip-shaped geometric characteristics of the target strip obstacle, reduces the redundancy rate and omission rate of the operation path, and improves the quality and efficiency of unmanned operations on the farm. At the same time, the use of a pure geometric algorithm also greatly improves the efficiency of the algorithm, and can quickly generate a global operation path that traverses each specified target strip obstacle. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0032] Figure 1 is a schematic diagram of a flow chart of an embodiment of the present invention;

[0033] Figure 2 2 is a schematic diagram of trajectory generation according to an embodiment of the present invention. DETAILED DESCRIPTION

[0034] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0035] like Figures 1 and 2 As shown:

[0036] The present invention primarily encompasses three key components: farm information preprocessing, work strip generation, and work strip connection. Farm information includes plot boundaries, target obstacle strip boundaries, and non-target obstacle boundaries. Farm information preprocessing primarily involves topological verification and correction of these three types of boundaries. Work strip generation involves target obstacle strip boundary segmentation, baseline generation, and baseline obstacle avoidance. Work strip connection involves generating a U-turn connection line for the work strip outside the target obstacle strip. The target obstacle strip can be, for example, a forest belt.

[0037] 1 Farm information preprocessing

[0038] In the farm information preprocessing part, it is necessary to perform topological correction and intersection processing on the boundaries of the plot, the boundaries of the target strip obstacles and the boundaries of non-target obstacles to reduce the complexity of the input data of the subsequent algorithm, filter out unnecessary intersection relationships in advance, and avoid abnormal results that may appear in subsequent geometric operations. The preprocessing steps specifically include the following points, among which bd orchard is the plot boundary, is the boundary of n target strip obstacles, are m non-target obstacle boundaries:

[0039] a) Check the input bd orchard 、 as well as Check whether the starting point and end point of each boundary array are the same. If they are not the same, add the starting point coordinates to the end of the boundary array to form a closed loop;

[0040] b) When bd orchard 、 as well as After all the boundaries in are closed, determine whether each closed boundary has self-intersection. If so, remove the small closed loops formed by self-intersection, and the remaining large closed loop coordinate set is used as the final boundary array;

[0041] c) Loop check bd orchard and The geometric relationship of The polygon is completely in bd orchard If the polygon is inside, it is normal; if with bd orchard If there is an intersection, then according to Cutting bd orchard , remove with bd orchard Overlay part to form a new plot boundary and delete the

[0042] d) Loop check bd orchard and The geometric relationship between bd orchard The polygon contains The polygon formed is normal; if with bd orchard If there is intersection, cut The polygons formed only retain those in bd orchard The part inside the polygon.

[0043] 2Job strip generation

[0044] BD orchard 、 as well as After preprocessing, work strip generation begins. On a farm, target strip obstacles are distributed as groups of approximately parallel long polygons, with the long sides of these polygons representing the boundaries to be worked on. Therefore, in addition to generating work strips along the long sides of the first and last target strip obstacles, work strips are primarily generated between two adjacent target strip obstacles, simultaneously working on both sides. Work strip generation specifically involves three steps: target strip obstacle boundary segmentation, work baseline generation, and baseline obstacle avoidance processing.

[0045] In the target strip obstacle boundary segmentation part, It needs to be divided into four parts Represent the two long boundaries and two short boundaries of the target strip obstacle respectively. The specific segmentation method is to calculate in sequence The slope k of the line segment formed by each two adjacent points in j , then the slope change will have the first four jumps, these jump points are the dividing points between the long side and the short side of the long target strip obstacle polygon. At these four jump points, The closed loop is broken into four broken line segments, and the segmentation can be completed by combining length judgment, etc. The length judgment is to determine which two are long sides and which two are short sides.

[0046] In the baseline generation part of the job, first obtain the first target obstacle strip The outer long side The minimum distance between the path of the given working vehicle and the target strip obstacle during normal operation is d min , then Outward deviation d min Get the first job baseline For the middle area with target strip obstacles on both sides, get each two adjacent groups of target strip obstacles The adjacent long side and And calculate the medians of these two sets of long sides The specific method is: 1) Get as well as The two outermost endpoints that are longer along the extension line of the long side 2) There are k1+1 long sides, If there are k2-k1 long sides, the left and right pairing will generate max(k1+1,k2-k1) non-intersecting midlines, where the outer endpoints of the two outermost midlines are 3) Connect all the midlines in order to get Finally get the last target strip obstacle The outer long side Offset it to the outside d min Get the last job baseline

[0047] Finally, after all the operation baselines are generated, baseline obstacle avoidance processing is required. The specific processing process is to first remove the non-target obstacle boundary d min Expand outward and recycle to determine the operation baseline After the expansion Do the polygons intersect? If they do intersect but are not tangent, calculate the intersection point of the two polygons and get The shorter side between the two intersection points, and Fusion obtains the final obstacle avoidance operation strip

[0048] 3Job strip connection

[0049] In the job stripe that generates the boot job After that, you need to connect each Generate a turning path. Since the present invention is aimed at working crawler vehicles, a straight line connection path can be directly generated. If it is other types of working vehicles, such as those that conform to the Ackerman steering model, arc turning or light bulb turning paths can also be used for connection. In the process, in addition to the basic reciprocating law, it is also necessary to consider and bd orchard and The influence of the relative position relationship on the connection order and stripe traversal direction is as follows:

[0050] e) For adjacent working strips and Target strip obstacles sandwiched between strips Both ends of are connectable regions, and the initial connection region is determined according to the specified clockwise / counterclockwise direction combined with the strip sorting;

[0051] f) Determine whether there is any inconsistency in the initial connection area and bd orchard If there is a collision line, calculate the line and and The straight line formed by the two intersection points is the connecting line

[0052] g) If there is no collision-free straight line in the initial connection area, change the initial specified direction from clockwise / counterclockwise to counterclockwise / clockwise, select another connectable area, and repeat step f);

[0053] h) If there is no collision-free line between the two connectable areas, the path planning is completed.

[0054] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0055] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art will be able to modify the technical solutions described in the aforementioned embodiments or substitute equivalents for some of the technical features. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A farm operation path planning method, characterized in that: The following steps are involved: S100: Farm information acquisition: The farm information includes the land boundary, the target strip obstacle boundary and the non-target obstacle boundary; wherein, bd orchard is the plot boundary, is the boundary of n target strip obstacles, are m non-target obstacle boundaries; S200: generating a job strip: specifically including: S201: Target strip obstacle boundary segmentation: for Divide it into four parts Represent the two long boundaries and two short boundaries of the target strip obstacle respectively. The specific segmentation method is to calculate in sequence The slope k of the line segment formed by each two adjacent points in j , then the slope change will have the first four jumps, these jump points are the dividing points between the long side and the short side of the long target strip obstacle polygon. At these four jump points, The closed loop is broken into four broken line segments to complete the segmentation; S202: Operation baseline generation: First obtain the first target obstacle strip The outer long side The minimum distance between the path of the given working vehicle and the target strip obstacle during normal operation is d min , then Outward deviation d min Get the first job baseline For the middle area with target strip obstacles on both sides, get each two adjacent groups of target strip obstacles The adjacent long side and And calculate the medians of these two sets of long sides The specific method is: 1) Get as well as The two outermost endpoints that are longer along the extension line of the long side 2) There are k1+1 long sides, If there are k2-k1 long sides, the left and right pairing will generate max(k1+1,k2-k1) non-intersecting midlines, where the outer endpoints of the two outermost midlines are 3) Connect all the midlines in order to get Finally get the last target strip obstacle The outer long side Offset it to the outside d min Get the last job baseline S203: Operation baseline obstacle avoidance processing: first set the non-target obstacle boundary d min Expand outward and recycle to determine the operation baseline After the expansion Do the polygons intersect? If they do intersect but are not tangent, calculate the intersection point of the two polygons and get The shorter side between the two intersection points, and Fusion obtains the final obstacle avoidance operation strip S300: Working strip connection: Connect each strip based on the reciprocating law Generate a turn-around route.

2. A farm operation path planning method according to claim 1, characterized in that: In the above-mentioned S300, the turning path is any one of a straight line connection path, an arc turning path or a light bulb turning path.

3. A farm operation path planning method according to claim 1, characterized in that: In the above-mentioned S300, e) For adjacent working strips and Target strip obstacles sandwiched between strips Both ends of the are connectable areas, according to The initial connection area is determined by combining the strip sorting in the specified clockwise / counterclockwise direction; f) Determine whether there is any inconsistency in the initial connection area and bd orchard If there is a collision line, calculate the line and and The straight line formed by the two intersection points is the connecting line g) If there is no collision-free straight line in the initial connection area, change the initial specified direction from clockwise / counterclockwise to counterclockwise / clockwise, select another connectable area, and repeat step f); h) If there is no collision-free line between the two connectable areas, the path planning is completed.

4. A farm operation path planning method according to claim 1, characterized in that: In the aforementioned S100, the farm information is preprocessed, and the preprocessing steps specifically include: a) Check the input bd orchard 、 as well as Check whether the starting point and end point of each boundary array are the same. If they are not the same, add the starting point coordinates to the end of the boundary array to form a closed loop; b) When bd orchard 、 as well as After all the boundaries in are closed, determine whether each closed boundary has self-intersection. If so, remove the small closed loops formed by self-intersection, and the remaining large closed loop coordinate set is used as the final boundary array; c) Loop check bd orchard and The geometric relationship of The polygon is completely in bd orchard If the polygon is inside, it is normal; if with bd orchard If there is an intersection, then according to Cutting bd orchard , remove with bd orchard Overlay part to form a new plot boundary and delete the d) Loop check bd orchard and The geometric relationship between bd orchard The polygon contains The polygon formed is normal; if with bd orchard If there is intersection, cut The polygons formed only retain those in bd orchard The part inside the polygon.

5. An application of a farm operation path planning method, characterized in that: A farm operation path planning method as described in any one of claims 1 to 4 is run for farm operation path planning.

6. Application of a farm operation path planning method according to claim 5, characterized in that: The farm operation described is orchard plant protection.

7. An automatic driving agricultural machine, characterized in that: include: Memory for storing computer programs; A processor, configured to implement a farm operation path planning method as described in any one of claims 1 to 4 when executing the computer program.

8. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, it implements a farm operation path planning method as described in any one of claims 1 to 4.

Images