Dynamic full-coverage path planning method and device, cleaning equipment, storage medium

Abstract

The present invention discloses a dynamic full-coverage path planning method and device. The method includes generating a first cleaning path according to the boundary of the area to be cleaned, and performing edge cleaning on the area to be cleaned according to the first cleaning path; splitting a first sub-region that meets the first preset condition from the area surrounded by the first cleaning path, performing full-coverage path planning on the split first sub-region, and performing a cleaning task on the first sub-region according to the full-coverage path planning result; determining a second sub-region in the area surrounded by the first cleaning path, performing local path planning on the second sub-region based on a first dynamic layer updated in real time, and performing a cleaning task on the second sub-region according to the local path planning result. The solution of the present invention realizes intelligent and dynamic zoning and dynamic local planning of the area to be cleaned, so that the cleaning task can be efficiently and dynamically completed, better adapting to the actual situation of dynamic environmental changes and greatly improving the cleaning efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of autonomous driving, and particularly to a dynamic full-coverage path planning method, a dynamic full-coverage path planning device, a cleaning device, and a storage medium. Background Art

[0002] In recent years, with the development of driverless technology in the cleaning field, we increasingly see some unmanned vehicles freely performing cleaning tasks in large squares, supermarkets, basements and other scenarios, greatly reducing the burden on cleaning workers. For this cleaning method, how to achieve full-coverage tasks in a specified scenario is a crucial technical problem that needs to be overcome. At present, to achieve full-coverage tasks in a specified scenario, it is mainly achieved by performing full-coverage path planning for the entire target working area to plan a working path that covers the entire target working area, with no gaps between the working paths, and avoiding obstacles in the working area at the same time. Among them, the existing full-coverage path planning methods mainly directly perform unit decomposition on the basis of an existing global map, decompose the target working area into non-overlapping and obstacle-free sub-units, then perform one-time path planning within each sub-unit respectively, and finally connect the planned paths between the sub-units in a certain connection order to obtain a working path that covers the entire target working area. However, in real scenarios, the environmental conditions are often dynamically changing. Therefore, when using the one-time generated working path, in the face of a complex dynamic environment, it is often easy to cause missed sweeping. In response to this, the existing methods mainly perform unified supplementary sweeping on the missed sweeping areas of the entire target working area after completing the cleaning task of the entire area to avoid missed sweeping and achieve full coverage.

[0003] However, the one-time unit decomposition based on the global map in the prior art often fails to apply to the actual scenario due to the movement or removal of obstacles in the target working area. In this case, the one-time planned working path may become completely invalid and unable to perform the cleaning task. Moreover, due to special situations such as unevenness generated by teaching at the edge of the target working area and irregular obstacles in the working area, the unit decomposition based on the global map is also prone to generate many fragmented areas. Therefore, using the one-time planning method, it is easy to miss sweeping these fragmented areas, affecting the coverage rate. In addition, since the working path generated by the one-time planning method cannot well adapt to the dynamically changing environmental conditions, there will be relatively many missed sweeping areas. If these missed sweeping areas are all processed in the unified supplementary sweeping stage, it will not only lead to a reduction in the full-coverage cleaning efficiency, but also seriously reduce the efficiency of the unified supplementary sweeping. Summary of the Invention

[0004] An embodiment of the present invention provides a dynamic full-coverage path planning solution to solve the problems in the prior art that the unit decomposition method and the one-time path planning method cannot adapt to the dynamic changes of the environment, resulting in overly fragmented unit partitions, easy omission of sweeping, and low cleaning efficiency.

[0005] In a first aspect, an embodiment of the present invention provides a dynamic full-coverage path planning method, which includes

[0006] Generating a first cleaning path according to the boundary of the area to be cleaned, and performing edge cleaning on the area to be cleaned according to the first cleaning path;

[0007] Dividing a first sub-area that meets the first preset condition from the area surrounded by the first cleaning path, performing full-coverage path planning on the divided first sub-area, and performing a cleaning task on the first sub-area according to the full-coverage path planning result;

[0008] Determining a second sub-area in the area surrounded by the first cleaning path, performing local path planning on the second sub-area based on a first dynamically updated layer, and performing a cleaning task on the second sub-area according to the local path planning result.

[0009] In a second aspect, an embodiment of the present invention provides a dynamic full-coverage path planning device, which includes a memory for storing executable instructions; and

[0010] A processor for executing the executable instructions stored in the memory, and when the executable instructions are executed by the processor, the processor executes the dynamic full-coverage path planning method as described in the first aspect.

[0011] In a third aspect, an embodiment of the present invention provides a dynamic full-coverage path planning device, which includes:

[0012] A first planning module for generating a first cleaning path according to the boundary of the area to be cleaned, and performing edge cleaning on the area to be cleaned according to the first cleaning path;

[0013] A second planning module for dividing a first sub-area that meets the first preset condition from the area surrounded by the first cleaning path, performing full-coverage path planning on the divided first sub-area, and performing a cleaning task on the first sub-area according to the full-coverage path planning result;

[0014] A third planning module for determining a second sub-area in the area surrounded by the first cleaning path, performing local path planning on the second sub-area based on a first dynamically updated layer, and performing a cleaning task on the second sub-area according to the local path planning result.

[0015] In a fourth aspect, an embodiment of the present invention provides a cleaning device, which includes:

[0016] a fuselage; and

[0017] the above-mentioned dynamic full-coverage path planning device is provided on the fuselage.

[0018] In a fifth aspect, the present invention provides a storage medium, on which a computer program is stored, and when the program is executed by a processor, the steps of the above method are implemented.

[0019] In a sixth aspect, the present invention provides a computer program product containing instructions, and when the computer program product runs on a computer, the computer is enabled to execute the above-mentioned dynamic full-coverage path planning method.

[0020] The beneficial effects of the embodiments of the present invention are as follows: The method provided by the embodiments of the present invention does not perform unit partitioning on the target area at the first moment. Instead, it first uses the area boundary to perform the first cleaning path planning to complete the edge cleaning. After the edge cleaning, according to the actual situation of the area enclosed within the boundary area, the area enclosed by the first cleaning path is intelligently and dynamically partitioned, and for different types of partitions, different planning strategies are adopted to execute the cleaning task. Thus, it can not only perform efficient full-coverage path planning for the first sub-area that meets the first preset condition, but also perform dynamic local path planning for the determined second sub-area according to the real-time environment situation and the driving path, so that the cleaning task can be completed efficiently and dynamically to better adapt to the actual situation of the dynamic change of the environment. Moreover, the solution of the embodiments of the present invention can also achieve timely supplementary cleaning of the missed cleaning area or the area where the obstacle is removed, and real-time obstacle avoidance for the area where the obstacle is added by performing local path planning on the determined second sub-area, greatly improving the cleaning efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0022] Figure 1 is a flowchart of the dynamic full-coverage path planning method according to an embodiment of the present invention;

[0023] Figure 2 is a flowchart of the method for generating the first cleaning path according to the boundary of the area to be cleaned according to an embodiment of the present invention;

[0024] Figure 3 schematically shows a comparison effect display diagram of the first cleaning paths formed in different ways;

[0025] Figure 4 Schematically shows an effect display diagram of determining an effective uncleaned area through the combination of a first dynamic layer and a second dynamic layer, where, Figure 4 -a represents the global map of the target area, Figure 4 -b represents the second dynamic layer saved according to the perception information during the previous autonomous driving, Figure 4 -c represents the first dynamic layer formed according to the perception information detected based on the current position of the vehicle itself;

[0026] Figure 5 Schematically shows the display effect diagram of the coverage status map of the embodiment of the present invention;

[0027] Figure 6 Schematically shows the method flow diagram of splitting a first sub-region that meets the first preset condition from the area surrounded by the first cleaning path in an embodiment of the present invention;

[0028] Figure 7 Is schematically shown by using Figure 6 The method shown for the effect display diagram of dynamically partitioning to split out the first sub-region;

[0029] Figure 8 Schematically shows the method flow diagram of performing full-coverage path planning on the split first sub-region in an embodiment;

[0030] Figure 9 Schematically shows the display effect diagram of the full-coverage path directly formed by "bow"-shaped filling;

[0031] Figure 10 Schematically shows the method flow diagram of determining the optimal connection order between "bow"-shaped reference paths in an embodiment;

[0032] Figure 11 Schematically shows the use of Figure 8 and 10 The method shown for the Figure 9 The optimized planned path effect diagram of the "bow"-shaped reference path shown;

[0033] Figure 12 Schematically shows the method flow diagram of performing local path planning on the second sub-region during the execution of the cleaning task in an embodiment;

[0034] Figure 13 Schematically shows the display effect diagram of performing local path planning through edge exploration on the encountered obstacles to achieve dynamic obstacle avoidance in a certain situation;

[0035] Figure 14Schematically shows a display effect diagram of local path planning for avoiding an encountered obstacle through side exploration to achieve dynamic obstacle avoidance in another scenario;

[0036] Figure 15 Schematically shows a method flow diagram of optimizing a second reference path using biased sampling to determine a second reference path that can be used as the basis for actual driving;

[0037] Figure 16 Schematically shows Figure 15 The display effect diagram of the influence of the method on the biased screening of immediate supplementary scanning;

[0038] Figure 17 Schematically shows Figure 15 The display effect diagram of the influence of the method on the biased screening of preferentially selecting uncleaned areas for obstacle avoidance;

[0039] Figure 18 Schematically shows Figure 15 The display effect diagram of the influence of the method on the biased screening of path smoothness;

[0040] Figure 19 Schematically shows the display effect diagram of dynamically determining the effective cleaning area, where Figure 19 A is the global map corresponding to the target area, Figure 19 B is the second dynamic map updated in real time, Figure 19 C is the coverage status map after cleaning is completed, Figure 19 D is the area that needs to be supplemented and scanned after being corrected by combining the second dynamic map;

[0041] Figure 20 Schematically shows the principle block diagram of the dynamic full-coverage path planning device according to an embodiment of the present invention;

[0042] Figure 21 Schematically shows the principle block diagram of the dynamic full-coverage path planning device according to another embodiment of the present invention;

[0043] Figure 22 Is the principle block diagram of the dynamic full-coverage path planning device according to still another embodiment of the present invention;

[0044] Figure 23 Is the principle block diagram of the dynamic full-coverage path planning device according to yet another embodiment of the present invention;

[0045] Figure 24 Is the principle block diagram of the dynamic full-coverage path planning device according to yet another embodiment of the present invention;

[0046] Figure 25 Is the principle block diagram of the cleaning device according to an embodiment of the present invention;

[0047] Figure 26 This is a schematic structural diagram of an embodiment of the electronic device of the present invention. Detailed implementation manners

[0048] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are some but not all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0049] It should be noted that, without conflict, the embodiments in the present application and the features in the embodiments may be combined with each other.

[0050] The present invention may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. The present invention may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including storage devices.

[0051] In the present invention, "module", "device", "system", etc. refer to related entities applied to a computer, such as hardware, a combination of hardware and software, software, or software in execution. Specifically, for example, an element may be, but is not limited to, a process running on a processor, a processor, an object, an executable element, an execution thread, a program, and / or a computer. Further, an application program or a script program running on a server, and the server may both be elements. One or more elements may be in a process and / or thread in execution, and the elements may be localized on one computer and / or distributed between two or more computers, and may be run by various computer-readable media. The elements may also communicate through local and / or remote processes according to signals having one or more data packets, for example, signals from data interacting with another element in a local system, a distributed system, and / or signals interacting with other systems through a network on the Internet.

[0052] Finally, it should also be noted that in this text, relative terms such as first and second are only used 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 "comprising" and "including" not only include those elements, but also other elements not explicitly listed, or elements inherent to such a process, method, article, or device. Without further limitation, the elements defined by the statement "comprising..." do not exclude the existence of additional identical elements in the process, method, article, or device that includes the said elements.

[0053] The dynamic full-coverage path planning method in the embodiments of the present invention can be applied in a dynamic full-coverage path planning device, so that a user can use the dynamic full-coverage path planning device to dynamically execute path planning, improving the flexibility and environmental adaptability of cleaning task execution. These dynamic full-coverage path planning devices include, but are not limited to, smartphones, smart tablets, personal PCs, computers, cloud servers, etc. In particular, the dynamic full-coverage path planning method in the embodiments of the present invention can also be applied in intelligent devices with cleaning functions (or mobile platforms provided with automatic cleaning functions), such as unmanned cleaning vehicles, floor cleaning robots, or autonomous driving cleaning vehicles, etc. The present invention does not make any limitations in this regard.

[0054] Figure 1 Schematically shows a dynamic full-coverage path planning method according to an embodiment of the present invention. This method is applicable to dynamic full-coverage path planning devices such as smartphones, personal computers, cloud servers, etc., so that these devices can achieve intelligent dynamic partitioning and dynamic path planning through the combination of multiple planning strategies, thereby assisting cleaning devices to efficiently execute cleaning tasks according to the real-time changes of the dynamic environment; it is also applicable to cleaning devices that execute cleaning tasks, so that these cleaning devices can better adapt to the changes of the dynamic environment during the execution of cleaning tasks. These cleaning devices can be, for example, unmanned sanitation vehicles, unmanned cleaning vehicles, unmanned floor sweeping vehicles, floor cleaning robots, unmanned floor scrubbers, etc. As Figure 1 shown, the method of the embodiment of the present invention includes:

[0055] Step S10: Generate a first cleaning path according to the boundary of the area to be cleaned, and perform edge cleaning on the area to be cleaned according to the first cleaning path;

[0056] Step S11: Divide a first sub-area that meets the first preset condition from the area surrounded by the first cleaning path, perform full-coverage path planning on the divided first sub-area, and execute cleaning tasks on the first sub-area according to the full-coverage path planning result;

[0057] Step S12: Determine a second sub-region within the area enclosed by the first cleaning path, perform local path planning on the second sub-region based on the real-time updated first dynamic layer, and execute a cleaning task on the second sub-region according to the local path planning result.

[0058] In the embodiments of the present invention, the area to be cleaned can be the target working area that needs to be cleaned. The map boundary can be determined according to the corresponding grid map of this area, and the map boundary is used as the area boundary. For the target working area, its area boundary may be generated by teaching or may be a hand-drawn straight line, but generally there are obstacles on the area boundary. Therefore, for the outermost circle, if the cleaning task is executed based on a one-time planned path, the cleaning efficiency will be seriously affected due to missed sweeping or the need to turn in place, etc. Therefore, in the embodiments of the present invention, when starting the task, the whole area will not be pre-divided into units and a one-time full-coverage path planning will not be performed. Instead, the edge cleaning will be first performed on the target area, that is, the area to be cleaned, through step S10 to form a clear cleaning boundary, and subsequent processing will be based on the cleaning boundary formed by the actual driving to ensure the cleaning efficiency. Specifically, in step S10, the first cleaning path can be generated according to the map boundary corresponding to the area to be cleaned, and the edge cleaning can be performed based on the first cleaning path. Exemplarily, generating the first cleaning path according to the map boundary corresponding to the area to be cleaned can be implemented by shrinking the map boundary of the grid map corresponding to the area to be cleaned by a preset width, and taking the shrunk route as the first cleaning path. Among them, the preset width of shrinking can be, for example, shrinking the map boundary by half of the vehicle width or half of the cleaning width. More preferably, in order to further improve the cleaning efficiency, in the preferred embodiments of the present invention, when generating the first cleaning path according to the map boundary corresponding to the area to be cleaned, not only the boundary of the grid map is shrunk to generate the cleaning path, but also the shrunk path is optimized on the basis of the shrunk generation of the path, so that the finally obtained first cleaning path is both collision-free and smooth and feasible. Figure 2 Schematically shows the method flow of generating the first cleaning path according to the boundary of the area to be cleaned in a preferred embodiment of the present invention, as Figure 2 shown, which can be implemented to include:

[0059] Step S101: Shrink the grid map boundary by a preset width on the grid map corresponding to the area to be cleaned to form a first reference path;

[0060] Step S102: Perform horizontal sampling based on the first reference path to obtain multiple sets of first sampling point sets, where each set of first sampling point sets corresponds to a horizontal direction;

[0061] Step S103: Screen each set of first sampling point sets according to the first screening condition to obtain multiple sets of second sampling point sets;

[0062] Step S104: Select the optimal curve segments from the multiple sampling curves formed by the multiple sets of second sampling point sets and splice them to form the first cleaning path.

[0063] Among them, the preset width in step S101 can be set according to requirements or actual situations, for example, set to half of the vehicle width or half of the width of the cleaning device, etc.

[0064] In step S102, the lateral sampling based on the first reference path can be, for example, to draw a perpendicular line along the first reference path at corresponding positions every preset distance, and take a preset number of points along the perpendicular line direction as a set of sampling points. Thus, by performing lateral sampling every preset distance along the first reference path, multiple sets of first sampling point sets can be formed. Since each set of sampling points is the points collected along the perpendicular line direction by drawing a perpendicular line to the first reference path at the corresponding position, each set of sampling points obtained is in one-to-one correspondence with the lateral direction identified by the corresponding perpendicular line.

[0065] In step S103, the embodiments of the present invention preferably perform a tendency screening on the points in the multiple sets of first sampling point sets obtained, so that the finally selected sampling points are more in line with the desired effect. Among them, the first screening condition can be set according to the actual tendency expectation. Exemplarily, in order to avoid collisions and ensure coverage, the first screening condition can be set to discard the sampling points in the obstacles or outside the working area. Thus, according to the first screening condition, the sampling points in the obstacles or outside the working area in the multiple sets of first sampling point sets can be discarded, so as to obtain multiple sets of second sampling point sets that meet the expectations.

[0066] In the process of forming a sampling curve based on the selected multiple sets of second sampling point sets, by connecting the sampling points between different sets, such as connecting them through a Bezier curve, multiple sets of sampling curves can be formed. In order to finally plan a first cleaning path that meets expectations, such as a cleaning path with good smoothness and high feasibility, in step S104, the multiple sets of sampling curves obtained based on the multiple sets of second sampling point sets will be further screened to select the optimal curve as the final first cleaning path. As a preferred implementation, the method of selecting the optimal curve can be implemented by first performing a collision test on the formed multiple sets of sampling curves and deleting the sampling curves that will cause a collision of the cleaning device model, such as a vehicle model; then evaluating the curvature, length, etc. of the remaining sampling curves. Based on the evaluation results, the greedy algorithm is used to select the optimal curve segment from the curve segments formed by connecting the sampling points between every two adjacent lateral directions, and the optimal curve segments between every two adjacent lateral directions are spliced together, and the splicing result is used as the first cleaning path. Among them, taking the example that the sampling points are connected by a Bezier curve, the evaluation of the curvature, length, etc. of the sampling curve can specifically be to assign weights to the curvature and length of the sampling curve respectively, and sum the product of the curvature and the weight and the product of the length and the weight to be used as the evaluation value of this curve segment. The greedy algorithm is used to select the curve segment with the smallest evaluation value from multiple sets of curve segments for the same adjacent lateral direction as the optimal curve segment in this adjacent lateral direction. Among them, the content about the greedy algorithm can be implemented with reference to the prior art and will not be elaborated here.

[0067] Thus, the embodiment of the present invention can use the above method to use the in-shrinking route as the reference path, and then adopt the idea of dynamic programming to select the optimal curve segments between every two adjacent lateral directions for splicing by performing lateral sampling at regular intervals and evaluating the sampling curves, so as to form the first cleaning path as the result route for performing boundary cleaning. Figure 3 Schematically shows the comparison effect diagram of the first cleaning path formed in the in-shrinking manner and the first cleaning path formed according to the Figure 2 method shown, as Figure 3 shown, compared with the first cleaning path 1 formed only by in-shrinking the boundary and the first cleaning path 2 formed after optimization according to the Figure 2 method shown, the latter path is significantly smoother and more fluent, with better smoothness.

[0068] As a preferred implementation, since the start of the self-cleaning task, two dynamic layers and an overlay status map can be dynamically constructed and maintained for the target area to assist the global grid map corresponding to the target area to achieve real-time dynamic path planning. Among them, the two dynamic layers constructed for the target area can include a first dynamic layer obtained according to real-time perception information and a second dynamic layer recording all static obstacle information in the target area. Thus, as the cleaning device moves and its position is updated in real time, a real-time dynamic local map within a certain range can be obtained at any time through real-time perception information (referred to as the first dynamic layer in the embodiments of the present invention), and based on the real-time dynamic local map, the latest static obstacle information can be obtained and recorded in the dynamically updated second dynamic layer for real-time dynamic path planning. Exemplarily, centered on the vehicle itself, a dynamic layer with a preset range size, such as 7.5m X 7.5m, can be maintained based on real-time perception information and saved in the memory as the first dynamic layer, and the first dynamic layer can be updated in real time according to the position of the vehicle itself; at the same time, a second dynamic layer recording static obstacle information is saved in the form of a temporary file in the map folder. Among them, the second dynamic layer is a global dynamic layer corresponding to the global grid map, and the second dynamic layer can be dynamically updated according to the first dynamic layer or the static obstacle information detected by perception. For example, each time a new perception input is received, through global coordinate conversion, the first dynamic layer or the static obstacle information detected by perception is updated to the second dynamic layer. Thus, in the embodiments of the present invention, when performing the cleaning task, the combination of the dynamically updated second dynamic layer and the global grid map can be used to determine the latest effective uncleaned area; among them, when combining the second dynamic map with the global grid map, the second dynamic layer can be updated in real time again according to the first dynamic layer currently obtained in real time, so as to utilize the real-time nature of the perception information. Through the combination of the two dynamic layers, static obstacles can be updated in a timely manner, such as timely recording newly added obstacles and deleting removed obstacles, to further improve the accuracy and effectiveness of determining the latest effective uncleaned area. Among them, Figure 4 Schematically shows the effect diagram of determining the effective uncleaned area through the combination of the first dynamic layer and the second dynamic layer, such as Figure 4 shown, Figure 4 -a represents the global map of the target area, which records the pre-existing obstacles and passable states in this area. Among them, all the obstacles outside and inside this area are black, and the passable area is white, while Figure 4 -b represents the second dynamic layer saved according to the perception information during the previous automatic driving, which records the newly added obstacle 3 detected in the previous frame of perception information, Figure 4-c represents the first dynamic layer formed based on the perception information detected according to the current position of the vehicle itself, which deletes the static obstacles 4 that have been removed according to the real-time detected perception information. Therefore, according to Figure 4 it can be seen that as the dynamic environment changes, the original global Figure 4 -a records that the effective area to be cleaned is not completely effective. Therefore, it is necessary to combine the dynamically updated first dynamic layer and the second dynamic layer with the global map to determine the effective area to be cleaned. For example, by comparing the first dynamic layer and the second dynamic layer to update the second dynamic layer, and then determining the effective area based on the combination of the second dynamic layer and the global map for full-coverage path planning; and / or by comparing the first dynamic layer and the second dynamic layer to update the second dynamic layer, and then performing local dynamic path planning based on the first dynamic layer during the execution of the cleaning task to achieve real-time obstacle avoidance when new obstacles are added and immediate supplementary cleaning when obstacles are removed. It should be noted that Figure 4 -b and Figure 4 the black boundaries in -c only represent the boundaries of the areas that need to be maintained.

[0069] In addition, according to the execution situation of the cleaning task, such as according to the path trajectory traveled by the cleaning device, a coverage status map can also be dynamically maintained to distinguish and mark the cleaning areas in the target area. Exemplarily, the cleaned areas in the target area can be marked as covered areas and the uncleaned areas can be marked as uncovered areas in the coverage status map, so as to use the coverage status map for dynamic path planning to improve the coverage and cleaning efficiency. Among them, Figure 5 schematically shows the display effect of the coverage status map of the embodiment of the present invention, such as Figure 5 shown, in the coverage status map, by counting the cleaned areas and the uncleaned areas and identifying them respectively, for example, the cleaned areas are marked as covered areas 5 by coloring, and the remaining areas are marked as uncovered areas 6 by white, so as to perform path planning according to the dynamically updated coverage status during the path planning process to avoid missed cleaning and achieve immediate supplementary cleaning.

[0070] Exemplarily, in order to perform efficient cleaning based on the effective uncleaned area, in step S11, the first sub-region that satisfies the first preset condition is segmented from the area surrounded by the first cleaning path, which can be realized based on the dynamically updated second dynamic layer corresponding to the target area and the coverage status map. Thus, the embodiment of the present invention can realize the dynamic partitioning of the target area based on the dynamically updated dynamic layer and the coverage status map. Figure 6 schematically shows the method flow of segmenting the first sub-region that satisfies the first preset condition from the area surrounded by the first cleaning path in an embodiment of the present invention, such as Figure 6As shown, it can be implemented to include:

[0071] Step S111: Determine the uncleaned areas within the area surrounded by the first cleaning path based on the second dynamic layer corresponding to the dynamically updated target area and the coverage status map;

[0072] Step S112: Determine whether the uncleaned areas meet the first preset condition, and use the uncleaned areas that meet the first preset condition as the first sub-areas.

[0073] Thus, after one round of boundary cleaning is performed, according to the uncovered areas marked on the coverage status map and the static obstacle information marked in the second dynamic layer, the first sub-areas that meet the first preset condition can be segmented from the area surrounded by the first cleaning path, so as to realize intelligent and dynamic partitioning of the target area.

[0074] Among them, in step S111, when determining the uncleaned areas within the area surrounded by the first cleaning path based on the second dynamic layer and the coverage status map, in addition to finding the uncovered areas according to the coverage status map, the addition or removal of obstacles will also be determined according to the second dynamic layer, so as to determine the uncleaned areas from the uncovered areas according to the addition or removal of obstacles. Exemplarily, for example, the uncovered areas with added obstacles are regarded as non-cleaning areas to be cleaned, while the uncovered areas with removed obstacles are regarded as newly added cleaning areas to be cleaned, so as to dynamically determine the latest effective uncleaned areas in the uncovered areas. It should be further noted that since the first dynamic layer in the embodiments of the present invention is a local dynamic map, although it has the characteristic of high real-time performance but limited coverage, the embodiments of the present invention preferably use the second dynamic layer to assist the global grid map for dynamic partitioning. During the entire path planning process, the embodiments of the present invention will maintain the real-time comparison relationship between the first dynamic layer and the second dynamic layer. For example, the second dynamic layer will be updated each time perception information is obtained, or the first dynamic layer and the second dynamic layer will be compared each time path planning is performed to determine the latest obstacle information, etc., so as to perform timely dynamic update of the second dynamic layer and ensure the timeliness of the static obstacle information in the second dynamic layer.

[0075] Since the input of the full-coverage path planning algorithm is the global grid map corresponding to the target area, and the global grid map is established in the site adaptation stage, it cannot be updated in time when new obstacles are added or existing obstacles are removed in the site. As a result, at least part of the path planned by the full-coverage planning is invalid. For example, each time a route is planned for the position of the newly added obstacle, when the cleaning device actually reaches that position, it can only pass by avoiding the obstacle, generating regional fragments. Finally, the position still needs to be reswept, resulting in a complex resweeping path. Moreover, for some spaces occupied by newly added static obstacles that do not require resweeping, path planning is still carried out, resulting in the generation of invalid paths and extremely low resweeping efficiency. Therefore, in the embodiments of the present invention, by constructing and maintaining the first dynamic layer and the second dynamic layer, the real-time perception feature can be fully utilized to record the static obstacles inside the target area as an auxiliary to the global grid map, so as to update the obstacle information in a timely manner. Then, based on this, the latest effective area to be cleaned can be determined, and the dynamic partitioning of the target area can be realized based on this. After that, in the embodiments of the present invention, the first sub-area obtained by partitioning can be used as the object of the full-coverage path planning, and the full-coverage path planning is carried out on the first sub-area dynamically segmented based on the effective area to be cleaned, so as to improve the efficiency of the full-coverage path planning and avoid planning meaningless cleaning paths.

[0076] As a preferred implementation manner, the first preset condition can be set according to the characteristics of the full-coverage path planning method adopted and the good adaptability between the first sub-area segmented and the full-coverage path planning method adopted. Exemplarily, taking the "bow" - shaped strategy as the full-coverage path planning method adopted, since the "bow" - shaped full-coverage can be well applied to relatively large and regular uncleaned areas to achieve fast and efficient full-coverage path planning, in this case, the first preset condition can be set to include that the area of the uncleaned area is greater than the first preset value, the length and width of the circumscribed rectangle of the uncleaned area are respectively greater than the second preset value and the third preset value, and the area ratio of the uncleaned area inside its circumscribed rectangle is greater than the fourth preset value. Thus, the first sub-area that meets this first preset condition can be segmented from the area surrounded by the first cleaning path, and the "bow" - shaped strategy can be used to perform full-coverage path planning on this type of first sub-area, so as to achieve the effects of improving the adaptability of the path planning method to the dynamic environment and improving the cleaning efficiency through dynamic partitioning and adaptive path planning. Among them, the first preset value, the second preset value, the third preset value, and the fourth preset value can be set according to requirements or prior experience. Figure 7 Schematically shows the use of Figure 6 the method shown to perform dynamic partitioning to show the effect diagram of the segmented first sub-area, as Figure 7As shown, the obstacles shown in the figure are the result of superimposing the static obstacle information recorded in the second dynamic layer and the global map. And according to the coverage status map, the covered area and the uncovered area in the map can be determined. Figure 7 They are respectively identified by the gray area and the white area in the figure. According to Figure 7 As can be seen from the superimposed effect diagram shown, there is an uncleaned area 7 that meets the first screening condition in the map, that is, the area of the uncleaned area meets the constraints and there are no obstacles inside. Therefore, this uncleaned area can be segmented as the first sub-area to separately use the "bow" - shaped route for full - coverage path planning.

[0077] It can be understood that in other embodiments, the full - coverage path planning method of the loop path can also be used to plan the path for the dynamically segmented first sub - area, and the embodiments of the present invention do not limit this. Although the cleaning path planned by the loop path can meet the turning radius requirements, it always needs to rotate and track. Especially when the area of the obstacle - free area is greater than a certain limit, the covering efficiency of the loop coverage is not as high as that of the "bow" - shaped coverage with more straight lines. Therefore, the embodiments of the present invention preferably determine whether the uncleaned area meets the first screening condition for dividing into a regular geometric area according to the dynamically updated dynamic layer to segment the first sub - area, and preferentially use the "bow" - shaped path covering strategy to perform full - coverage path planning on the segmented first sub - area.

[0078] In another preferred embodiment, the first preset condition can also be custom - set according to user needs and expected goals in combination with the actual situation. For example, the first preset condition can be set according to the convexity and concavity of the actual driving path after the edge - cleaning according to the first cleaning path, so as to realize dynamic zoning based on the convexity and concavity of the actual driving route.

[0079] Taking the full - coverage path planning method for the segmented first sub - area as the "bow" - shaped strategy as an example, Figure 8 Schematically shows the method flow of full - coverage path planning for the segmented first sub - area in an embodiment, as Figure 8 shown, which is implemented to include:

[0080] Step S113: Perform "bow" - shaped filling on the first sub - area to form a "bow" - shaped reference path;

[0081] Step S114: Determine the optimal connection order between the straight - line routes in the "bow" - shaped reference path;

[0082] Step S115: Modify the connection mode between the straight - line routes in the "bow" - shaped reference path according to the determined optimal connection order to form an optimized "bow" - shaped cleaning path as the full - coverage path planning result for the first sub - area.

[0083] by Figure 7 Taking the display effect shown in the figure as an example, the first sub-area separated out can be filled with a "bow" shape to form a "bow" shaped reference path. Figure 9 The schematic diagram shows the display effect of the fully covered path thus formed, namely the “bow”-shaped reference path. Figure 9 As shown, the first sub-area is separated out separately and filled in a "bow" shape to form a "bow"-shaped reference path A covering the entire area.

[0084] After the traditional bow-shaped covering strategy of step S113, Figure 9 As shown, the adjacent straight line routes A1 are directly connected to form a coverage path. In this way, for large cleaning equipment, since the turning radius of the cleaning equipment is limited, it needs to rotate on the spot when switching between adjacent straight line routes, which will seriously affect the cleaning efficiency. Therefore, as a preferred embodiment, after the first sub-area is filled in a "bow" shape, the present invention will further optimize the formed "bow" shaped reference path to improve the smoothness of the final full coverage path planning result and improve the cleaning efficiency of the cleaning equipment. Among them, step S114 and step S115 describe the specific concept of optimizing the formed "bow" shaped reference path, and its main implementation idea is to improve the connection method between "bow" shaped routes, including but not limited to improving the type of connecting lines between straight lines (such as avoiding direct straight line connections), optimizing the connection order, etc. The connection order refers to the connection sequence between the straight lines in the planned "bow"-shaped route. For example, the parameter n can be set to represent the connection order, such as when n=1, it means that adjacent straight lines are connected to each other; when n=2, it means that there is a straight line between every two connected straight lines; and so on. Therefore, the embodiment of the present invention can optimize the "bow"-shaped reference route by finding the optimal solution for the connection order.

[0085] As an implementation method, the optimal connection order can be adaptively calculated based on the circumscribed rectangle area and cleaning width of the first sub-region and the evaluation function. Taking the connection order represented by parameter n as an example, the purpose of calculating the optimal connection order is to find the optimal n so that the cost of the cleaning path finally planned is the lowest. Figure 10 The method flow of determining the optimal connection order between the “bow”-shaped reference paths according to one embodiment is schematically shown. Figure 10 As shown, its implementation includes:

[0086] Step S1141: Determine all possible connection sequences between the straight lines in the “bow”-shaped reference path;

[0087] Step S1142: Evaluate all possible connection orders according to a preset evaluation function, and filter out the optimal connection order among the straight-line routes based on the evaluation results.

[0088] Among them, all possible connection orders among the straight-line routes in the "bow"-shaped reference path refer to all possible sorting methods for connecting the straight-line routes in the "bow"-shaped reference path. Since these straight-line routes can be connected in adjacent order to form a coverage path, or can be connected in an order with a preset number of straight-line routes spaced apart to form a coverage path, such as spaced apart by one or two, etc. Therefore, in the embodiments of the present invention, in order to filter out the optimal connection order therefrom, all possible connection orders among the straight-line routes in the "bow"-shaped reference path will be determined according to possible permutation and combination methods, that is, all possible values of the parameter n will be determined. Then, in step S1142, all possible connection orders will be evaluated to filter out the optimal connection order based on the evaluation results, that is, the optimal solution of the parameter n will be adaptively determined through evaluation.

[0089] As a preferred embodiment, the evaluation function can be set to be related to the connection method among the straight-line routes in the "bow"-shaped reference path. Taking the connection method including the connection line type and connection order between the straight-line routes in the "bow"-shaped route, where the connection line type is a Bezier curve and the evaluation function is used to evaluate the total cost of the connection order as an example, the preset evaluation function can be determined by the sum of the lengths of all straight-line routes in the "bow"-shaped reference path, the sum of the curvatures of the Bezier curves used for connection between each straight-line route according to the current connection order, and the total length of the Bezier curves used to connect all straight-line routes according to the current connection order. Exemplarily, the evaluation function can be specifically set to be represented by the following formula:

[0090] Cost = k len *C len +k curcature *C curvature +K b_len *C b_len

[0091] Among them, Cost is used to represent the total cost of the cleaning path formed according to the corresponding connection order, C len is used to represent the sum of the lengths of all straight-line routes in the "bow"-shaped reference path, C curvature is used to represent the sum of the curvatures of the Bezier curves used for connection between each straight-line route in the "bow"-shaped reference path, C b_len is used to represent the total length of the Bezier curves connecting all straight-line routes, k len 、k curvature and k b_lenUsed to represent the weight values assigned to each evaluation item. In practical applications, corresponding weights can be assigned to each evaluation item in the evaluation function according to requirements to adjust the proportion of each evaluation item in the total cost. In addition, it should be noted that for the case where adjacent straight-line routes need to be directly connected or a U-turn in place is required, a reasonable cost value can be customized according to the actual situation as the value of the curvature of the Bezier curve connecting the corresponding straight-line routes for this index item.

[0092] Thus, all possible connection orders are evaluated through this evaluation function, and the parameter n that minimizes the total cost, i.e., Cost, is determined from them, so as to screen out the optimal connection order. Then, according to the determined optimal connection order, the straight-line routes in the bow-shaped reference path are reconnected using Bezier curves to form an optimized bow-shaped cleaning path. Figure 11 Schematically shows the use of Figure 8 and 10 The method shown in Figure 9 to optimize the bow-shaped reference path shown in Figure 11 That is, the effect diagram of the planning result obtained by using the above evaluation function to find the optimal connection order and connecting the straight-line routes according to the optimal connection order, as shown in

[0093] In other embodiments, the optimal connection order between the straight-line routes in the bow-shaped reference path can also be obtained by receiving user-defined configurations. In another embodiment, it can also be to allow the user to define all possible connection orders between the straight-line routes through user-defined configurations, and use the above evaluation function to evaluate the received user-defined connection orders to determine the optimal connection order. More preferably, in other embodiments, the above evaluation function can also be customized by the user according to user requirements and expected goals.

[0094] As a preferred embodiment, after dynamically dividing the first sub-region, the present invention further determines a second sub-region in the region surrounded by the first cleaning path according to the division result of the first sub-region. Specifically, in step S112, according to the judgment result of whether the uncleaned region in the region surrounded by the first cleaning path meets the first preset condition, the uncleaned regions that do not meet the first preset condition are all used as the second sub-region.

[0095] In practical applications, there may be more than one first sub-region in the area surrounded by the first cleaning path. In this case, the embodiments of the present invention preferably perform full-coverage path planning and cleaning tasks in descending order according to the area sizes of the first sub-regions that can be segmented. More preferably, the processing of the second sub-region is performed after corresponding cleaning tasks are executed for all the segmented first sub-regions.

[0096] Since the segmented second sub-regions are all irregular or small-area uncovered regions that do not meet the first screening condition, and may also contain obstacles, and the obstacles may change in real time in the actual scenario, therefore, the embodiments of the present invention preferably perform local path planning for the second sub-region based on the first dynamic layer updated in real time when performing the cleaning task for the second sub-region, so as to perform dynamic planning and dynamic cleaning for the second sub-region in combination with the local path planning result, so as to effectively avoid obstacles in real time or achieve immediate supplementary cleaning. Figure 12 Schematically shows a method flow of performing local path planning for the second sub-region during the execution of the cleaning task, as Figure 12 shown, which is implemented to include:

[0097] Step S121: Determine the second reference path of the second sub-region;

[0098] Step S122: Determine whether a newly added first obstacle or a removed second obstacle is encountered according to the second dynamic layer and the first dynamic layer updated in real time. When it is determined that a newly added first obstacle is encountered, execute step S123. When it is determined that a removed second obstacle is encountered, execute step S124;

[0099] Step S123: Perform local path planning according to the first dynamic layer updated in real time and the coverage status map, determine the exploration path for avoiding the newly added first obstacle, and correct the second reference path according to the exploration path;

[0100] Step S124: Perform local path planning according to the first dynamic layer updated in real time and the coverage status map, determine the supplementary cleaning path for covering the area where the removed second obstacle is located, and correct the second reference path according to the supplementary cleaning path.

[0101] When performing a cleaning task on the second sub-region, there are two situations: In the first situation, when judging the uncleaned area surrounded by the first cleaning path, no area that meets the first screening condition is found, that is, no first sub-region can be divided within the area surrounded by the first cleaning path; In the second situation, by judging the uncleaned area surrounded by the first cleaning path, at least one first sub-region is found. At this time, the cleaning tasks will be sequentially performed on these first sub-regions first, and then the remaining second sub-region will be cleaned. Therefore, for step S121, the second reference path of the second sub-region can be determined according to the specific situation. For the first situation, the area surrounded by the first cleaning path will be regarded as the second sub-region. At this time, the actual driving path during the previous cleaning task, such as the actual first cleaning path, can be used to shrink the actual driving path of the previous cleaning by a preset width, such as shrinking by the width of a vehicle or a cleaning width, to form the reference path for this cleaning task; For the second situation, the first dynamic layer, the second dynamic layer and the coverage status map will be combined to select the second sub-region of the uncleaned area closest to the current position. The reference path for this circle of the second sub-region will be dynamically planned through the "return" shaped strategy based on the second dynamic map and the global map. During the subsequent cleaning of the same second sub-region, the actual driving path of the previous circle in the second sub-region will be shrunk by a preset width as the reference path for this circle.

[0102] Since the second sub-region may be an uncleaned area with possible obstacles or an irregular uncleaned area, during the actual execution of the cleaning task, there may be dynamic changes such as the emergence of new obstacles or the removal of obstacles. At this time, there will be a real-time local obstacle avoidance requirement or a requirement for supplementary cleaning due to missed cleaning. Therefore, in the embodiments of the present invention, in order to ensure more effective cleaning efficiency, local dynamic planning will be performed on the second sub-region according to the obstacle information in the first dynamic layer obtained in real time. Among them, in the process of performing local dynamic planning on the second sub-region according to the obstacle information in the first dynamic layer obtained in real time, each time the first dynamic layer is obtained (that is, each time new sensing information is detected or new sensing input information is received), the real-time obstacle information in the latest obtained first dynamic layer is compared with the static obstacle information recorded in the second dynamic layer, and the second dynamic layer is compared with the global grid map to determine whether there is a newly added first obstacle or a removed second obstacle. Since the first dynamic layer is a local dynamic map obtained according to real-time sensing information, the second dynamic layer is a dynamic map that records static obstacle information and is dynamically updated according to environmental changes, and the global grid map is pre-constructed. Therefore, in the embodiments of the present invention, during the execution of the cleaning task, by maintaining the comparison relationship between the first dynamic layer and the second dynamic layer, and between the second dynamic layer and the global grid map, the real-time changes in the environment, especially the changes in obstacles, are determined to achieve accurate segmentation and determination of the effective area to be cleaned.

[0103] Preferably, the embodiments of the present invention further update the second dynamic layer according to the first dynamic layer, so that the determination of the uncleaned area and the division of the first sub-region in the present invention are also effective and real-time, so that the entire planning method of the present invention can effectively adapt to dynamic environmental changes.

[0104] Among them, when performing local path planning according to the first dynamic layer, for the case of newly added obstacles, the purpose of local path planning is to achieve real-time local obstacle avoidance, while for the case of removed obstacles, the purpose of local path planning is to achieve immediate supplementary cleaning. For immediate supplementary cleaning, mainly when the latest first dynamic layer is obtained in real time, according to the real-time comparison result of the first dynamic layer and the second dynamic layer, when a missed cleaning area is found near the current driving position due to the existence of removed second obstacles, the missed cleaning area will be included in the scope of execution of the current cleaning task to achieve immediate supplementary cleaning, reduce the pressure in the later unified supplementary cleaning stage, and improve the cleaning efficiency. Exemplarily, it can be implemented as first locally sampling the area according to the first dynamic layer and the coverage status map and planning a supplementary cleaning path, and then connecting the supplementary cleaning path into the current cleaning path being executed, such as the second reference path, according to the planning result, so as to form a corrected second reference path, and performing the cleaning task along the corrected second reference path. For the case of newly added first obstacles encountered during the cleaning device's travel along the second reference path, the embodiments of the present invention, according to the first dynamic layer and the coverage status map, preferentially use the side-attaching algorithm to explore through the newly added first obstacle from the uncovered area, so as to determine the exploration path. After determining the exploration path, the embodiments of the present invention will correct the current cleaning path being executed, such as the second reference path, according to the exploration path, so as to form a corrected second reference path, and perform the cleaning task along the corrected second reference path. Since in the case of newly added first obstacles, the exploration path formed by side-attaching exploration will eventually return to the second reference path, that is, intersect with the reference path, therefore, the exploration path can be incorporated into the second reference path at the position where the exploration path intersects with the second reference path to replace the invalid path part in the second reference path, where the invalid path part refers to the part of the path that has not been actually executed due to avoiding obstacles through local dynamic planning.

[0105] In an actual scenario, when encountering newly added first obstacles, two situations will occur: The first situation is that the volume of the first obstacle is small, and after side-attaching exploration, the exploration path will eventually return to the second reference path, as Figure 13 shows this situation; The second situation is that the volume of the first obstacle is large enough to span the uncovered area. When the side-attaching exploration path finally returns to the second reference path, the new closed path formed by the exploration path and the second reference path dynamically divides the second sub-region, as Figure 14 shows this situation. For the first situation, after correcting the second reference path through the exploration path, the exploration path of local dynamic planning will be followed near the first obstacle, while in other sections, the second reference path will continue to be followed to perform the cleaning task, as Figure 13As shown in , the second reference path B1 and the exploration path B2 together form a new actual driving path, and finally the cleaning task can be performed according to the modified second reference path to achieve obstacle avoidance cleaning. Figure 14 As shown, after the second reference path is corrected by the exploration path, due to the large size of the first obstacle, when the exploration path B2 coincides with the second reference path B1, the two form a new closed path, thereby dividing the current second sub-area into two areas, namely, forming a new sub-area C1 and a new sub-area C2. In this case, after avoiding the obstacle, the second reference path after correction, that is, the actual driving path, is used as a reference, and a new second reference path for the next circle is formed by shrinking, and the cleaning task is continued to be performed on the divided area. For the other sub-area C2 generated by the obstacle avoidance segmentation, it will be planned as a new uncleaned area according to the aforementioned planning strategy and the cleaning task will be performed according to the planning results. Therefore, the embodiment of the present invention also realizes that in the process of local path planning for the second sub-area based on the first dynamic layer updated in real time, according to the size of the newly encountered first obstacle, when the newly encountered first obstacle is greater than the preset standard, the second sub-area is segmented according to the revised second reference path. Therefore, the embodiments of the present invention can not only realize intelligent partitioning in accordance with preset rules and conditions according to dynamic environmental changes, but also dynamically partition irregular areas according to obstacle conditions, thereby improving adaptability to dynamic environments and ensuring higher cleaning efficiency.

[0106] Since the second reference path is determined based on the inward or "U"-shaped strategy planning, the second reference path determined will be uneven and have missed scans due to the obstacle avoidance behavior in the previous lap of actual driving, the irregular boundary of the second sub-area, etc. Therefore, if the second reference path determined in this way is directly used as the cleaning path for strict tracking, it will inevitably lead to the defect of low cleaning efficiency. In addition, in actual scenes, when frequent obstacle avoidance occurs due to the movement of dynamic obstacles, the actual cleaning path will also generate a large number of small fragmented areas. If all small fragmented areas are placed in a unified re-sweeping stage for re-sweeping, it will inevitably lead to the need to connect a large number of small fragmented areas in series, making the cleaning path complicated and difficult to understand. Therefore, in order to solve the problems of low cleaning efficiency caused by dynamic changes in the environment, the present invention, when planning the second reference path, also utilizes the idea of dynamic programming and uses biased sampling to optimize the second reference path initially generated based only on the retraction method or the "U"-shaped strategy, so as to determine the optimal result that can simultaneously ensure smoothness, overlap with the covered area, and the effect of re-scanning the missed areas in the outer circle. This ensures that when driving according to the dynamically planned second reference path, the path can be smooth, the repetition rate can be low, and the effect of instant re-scanning can be achieved. Figure 15The method process of optimizing the second reference path by using the sampling with tendency to determine the second reference path that can be used as the basis for actual driving is shown schematically. Figure 15 As shown, its implementation includes:

[0107] Step S121A: performing lateral sampling based on the initially generated second reference path to obtain multiple sets of third sampling point sets, wherein each set of third sampling point sets corresponds to a lateral direction;

[0108] Step S121B: screening each group of third sampling point sets based on the second screening condition to obtain multiple groups of fourth sampling point sets;

[0109] Step S121C: setting a cost value for each sampling point according to the position of each sampling point in the fourth sampling point set;

[0110] Step S121D: selecting the best sampling curve from the multiple groups of sampling curves respectively formed based on the multiple groups of fourth sampling point sets according to a preset evaluation function to form an optimized second reference path.

[0111] Among them, the specific implementation process of step S121A and step S121B is similar to the aforementioned step S102 and step S103, and can refer to the description of the corresponding parts, which will not be repeated here. In step S121B, the second screening condition is specifically set to select the sampling point where the sampling point is connected to the covered area. The concept of connection here is: when the body of the cleaning equipment is located on the sampling point, the cleaning area of the cleaning equipment cannot have a gap with the covered area, but overlap with the covered area is allowed. In this way, by using the second screening condition to perform a tendency screening of the sampling points, the points that are not connected to the covered area can be filtered out, and through such screening, it is possible to ensure as few gaps as possible between the formed cleaning paths, improve the cleaning coverage rate, and reduce missed scans.

[0112] In step S121C, in order to further implement the tendency screening of the sampling points so that the cleaning path obtained based on the sampling points meets the desired goal, the cost value of each sampling point can be configured according to the desired goal and requirements to increase the tendency identification of the sampling points. With the desired goal of ensuring smoothness, overlap with the covered area, and the effect of filling in the outer circle missed area at the same time, C sample Taking the cost value set for the sampling point as an example, the cost value can be set for the sampling point based on the location of the sampling point. This can be achieved in the following ways:

[0113] First, set the three-level cost value C for the sampling point supplement , C uncleaned , C cleaned And give the three-level cost values set to a size relationship of C supplement<C uncleaned <C cleaned , where C supplement represents the cost of supplementary scanning at the location of the sampling point, C uncleaned represents the cost of the sampling point passing through an uncleaned area, and C cleaned represents the cost of the sampling point passing through a cleaned area. Then, the cost value is set for the sampling point according to the position of the sampling point. Specifically, the second reference path can be widened by a preset value first, such as widening half of the vehicle width or half of the cleaning width to the left and right. Then, the position of the sampling point is judged according to the second reference path, its widened range and the coverage status map, and the cost value of the sampling point is set according to the judgment result. When it is judged that the sampling point is outside the widened area of the second reference path and the coverage status of the position where the sampling point is located is an uncovered area, the position of the sampling point is determined as an outer-ring missed-scanning area. At this time, the cost value of the sampling point is set according to the set supplementary scanning cost; when it is judged that the sampling point is within the widened area of the second reference path and the coverage status is an uncovered area, the cost value of the sampling point is set according to the set cost of the sampling point passing through an uncleaned area; when it is judged that the sampling point is within the widened area of the second reference path and the coverage status is a covered area, the cost value of the sampling point is set according to the set cost of the sampling point passing through a cleaned area. Among them, when setting the cost value of the sampling point according to the set supplementary scanning cost, the distance between the sampling point and the central sampling point of the second reference path can be considered at the same time, so as to add a cost to the degree of deviation of the sampling point to the inner ring at the same time. Exemplarily, taking the distance between the sampling point and the central sampling point of the second reference path as d and the widened range of the second reference path as [d min , d max as an example, when the position of the sampling point is in the outer-ring missed-scanning area, the cost value of the sampling point can be set to Exemplarily, when the sampling point is within the widened range and in an uncleaned area, the cost value of the sampling point can be set to C sample = C uncleaned . Exemplarily, when the sampling point is within the widened range and in a cleaned area, the cost value of the sampling point can be set to C sample = C cleaned .

[0114] In step S121D, the evaluation function can be set according to the desired goal. Taking the desired goal as simultaneously ensuring smoothness, overlap with the covered area, and supplementary scanning effect on the outer-ring missed-scanning area as an example, the cost value of the sampling point, the smoothness of the cleaning path formed based on the sampling point, and the overlap rate of the cleaning path formed based on the sampling point are used to determine the evaluation function. Exemplarily, the evaluation function can be set to be represented by the following formula:

[0115] Cost = k sample * C sample+k curvature *C curvature +k repeat *C repeat

[0116] Among them, Cost is used to represent the total cost of the cleaning path formed according to the sampling points, k sample 、k curvature and k repeat are used to represent the weight values assigned to each evaluation item, C curvature represents the cost of the path smoothness, C repeat represents the cost of the path overlap rate. Among them, the value of C curvature can be set to the sum of the curvatures of the curves connecting the waypoints on the cleaning path, and C repeat can be specifically determined by the overlapping area of the widened area of the cleaning path and the covered area and the widened area. Exemplarily, it can be set to be calculated by the following method: First, widen the obtained cleaning path according to the cleaning width, and calculate the area A path occupied by the widened cleaning path; then, calculate the overlapping area A repeat between the area occupied by the widened cleaning path and the covered area; finally, calculate the overlap rate as A repeat / A path .

[0117] Thus, by sampling through the second reference path formed based on the shrinking method or the "return" - shaped strategy, and performing a tendency screening on the sampling points through the screening conditions, and performing a tendency evaluation on the sampling curve formed by the screened sampling points, the optimal sampling curve can be finally determined as the second reference path that can be used as the actual driving basis. Among them, the determined optimal sampling curve is the sampling curve with the minimum cost value. In addition, the specific method of forming a sampling curve from the screened sampling points can refer to the description in the previous Figure 2 section and will not be elaborated here. By designing a cost function based on the cost value set for the sampling points, the cost value of the path smoothness, and the cost value of the path overlap rate to evaluate the sampling curve determined by the sampling points and selecting the optimal curve as the second reference path, not only can immediate supplementary cleaning be achieved, but also the un - cleaned area can be selected for obstacle avoidance, and at the same time, the smoothness of the finally generated second reference path can be ensured. Taking the evaluation of the sampling curve composed of sampling points through the above - mentioned evaluation function to determine the optimal second reference path as an example, Figure 16 shows its tendency screening for immediate supplementary cleaning, as Figure 16As shown, the shaded part M marked in the figure is the cleaned area, the white part N is the uncleaned area, the coloring route O is the second reference route generated after the actual route of the outer circle is shrunk, and the dotted line P is the dynamic programming sampling point obtained by horizontally sampling based on the second reference route. When screening and evaluating the sampling points as described above, the uncleaned area of the outer circle is preferably selected. Therefore, after the above screening and evaluation, the final generated route after widening is the colored route Q, which can achieve immediate supplementary cleaning of the uncleaned area of the outer circle. Taking the example of evaluating the sampling curve formed by the sampling points through the above evaluation function to determine the optimal second reference route, Figure 17 It shows the tendency screening for obstacle avoidance by preferentially selecting uncleaned areas, such as Figure 17 As shown, since there is no uncleaned area in the outer circle, when screening and evaluating the sampling points, the uncleaned area of the inner circle is preferably selected. Therefore, the final generated route after widening is as shown by the route Q, which preferably avoids obstacles through the inner circle. Taking the example of evaluating the sampling curve formed by the sampling points through the above evaluation function to determine the optimal second reference route, Figure 18 It shows the tendency screening for the smoothness of the route, such as Figure 18 , even if the initially determined second reference route O is uneven, a smoother route is preferably selected during local route planning to determine the optimal second reference route Q. This route Q overlaps with the cleaned area and can improve the flatness of the route, thus improving the coverage rate and cleaning efficiency.

[0118] In other preferred embodiments, after the cleaning task is completed based on the above planning strategy, if there are still missed cleaning areas in the target area, the target area will be further uniformly supplemented with cleaning. As a preferred embodiment, when performing the uniform supplementary cleaning, the embodiment of the present invention further determines the effective area to be cleaned according to the second dynamic layer to reduce the ineffective cleaning area in the target area and increase the effective cleaning area in the target area. Specifically, the effective area to be cleaned in the target area can be corrected according to the changes in the static obstacles recorded in the dynamically updated second dynamic map, and the uniform supplementary cleaning can be performed according to the corrected effective area to be cleaned to avoid ineffective supplementary cleaning planning and improve the cleaning efficiency. Exemplarily, the area to be cleaned that needs to be supplemented in the current target area can be corrected according to the second dynamic map, the global map corresponding to the target area, and the coverage status map. For example, according to the comparison result of the second dynamic map and the global map, the latest obstacle information in the global map can be determined, and according to the latest obstacle information and the coverage status map, the current effective area to be cleaned in the global map can be determined to perform supplementary cleaning on the determined current effective area to be cleaned. Figure 19Schematically shows the display effect of correcting the area to be cleaned that needs to be reswept in the current target area according to the second dynamic map, the global map corresponding to the target area, and the coverage status map to achieve effective resweeping and avoid ineffective resweeping. As Figure 19 shown, Figure 19 A is the global map corresponding to the target area, Figure 19 B is the second dynamic map after real-time update, which records the static obstacles not in the global map; Figure 19 C is the coverage status map after cleaning. In Figure 19 the state of A, if the judgment is not combined with the second dynamic map, since there is a static obstacle J only recorded in the second dynamic map but not in the global map in the global map, the area corresponding to the obstacle J not recorded by the global map also needs to be reswept. However, after combining the second dynamic map, the global map, and the coverage status map for judgment, Figure 19 D is the area that needs to be reswept after being corrected by combining the second dynamic map judgment. The area of the static obstacle J covered in the corrected area to be cleaned is finally reduced, avoiding ineffective resweeping.

[0119] Thus, the solution provided by the embodiment of the present invention realizes autonomous dynamic zoning during the edge cleaning process based on the local dynamic map, and intelligently determines to use a more efficient coverage method to generate a path according to the zoning constraint conditions, improving the path coverage rate and ensuring the cleaning efficiency; at the same time, the solution of the embodiment of the present invention also makes the reference paths planned in different zones smooth and feasible, the curvature meet the turning radius constraint, and reduces the number of in-place turns through the method of secondary optimization of the path and adaptive calculation of key parameters, further improving the cleaning efficiency and path coverage rate; in addition, the embodiment of the present invention also realizes instant resweeping during local path planning by performing dynamic local path planning in the zone, reduces the ineffective resweeping area and increases the effective resweeping area according to the local dynamic map during the unified resweeping stage, and improves the resweeping efficiency and coverage rate.

[0120] Figure 20 Schematically shows a dynamic full-coverage path planning device according to an embodiment of the present invention. As Figure 20 shown, the device includes:

[0121] a memory 60 for storing executable instructions; and

[0122] a processor 61 for executing the executable instructions stored in the memory.

[0123] Wherein, as a preferred implementation example, the executable instructions stored in the memory 60, when executed by the processor, cause the processor to execute the dynamic full-coverage path planning method described in any embodiment of the present invention.

[0124] Figure 21 Schematically shows a dynamic full-coverage path planning device according to an embodiment of the present invention, as Figure 21 shown. The device includes:

[0125] A first planning module 100, configured to generate a first cleaning path according to the boundary of the area to be cleaned, and perform edge cleaning on the area to be cleaned according to the first cleaning path;

[0126] A second planning module 200, configured to divide a first sub-area that meets a first preset condition from the area surrounded by the first cleaning path, perform full-coverage path planning on the divided first sub-area, and execute a cleaning task on the first sub-area according to the full-coverage path planning result;

[0127] A third planning module 300, configured to determine a second sub-area in the area surrounded by the first cleaning path, perform local path planning on the second sub-area based on a first dynamic layer updated in real time, and execute a cleaning task on the second sub-area according to the local path planning result.

[0128] Wherein, in a preferred embodiment of the present invention, the area to be cleaned preferably refers to the target area to be cleaned, and its boundary can be obtained through the map boundary of the grid map corresponding to the target area to be cleaned. The first preset condition preferably adapts to the full-coverage path planning method adopted for the divided first sub-area. Exemplarily, taking the "bow" - shaped strategy as the full-coverage path planning method adopted, since the "bow" - shaped full-coverage can well apply to relatively large and regular uncleaned areas and achieve fast and efficient full-coverage path planning, in this case, the first preset condition can be set to include that the area of the uncleaned area is greater than a first preset value, the length and width of the circumscribed rectangle of the uncleaned area are respectively greater than a second preset value and a third preset value, and the area ratio of the uncleaned area inside its circumscribed rectangle is greater than a fourth preset value. Thus, a first sub-area that meets this first preset condition is divided from the area surrounded by the first cleaning path, and the "bow" - shaped strategy is used to perform full-coverage path planning on this type of first sub-area, thereby achieving the effects of improving the adaptability of the path planning method to the dynamic environment and improving the cleaning efficiency through dynamic zoning and adaptive path planning. Among them, the first preset value, the second preset value, the third preset value, and the fourth preset value can be set according to requirements or prior experience.

[0129] As a preferred implementation manner, the second planning module 200 is preferably configured to determine the uncleaned area within the area surrounded by the first cleaning path based on the second dynamic layer corresponding to the dynamically updated target area and the coverage status map, and determine whether the uncleaned area meets a first preset condition, and use the uncleaned area that meets the first preset condition as the first sub-area. Specifically, since the start of the self-cleaning task, two dynamic layers and a coverage status map can be dynamically constructed and maintained for the target area to assist the global grid map corresponding to the target area to achieve real-time dynamic path planning. Among them, the two dynamic layers constructed for the target area may include a first dynamic layer obtained according to real-time perception information and a second dynamic layer recording all static obstacle information in the target area. Thus, as the cleaning device moves and its position is updated in real time, a real-time dynamic local map (referred to as the first dynamic layer in the embodiments of the present invention) within a certain range can be obtained at any time through real-time perception information, and the latest static obstacle information can be obtained according to the real-time dynamic local map and recorded in the dynamically updated second dynamic layer for real-time dynamic path planning. Exemplarily, a dynamic layer with a preset range size, such as 7.5m × 7.5m, can be maintained centered on the vehicle itself based on real-time perception information and saved in the memory as the first dynamic layer, and the first dynamic layer is updated in real time according to the position of the vehicle's movement; at the same time, a second dynamic layer recording static obstacle information is saved in the map folder in the form of a temporary file. Among them, the second dynamic layer is a global dynamic layer corresponding to the global grid map, and the second dynamic layer can be dynamically updated according to the first dynamic layer or the static obstacle information detected by perception. For example, when a new perception input is received each time, through global coordinate conversion, the first dynamic layer or the static obstacle information detected by perception is updated to the second dynamic layer.

[0130] As a preferred implementation manner, after the first sub-area is dynamically segmented in the present invention, the second sub-area within the area surrounded by the first cleaning path is further determined according to the segmentation result of the first sub-area. Specifically, according to the judgment result of whether the uncleaned area within the area surrounded by the first cleaning path meets the first preset condition, the uncleaned areas that do not meet the first preset condition are all used as the second sub-areas.

[0131] Since in practical applications, there may be more than one first sub-area within the area surrounded by the first cleaning path. In this case, the embodiments of the present invention preferably perform full-coverage path planning and execute cleaning tasks in descending order of the area size of the first sub-areas that can be segmented. More preferably, the processing of the second sub-areas is performed after the corresponding cleaning tasks are executed for all the first sub-areas segmented.

[0132] Therefore, the device according to the embodiments of the present invention can not only dynamically partition the target area during the task execution, but also perform full coverage and local path planning based on the dynamically updated dynamic layer to determine a more accurate effective area to be cleaned, and achieve instant supplementary cleaning and avoid missed cleaning by instantaneously adjusting the cleaning path, thereby improving the cleaning efficiency.

[0133] Figure 22 Schematically shows a dynamic full-coverage path planning device according to an embodiment of the present invention, as Figure 22 shown. In this device, the first planning module 100 specifically includes:

[0134] A reference path generation unit 100A, configured to shrink the grid map boundary by a preset width on the grid map corresponding to the area to be cleaned to form a first reference path;

[0135] A sampling unit 100B, configured to perform horizontal sampling based on the first reference path to obtain multiple sets of first sampling point sets, where each set of first sampling point sets corresponds to a horizontal direction;

[0136] A tendency screening unit 100C, configured to screen each set of first sampling point sets according to a first screening condition to obtain multiple sets of second sampling point sets;

[0137] An optimal solution obtaining unit 100D, configured to splice the optimal curve segments in each set of sampling curves formed by multiple sets of second sampling point sets to form a first cleaning path.

[0138] Figure 23 Schematically shows a dynamic full-coverage path planning device according to another embodiment of the present invention, as Figure 23 shown. In the device according to the embodiment of the present invention, the second path planning module 200 is further implemented to include:

[0139] A path coverage unit 200A, configured to perform a "bow"-shaped filling on the first sub-region to form a "bow"-shaped reference path;

[0140] A path optimization unit 200B, configured to determine the optimal connection order between the "bow"-shaped reference paths, and correct the connection mode of the "bow"-shaped reference paths according to the determined optimal connection order to form an optimized "bow"-shaped cleaning path as the full-coverage path planning result for the first sub-region.

[0141] As a preferred embodiment, the path optimization unit 200B can specifically screen out the optimal connection order between each straight-line route according to all possible connection orders between the straight-line routes in the "bow"-shaped reference path and a preset evaluation function. Among them, the preset evaluation function can be set according to the type of the connection line between the selected straight-line routes. Exemplarily, taking the connection between straight-line routes by a Bezier curve as an example, the preset evaluation function can be set to be determined by the sum of the lengths of all straight-line routes in the "bow"-shaped reference path, the sum of the curvatures of the Bezier curves used to connect the straight-line routes according to the current connection order, and the total length of the Bezier curves used to connect all straight-line routes.

[0142] Figure 24 Schematically shows a dynamic full-coverage path planning device according to another embodiment of the present invention, as Figure 24 shown, in the device of the embodiment of the present invention, the third path planning module is implemented to further include:

[0143] A path determination unit 300A, configured to determine a second reference path for the second sub-region;

[0144] A dynamic scheduling unit 300B, configured to determine whether a newly added first obstacle or a removed second obstacle is encountered according to the second dynamic layer and the real-time updated first dynamic layer. When it is determined that a newly added first obstacle is encountered, the first dynamic programming unit is called for local path planning. When it is determined that there is a removed second obstacle, the second dynamic programming unit is called for local path planning;

[0145] A first dynamic programming unit 300C, configured to perform local path planning according to the real-time updated first dynamic layer and the coverage status map, determine an exploration path for avoiding the newly added first obstacle, and correct the second reference path according to the exploration path;

[0146] A second dynamic programming unit 300D, configured to perform local path planning according to the real-time updated first dynamic layer and the coverage status map, determine a supplementary scanning path for covering the area where the removed second obstacle is located, and correct the second reference path according to the supplementary scanning path.

[0147] As a preferred implementation manner, the path determination unit 300A can specifically be implemented to:

[0148] Perform horizontal sampling based on the initially generated second reference path to obtain multiple groups of third sampling point sets, where each group of third sampling point sets corresponds to a horizontal direction, and the initially generated second reference path is generated according to the actual driving path of the previous lap or the "return"-shaped path planning of the second sub-region;

[0149] Screen each set of third sampling point sets based on the second screening condition to obtain multiple sets of fourth sampling point sets;

[0150] Set evaluation values for each sampling point according to the positions of the sampling points in the fourth sampling point set;

[0151] Select the optimal sampling curve from multiple sets of sampling curves respectively formed based on multiple sets of fourth sampling point sets according to a preset evaluation function to form an optimized second reference path.

[0152] As another preferred embodiment, the third path planning module 300 may further include:

[0153] A dynamic partitioning unit, configured to perform regional segmentation on the second sub-region according to the size of the newly encountered obstacle when the newly encountered obstacle is greater than a preset value according to the corrected second reference path. It should be noted that the specific implementation processes and implementation principles of the various modules and units of the dynamic full-coverage path planning device in the embodiments of the present invention can be specifically referred to the corresponding descriptions in the above method embodiments, so they will not be elaborated here. Exemplarily, the dynamic full-coverage path planning device in the embodiments of the present invention can be any intelligent device with a processor, including but not limited to a computer, a smart phone, a personal computer, a robot, a cloud server, etc.

[0154] Figure 25 Schematically shows a cleaning device according to an embodiment of the present invention, such as Figure 25 shown, the cleaning device includes:

[0155] A fuselage 70;

[0156] A dynamic full-coverage path planning device 71, disposed on the fuselage 70.

[0157] Among them, the dynamic full-coverage path planning device 71 can select the device of any of the above embodiments. Among them, the specific implementation process and implementation principle of the dynamic full-coverage path planning device can be specifically referred to the corresponding descriptions in the above embodiments, and will not be elaborated here. It should be noted that the cleaning device in the embodiments of the present invention can be an unmanned cleaning vehicle, an unmanned sweeper, a floor sweeping robot, etc. with an automatic cleaning function.

[0158] In some embodiments, the present invention provides a non-volatile computer-readable storage medium, in which one or more programs including execution instructions are stored, and the execution instructions can be read and executed by an electronic device (including but not limited to a computer, a server, or a network device, etc.) for executing the dynamic full-coverage path planning method of any one of the above embodiments of the present invention.

[0159] In some embodiments, the embodiments of the present invention further provide a computer program product, which includes a computer program stored on a non-volatile computer-readable storage medium. The computer program includes program instructions that, when executed by a computer, cause the computer to execute the dynamic full-coverage path planning method of any of the above embodiments.

[0160] In some embodiments, the embodiments of the present invention further provide an electronic device, which includes: at least one processor, and a memory communicatively connected to the at least one processor. Wherein, the memory stores instructions executable by the at least one processor, and when the instructions are executed by the at least one processor, the at least one processor is enabled to execute the dynamic full-coverage path planning method of any of the above embodiments.

[0161] In some embodiments, the embodiments of the present invention further provide a storage medium, on which a computer program is stored. The feature is that when the program is executed by a processor, it implements the dynamic full-coverage path planning method of any of the above embodiments.

[0162] Figure 26 is a schematic hardware structure diagram of an electronic device for executing the dynamic full-coverage path planning method provided by another embodiment of the present invention. As Figure 26 shown, the device includes:

[0163] One or more processors 610 and a memory 620, Figure 26 Taking one processor 610 as an example.

[0164] The device for executing the dynamic full-coverage path planning method may further include: an input device 630 and an output device 640.

[0165] The processor 610, the memory 620, the input device 630, and the output device 640 may be connected through a bus or other means, Figure 26 Taking connection through a bus as an example.

[0166] The memory 620, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules, such as the program instructions / modules corresponding to the dynamic full-coverage path planning method in the embodiments of the present invention. The processor 610 executes various functional applications and data processing of the server by running the non-volatile software programs, instructions, and modules stored in the memory 620, that is, implements the dynamic full-coverage path planning method of the above method embodiments.

[0167] The memory 620 may include a program storage area and a data storage area. The program storage area may store an operating system and application programs required for at least one function. The data storage area may store data created according to the use of the dynamic full-coverage path planning method, etc. In addition, the memory 620 may include high-speed random access memory and may also include non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, or other non-volatile solid-state storage devices. In some embodiments, the memory 620 may optionally include a memory remotely disposed relative to the processor 610, and these remote memories may be connected to the electronic device through a network. Examples of the above network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0168] The input device 630 may receive input digital or character information and generate signals related to user settings and function control of the dynamic full-coverage path planning device. The output device 640 may include a display device such as a display screen.

[0169] The one or more modules are stored in the memory 620 and, when executed by the one or more processors 610, execute the dynamic full-coverage path planning method in any of the above method embodiments.

[0170] The above product may execute the method provided by the embodiments of the present invention, and has corresponding functional modules and beneficial effects for executing the method. Technical details not described in detail in this embodiment may be referred to the method provided by the embodiments of the present invention.

[0171] The electronic device according to the embodiments of the present invention exists in various forms, including but not limited to:

[0172] (1) Mobile communication devices: Such devices are characterized by having mobile communication functions and mainly aim to provide voice and data communication. Such terminals include: smart phones (such as iPhone), multimedia phones, functional phones, and low-end phones, etc.

[0173] (2) Ultra-mobile personal computer devices: Such devices belong to the category of personal computers, have computing and processing functions, and generally also have the characteristic of mobile Internet access. Such terminals include: PDAs, MIDs, and UMPC devices, etc., such as iPad.

[0174] (3) Portable entertainment devices: Such devices can display and play multimedia content. Such devices include: audio and video players (such as iPod), handheld game consoles, e-books, and smart toys and portable vehicle navigation devices.

[0175] (4) Server: A device that provides computing services. The server consists of a processor, hard disk, memory, system bus, etc. The server is similar to a general computer architecture, but due to the need to provide highly reliable services, it has higher requirements in terms of processing power, stability, reliability, security, scalability, manageability, etc.

[0176] (5) Other electronic devices with data interaction functions.

[0177] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0178] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence, or the parts that contribute to the relevant technologies, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0179] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of each embodiment of the present application.

Claims

1. A dynamic full-coverage path planning method, characterized in that, The method includes: Generating a first cleaning path according to the boundary of the area to be cleaned, and performing edge cleaning on the area to be cleaned according to the first cleaning path; Dividing a first sub-region that meets the first preset condition from the area surrounded by the first cleaning path, performing full-coverage path planning on the divided first sub-region, and performing a cleaning task on the first sub-region according to the full-coverage path planning result, where the first sub-region is an uncleaned area that meets the first preset condition within the area surrounded by the first cleaning path, and the uncleaned area is determined based on a second dynamic layer corresponding to the dynamically updated target area and a coverage status map; Determining a second sub-region in the area surrounded by the first cleaning path, performing local path planning on the second sub-region based on the real-time updated first dynamic layer, and performing a cleaning task on the second sub-region according to the local path planning result, where the second sub-region is an uncleaned area within the area surrounded by the first cleaning path that does not meet the first preset condition.

2. The method according to claim 1, wherein The generating a first cleaning path according to the boundary of the area to be cleaned includes: Shrinking the grid map boundary by a preset width on the grid map corresponding to the area to be cleaned to form a first reference path; Performing horizontal sampling based on the first reference path to obtain multiple sets of first sampling point sets, where each set of first sampling point sets corresponds to a horizontal direction; Screening each set of first sampling point sets according to the first screening condition to obtain multiple sets of second sampling point sets; Selecting the optimal curve segments from the multiple sets of sampling curves formed by the multiple sets of second sampling point sets and splicing them to form a first cleaning path.

3. The method according to claim 1, wherein The first preset condition includes that the area of the uncleaned area is greater than a first preset value, the length and width of the circumscribed rectangle of the uncleaned area are respectively greater than a second preset value and a third preset value, and the area ratio of the uncleaned area inside its circumscribed rectangle is greater than a fourth preset value.

4. The method according to claim 3, wherein The dividing a first sub-region that meets the first preset condition from the area surrounded by the first cleaning path includes: Determining the uncleaned area within the area surrounded by the first cleaning path based on a second dynamic layer corresponding to the dynamically updated target area and a coverage status map, where the current static obstacle information in the target area is marked in the second dynamic layer, and the cleaned area in the target area is marked as a covered area and the uncleaned area is marked as an uncovered area in the coverage status map; Judging whether the uncleaned area meets the first preset condition, and taking the uncleaned area that meets the first preset condition as the first sub-region; The determining a second sub-region in the area surrounded by the first cleaning path includes: Taking the uncleaned area that does not meet the first preset condition as the second sub-region.

5. The method according to claim 4, wherein The performing full-coverage path planning on the divided first sub-region includes: Performing a "bow"-shaped filling on the first sub-region to form a "bow"-shaped reference path; Determining the optimal connection order between the "bow"-shaped reference paths, and correcting the connection mode of the "bow"-shaped reference path according to the determined optimal connection order to form an optimized "bow"-shaped cleaning path as the full-coverage path planning result for the first sub-region.

6. The method according to claim 5, characterized in that, Determining the optimal connection order between the "bow"-shaped reference paths includes: Determining all possible connection orders between the straight-line routes in the "bow"-shaped reference paths; Evaluating all possible connection orders respectively according to a preset evaluation function, and screening out the optimal connection order between the straight-line routes according to the evaluation results.

7. The method according to claim 6, wherein, The preset evaluation function is determined by the sum of the lengths of all the straight-line routes in the "bow"-shaped reference path, the sum of the curvatures of the Bezier curves used for connection between the straight-line routes determined according to the current connection order, and the total length of the Bezier curves used for connecting all the straight-line routes; Modifying the connection mode of the "bow"-shaped reference path according to the determined optimal connection order to form an optimized "bow"-shaped cleaning path, including: Reconnecting the straight-line routes in the "bow"-shaped reference path with Bezier curves according to the optimal connection order to form an optimized "bow"-shaped cleaning path.

8. The method according to claim 7, wherein The local path planning for the second sub-region is carried out after the corresponding cleaning tasks have been completed for all the first sub-regions segmented.

9. The method according to claim 4, characterized in that, The local path planning for the second sub-region based on the first dynamically updated layer includes: Determining a second reference path for the second sub-region; Determining whether a newly added first obstacle or a removed second obstacle is encountered according to the second dynamic layer and the first dynamically updated layer. When it is determined that a newly added first obstacle is encountered, local path planning is carried out according to the first dynamically updated layer and the coverage status map to determine an exploration path for avoiding the newly added first obstacle, and the second reference path is corrected according to the exploration path; When it is determined that there is a removed second obstacle, local path planning is carried out according to the first dynamically updated layer and the coverage status map to determine a supplementary cleaning path for covering the area where the removed second obstacle is located, and the second reference path is corrected according to the supplementary cleaning path.

10. The method according to claim 9, wherein During the process of local path planning for the second sub-region based on the first dynamically updated layer, it also includes: According to the size of the newly added obstacle encountered, when the newly added obstacle encountered is greater than a preset value, the second sub-region is segmented according to the corrected second reference path.

11. The method according to claim 9 or 10, characterized in that, Determining the second reference path for the second sub-region includes: Performing horizontal sampling on the initially generated second reference path to obtain multiple sets of third sampling point sets, where each set of third sampling point sets corresponds to a horizontal direction, and the initially generated second reference path is generated according to the actual driving path of the previous lap or the "return"-shaped path planning for the second sub-region; Screening each set of third sampling point sets respectively according to the second screening condition to obtain multiple sets of fourth sampling point sets; Setting evaluation values for each sampling point according to the positions of the sampling points in the fourth sampling point sets; Selecting the optimal sampling curve from multiple sets of sampling curves respectively formed based on multiple sets of fourth sampling point sets according to the preset evaluation function to form an optimized second reference path.

12. Dynamic full-coverage path planning device, characterized in that, Including a memory for storing executable instructions; and A processor for executing executable instructions stored in a memory, the executable instructions, when executed by the processor, causing the processor to execute the dynamic full-coverage path planning method according to any one of claims 1 to 11.

13. Dynamic full-coverage path planning device, characterized in that, The device includes: A first planning module for generating a first cleaning path according to the boundary of the area to be cleaned and performing edge cleaning on the area to be cleaned according to the first cleaning path; A second planning module for dividing a first sub-region that meets a first preset condition from the area surrounded by the first cleaning path, performing full-coverage path planning on the divided first sub-region, and executing a cleaning task on the first sub-region according to the full-coverage path planning result, where the first sub-region is an uncleaned area that meets the first preset condition within the area surrounded by the first cleaning path, and the uncleaned area is determined based on a second dynamic layer corresponding to a dynamically updated target area and a coverage status map; A third planning module for determining a second sub-region in the area surrounded by the first cleaning path, performing local path planning on the second sub-region based on a real-time updated first dynamic layer, and executing a cleaning task on the second sub-region according to the local path planning result, where the second sub-region is an uncleaned area that does not meet the first preset condition within the area surrounded by the first cleaning path.

14. Cleaning device, characterized in that, Includes: A fuselage; And The dynamic full-coverage path planning device according to claim 12 or 13, provided on the fuselage.

15. A storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the method according to any one of claims 1 to 11.

16. A computer program product comprising instructions, characterized in that, When the computer program product runs on a computer, it causes the computer to execute the dynamic full-coverage path planning method according to any one of claims 1 to 11.

Images