Multi-region task planning method and device, electronic equipment and storage medium

Abstract

The invention relates to a multi-area task planning method and device, electronic equipment and a storage medium. The method comprises the steps that a robot is controlled to scan all channel labels in a current area along the edge; constructing a connection relationship between the current area and the adjacent area by reading the number information in the channel label, and determining the number of the current area; after a target channel is determined based on the connection relation, the robot is controlled to travel to the non-scanned adjacent area through the target channel, and time consumed for passing through the target channel is recorded; after all the areas are traversed, a weighted graph is generated based on the connection relation between the areas and the consumed time of passing through the channel, a multi-area task is planned according to the weighted graph, each vertex of the weighted graph represents one area, the serial number of the vertex is the serial number of the corresponding area, and the weight of each edge represents the consumed time of passing through the channel. According to the invention, it is ensured that the robot can plan a multi-region task without global positioning and a global map.

Description

Technical Field

[0001] This application relates to the technical field of robots, and particularly to a method, device, electronic device, and storage medium for planning multi-region tasks. Background Art

[0002] In many practical application scenarios, robots need to be able to autonomously plan and execute tasks in multiple regions. For example, in garden maintenance, a lawn mowing robot needs to work efficiently and orderly between multiple lawn areas; in household cleaning, a floor sweeping robot needs to clean between multiple rooms. To achieve the effective planning and execution of these tasks, it usually relies on the fusion positioning technology of multiple sensors to establish a detailed regional map and execute the tasks of each region one by one according to this regional map.

[0003] This method highly depends on the global positioning system (such as GPS) and a detailed regional map. If a monocular vision robot is used, and the robot lacks effective positioning means (such as weak or unavailable GPS signals in an indoor environment) and cannot accurately establish a regional map, then the planning and execution of tasks will become very difficult. This results in the inability of the robot to effectively plan and execute multi-region tasks without global positioning and a global map. Summary of the Invention

[0004] This application provides a method, device, electronic device, and storage medium for planning multi-region tasks to solve the problem of being unable to plan and execute multi-region tasks without global positioning and a global map.

[0005] In a first aspect, this application provides a method for planning multi-region tasks, and the method includes:

[0006] Controlling the robot to scan all channel tags in the current region along the edge, where the channel tags are set at the channel ports between every two adjacent regions;

[0007] Constructing the connection relationship between the current region and adjacent regions by reading the number information in the channel tags, and determining the number of the current region, where the number information is composed of the current region number and the number of the adjacent region to which it leads;

[0008] After determining the target channel based on the connection relationship, controlling the robot to travel through the target channel to an un-scanned adjacent region, and recording the time taken to pass through the target channel;

[0009] After traversing all areas, a weighted graph is generated based on the connection relationship between areas and the time consumed to pass through the channel, and multi-area tasks are planned according to the weighted graph, where each vertex of the weighted graph represents an area, the vertex number is the number of the corresponding area, and the weight of each edge represents the time consumed to pass through the channel.

[0010] Optionally, after determining the target channel based on the connection relationship, controlling the robot to travel through the target channel to an unscanned adjacent area includes:

[0011] After determining the target channel based on the connection relationship and the depth-first strategy, the robot is controlled to move through the target channel to an unscanned adjacent area according to the depth-first strategy.

[0012] Optionally, after determining the target channel based on the connection relationship and the depth-first strategy, controlling the robot to travel through the target channel to an unscanned adjacent area according to the depth-first strategy includes:

[0013] In a case where it is determined based on the connection relationship that there is at least one untraversed channel in the current area, controlling the robot to randomly or nearest select a target channel from the at least one untraversed channel, and travel to an unscanned adjacent area through the target channel;

[0014] Taking the unscanned adjacent area as the current area to construct an area connection relationship, repeating the above operation to continuously go deeper into the unscanned adjacent area, wherein the travel path of the robot is stored in a database;

[0015] If there is only one traversed channel in the current area, return to the previous area according to the stored travel path;

[0016] If there is an untraversed channel in the previous area, performing the steps of selecting and traveling to an unscanned adjacent area through a target channel among the untraversed channels;

[0017] If there is no untraversed passage in the previous area, the area is backtracked according to the stored travel path until an area with an untraversed passage is found.

[0018] Optionally, determining that there is at least one untraversed channel in the current area based on the connection relationship includes:

[0019] Determine all channels connected to the current area based on the connection relationship;

[0020] determining a traversed channel based on the travel path in the database;

[0021] Compare all the channels connected through the current area with the traversed channels to determine at least one un-traversed channel existing in the current area.

[0022] Optionally, comparing all the channels connected through the current area with the traversed channels to determine at least one un-traversed channel existing in the current area includes:

[0023] Determine the channel identifiers of all the channels connected through the current area by identifying the number information of all the channel labels in the current area, where the number information is used to indicate the channel identifier between two areas;

[0024] Determine the set of channel identifiers of the traversed channels by analyzing the travel path;

[0025] Compare all the channel identifiers corresponding to the current area with the set of channel identifiers of the traversed channels;

[0026] If there is a set channel identifier in the channel identifiers of the current area that does not appear in the set of channel identifiers, then regard the channel corresponding to the set channel identifier as an un-traversed channel.

[0027] Optionally, recording the time taken to pass through the target channel includes:

[0028] When the robot switches areas, record the first moment when the channel label at the first port of the target channel is scanned;

[0029] When the robot travels through the target channel to the next un-scanned adjacent area, record the second moment when the channel label at the second port of the target channel is scanned;

[0030] Take the time difference between the second moment and the first moment as the time taken for the robot to pass through the target channel.

[0031] Optionally, planning a multi-area task according to the weighted graph includes:

[0032] Arrange and combine the numbers of multiple vertices in the weighted graph to generate multiple area access sequences, where the area access sequence is used to indicate the set of number sequences for accessing areas in order;

[0033] Accumulate the weights of the corresponding edges in the weighted graph according to the area access order in the area access sequence to obtain the total weight of the edges of the area access sequence;

[0034] Compare the total weights of multiple area access sequences, determine the access order corresponding to the extreme total weight, and plan a multi-area task according to the access order, where the extreme total weight is the maximum total weight or the minimum total weight.

[0035] In a second aspect, the present application provides a planning device for multi-region tasks, the device comprising:

[0036] A scanning module, configured to control the robot to scan all channel tags along the edge in the current region, wherein the channel tags are set at the channel ports between every two adjacent regions;

[0037] A building and determining module, configured to build the connection relationship between the current region and the adjacent regions by reading the number information in the channel tags, and determine the number of the current region, wherein the number information is composed of the current region number and the number of the adjacent region to which it leads;

[0038] A traveling and recording module, configured to control the robot to travel through the target channel to an unscanned adjacent region after determining the target channel based on the connection relationship, and record the time taken to pass through the target channel;

[0039] A generating and planning module, configured to generate a weighted graph based on the connection relationship between regions and the time taken to pass through channels after traversing all regions, and plan multi-region tasks according to the weighted graph, wherein each vertex of the weighted graph represents a region, the number of the vertex is the number of the corresponding region, and the weight of each edge represents the time taken to pass through the channel.

[0040] In a third aspect, the present application provides an electronic device, comprising: at least one communication interface; at least one bus connected to the at least one communication interface; at least one processor connected to the at least one bus; at least one memory connected to the at least one bus.

[0041] In a fourth aspect, the present application further provides a computer storage medium, storing computer-executable instructions, the computer-executable instructions being used to execute the multi-region task planning method described in any one of the above of the present application.

[0042] The above technical solutions provided by the embodiments of the present application have the following advantages compared with the prior art: By setting channel tags containing number information at the channel ports, the connection relationship between adjacent regions and the number of the current region can be determined by using the robot to scan the channel tags along the edge. Then, the target channel is selected according to the connection relationship, and all regions are traversed without omission along the target channel to build a complete regional connection relationship graph. The robot also records the time taken to pass through the channel. Finally, a weighted graph is generated based on the regional connection relationship and the channel time, and multi-region tasks are planned through the weighted graph. The present application omits the complex multi-sensor fusion positioning and map building process, and can directly plan multi-region tasks according to the weighted graph, ensuring that the robot can plan multi-region tasks without global positioning and global map. Description of the Drawings

[0043] The accompanying drawings herein are incorporated into the specification and constitute a part of this specification, showing embodiments consistent with the present application, and are used together with the specification to explain the principles of the present application.

[0044] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the accompanying drawings required for use in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, without creative efforts, other accompanying drawings can also be obtained based on these drawings.

[0045] One or more embodiments are exemplarily illustrated by the pictures in the corresponding accompanying drawings. These exemplary illustrations do not constitute limitations on the embodiments. Elements with the same reference numerals in the accompanying drawings are represented as similar elements. Unless otherwise stated, the drawings in the accompanying drawings do not constitute a scale limitation.

[0046] Figure 1 A schematic diagram of constructing a regional map during multi-region mowing provided for the prior art;

[0047] Figure 2 A flowchart of a method for planning a multi-region task provided for an embodiment of the present application;

[0048] Figure 3 A schematic diagram of establishing the regional connection relationship of Region 5 provided for an embodiment of the present application;

[0049] Figure 4 A schematic diagram of establishing the regional connection relationship of Region 2 provided for an embodiment of the present application;

[0050] Figure 5 A schematic diagram of Region 4 returning to Region 2 provided for an embodiment of the present application;

[0051] Figure 6 A schematic diagram of establishing the regional connection relationship of Region 3 provided for an embodiment of the present application;

[0052] Figure 7 A schematic diagram of Region 3 returning to Region 5 provided for an embodiment of the present application;

[0053] Figure 8 A schematic diagram of a weighted graph provided for an embodiment of the present application;

[0054] Figure 9 A schematic diagram of the placement position of channel labels provided for an embodiment of the present application;

[0055] Figure 10 A schematic diagram of the structure of a multi-region task planning device provided for an embodiment of the present application;

[0056] Figure 11 A schematic structural diagram of an electronic device provided by an embodiment of the present application. Specific embodiments

[0057] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts shall fall within the scope of protection of the present application.

[0058] The following disclosure provides many different embodiments or examples for implementing different structures of the present application. To simplify the disclosure of the present application, components and settings of specific examples are described below. Of course, they are merely examples and are not intended to limit the present application. In addition, the present application may repeat reference numerals and / or letters in different examples. Such repetition is for the purpose of simplification and clarity and does not itself indicate the relationship between the various embodiments and / or settings discussed.

[0059] To solve the problem that multi-region tasks cannot be planned and executed when there is a lack of global positioning and a global map as mentioned in the background art, the robot scans the channel tags along the edge to construct the regional connection relationship, records the time consumed for passing through the channels during the process of exploring all un-scanned regions, and finally generates a weighted graph to plan multi-region tasks.

[0060] The embodiments of the present application are mainly applied to the task planning of multiple regions, and the application scenarios include but are not limited to: lawn management, house cleaning, shelf object handling, farmland management, etc.

[0061] A method for planning a multi-region task in the embodiments of the present application can be executed by a robot, specifically by a monocular vision robot lacking multiple sensors. The following will combine specific embodiments to provide a detailed description of a method for planning a multi-region task provided by the embodiments of the present application. Taking the application to a robot as an example, as Figure 2 shown, the specific steps are as follows:

[0062] Step 201: Control the robot to scan all the channel tags along the edge in the current region, where the channel tags are set at the channel ports between every two adjacent regions;

[0063] Step 202: Construct the connection relationship between the current region and the adjacent regions by reading the number information in the channel tags, and determine the number of the current region, where the number information is composed of the current region number and the number of the adjacent region to which it leads;

[0064] Step 203: After determining the target channel based on the connection relationship, control the robot to travel through the target channel to an un-scanned adjacent area, and record the time taken to pass through the target channel;

[0065] Step 204: After traversing all areas, generate a weighted graph based on the connection relationship between areas and the time taken to pass through channels, and plan multi-area tasks according to the weighted graph. Each vertex of the weighted graph represents an area, the number of the vertex is the number of the corresponding area, and the weight of each edge represents the time taken to pass through the channel.

[0066] Some terms mentioned in the embodiments of the present application are explained below, including the following content.

[0067] Channel label: Set at the channel port between every two adjacent areas, containing number information (e.g., "1-3" means from area 1 to area 3).

[0068] Number information: Consists of two groups of numbers. The first group of numbers is the current area number, and the second group of numbers is the number of the adjacent area to which it leads.

[0069] In step 201, after the robot enters a new area, it moves along the edge of the area and uses the camera equipped on it to scan all channel labels set at the channel ports in the area. Each channel label is placed at the entrance and exit positions of the channel for easy identification and scanning by the robot. These channel labels are used to identify the connection relationship between adjacent areas. In this way, the robot can comprehensively obtain all exit information of the current area.

[0070] In step 202, during the scanning process, the visual system of the robot identifies the channel label and obtains the number information in the channel label. The number information of each channel label consists of two groups of numbers. The first group is the current area number, and the second group is the number of the adjacent area to which it leads (for example, "1-3" means from area 1 to area 3). According to the scanned number information, the robot can determine the number of the current area and can also construct the connection relationship between the current area and other adjacent areas. For example, if "1-3" is scanned, it means that the number of the current area is 1, and it is recorded that there is a channel between area 1 and area 3.

[0071] Among them, the content on the channel label can be a two-dimensional code representing number information or a number representing number information. By scanning the channel label to obtain the number information, the robot can clearly know which areas around the current area have connection channels, that is, understand the topological structure between the current area and adjacent areas, so as to establish an initial area connection relationship graph.

[0072] In step 203, once all the adjacent regions of the current region are identified and recorded, that is, after the connection relationship of the current region is determined, the robot determines the channels between the current region and each adjacent region according to the connection relationship, then selects a target channel from the un-traversed channels, and then controls itself to travel along the target channel to the un-scanned adjacent region. When entering a new adjacent region, the robot records the time required to pass through the target channel (i.e., the time consumption). This time recording not only helps to generate a weighted graph, but also provides a time reference for actual channel walking, further optimizing subsequent task planning.

[0073] Among them, the robot can select channels and traverse regions through a strategy, which can be a depth-first strategy, a breadth-first strategy, or a random strategy. These three strategies will be described in detail below and will not be elaborated here.

[0074] In step 204, as the robot continuously shuttles and explores between regions, after all regions are traversed, the robot generates a comprehensive and intuitive weighted graph based on the previously recorded regional connection relationships and the time consumption data for passing through each channel. In the weighted graph, each vertex represents a region, and the vertex numbers correspond one-to-one with the actual region numbers, facilitating region identification and management. The edges connecting the vertices in the weighted graph represent the channels between adjacent regions, and the weight of the edge reflects the time consumed by the robot passing through this channel. The larger the weight value, the longer the time required to pass through this channel; conversely, the smaller the weight, the shorter the passing time.

[0075] After generating the weighted graph, the robot uses built-in algorithm modules (such as greedy algorithm, dynamic programming algorithm, or brute-force search algorithm, etc.) to analyze and calculate the weighted graph to plan multi-region tasks. During the planning process, the algorithm comprehensively considers the connection relationships between regions and the channel weights, determines the optimal multi-region planning order, and generates a detailed task execution plan, such as including the mowing sequence of each region and the movement path between regions.

[0076] In this application, by setting channel labels containing number information at the channel entrances, the robot can determine the connection relationships between adjacent regions and the number of the current region by scanning the channel labels along the edge, then select a target channel according to the connection relationship, and traverse all regions without omission along the target channel to construct a complete regional connection relationship graph. The robot also records the time consumption of passing through the channels, and finally generates a weighted graph based on the regional connection relationships and channel time consumption. By planning multi-region tasks through the weighted graph, this application omits the complex multi-sensor fusion positioning and map construction processes, and can directly plan multi-region tasks according to the weighted graph, ensuring that the robot can plan multi-region tasks without global positioning and a global map.

[0077] In the embodiments of this application, the strategies that the robot can adopt when selecting channels and traversing areas include but are not limited to the following three.

[0078] 1. Depth-first strategy. During the process of exploring an area, the depth-first strategy ensures that the robot explores each area as deeply as possible in one direction until it cannot continue to go deeper, and then backtracks to the previous area to continue exploring other areas, avoiding missing any area. The depth-first strategy can ensure quickly diving into a certain area, reducing repeated exploration. When the area connection is relatively complex, it can efficiently construct area connection relationships, avoid wandering repeatedly in the explored areas, and save exploration time and resources.

[0079] 2. Breadth-first strategy. The breadth-first strategy means taking the current area as the center and accessing the un-scanned adjacent areas in the order of distance from near to far. That is, it preferentially accesses the un-scanned areas directly connected to the current area, and then accesses the adjacent un-scanned areas of these areas, spreading out layer by layer. The breadth-first strategy can explore each area evenly, but when the area connection is complex and the range is large, due to the need to continuously process a large number of short-distance channels, the exploration efficiency is low.

[0080] 3. Random strategy. When facing multiple un-traversed channels, the robot randomly selects a channel to travel to an un-scanned adjacent area, without relying on a specific order or rule, and each time the channel selection is determined by a random mechanism. This exploration process can also complete the task, but it lacks systematicness, may lead to repeated exploration of the same area multiple times, and the task completion efficiency is low.

[0081] Therefore, based on the above strategy analysis, the embodiments of this application preferentially select the depth-optimal strategy, that is, determine the target channel based on the connection relationship and the depth-first strategy, and control the robot to travel through the target channel to an un-scanned adjacent area according to the depth-first strategy. The process of selecting the target channel and traversing the area using the depth-first strategy includes the following:

[0082] Step S11: When it is determined based on the connection relationship that there is at least one un-traversed channel in the current area, control the robot to randomly or nearby select a target channel from at least one un-traversed channel, and travel through the target channel to an un-scanned adjacent area;

[0083] Step S12: Take the un-scanned adjacent area as the current area to construct an area connection relationship, and repeat the above operation to continuously dive into the un-scanned adjacent area, where the travel path of the robot is stored in the database;

[0084] Step S13: Among them, if there is only one traversed channel in the current area, return to the previous area according to the stored travel path;

[0085] Step S14: If there is an unvisited passage in the previous area, execute the step of selecting and passing through the target passage in the unvisited passages to move to an un-scanned adjacent area;

[0086] Step S15: If there is no unvisited passage in the previous area, perform area backtracking according to the stored travel path until an area with an unvisited passage is found.

[0087] During the multi-area exploration of the robot, the depth-first strategy is the core method to guide its efficient traversal of each area. The specific steps are as follows.

[0088] When the robot is in a certain current area, it will first detect the passage conditions in this area based on the area connection relationship. If it is found that there is an unvisited passage in the current area, which means there are still new areas waiting to be explored, then the robot will select a target passage from the unvisited passages and control itself to move forward along this target passage to an un-scanned adjacent area.

[0089] If the number of unvisited passages is only one, then directly use this unvisited passage as the target passage; if the number of unvisited passages is more than one, then the robot will randomly or nearby select one of the unvisited passages as the target passage. The specific methods of randomly or nearby selecting passages are as follows.

[0090] When adopting the random selection strategy, the robot will start a random number generator. This generator will generate a corresponding random index value according to the number of passages in the unvisited passage list. The random index value is similar to a random pointer, pointing to a certain passage in the unvisited passage list. Through this random index, the robot selects a passage from the list as the target passage. When the robot has no specific exploration preference for each unvisited area, it can avoid a certain fixed selection pattern, making the exploration process more evenly cover each area and reducing the problem of exploration delay in some areas caused by a fixed selection order.

[0091] If the target passage is selected according to the principle of proximity, the robot will first use its distance sensor to measure the distance between the entrance of each unvisited passage and the current position of the robot. These distance data will be collected and sorted, and the robot will preferentially select the unvisited passage closest to the current position as the target passage. This selection method is very practical in practical applications, especially in the case of limited energy or tight time. For example, for a lawn mowing robot, selecting a passage with a short distance can reduce the energy and time consumed in moving between passages, improve the overall work efficiency, and enable it to cover more areas for lawn mowing operations in a shorter time.

[0092] Once the robot enters a new adjacent area, a series of information recording operations will be triggered. The robot will store the information of the current area it is in, the specific information of the passage it has just passed through, and the newly scanned area information as the travel path in the database. The travel path completely records the exploration process of the robot in the order in which the information enters. After that, the robot will set the newly entered and unscanned area as the current area, build a new area connection relationship based on this, and then check whether there are un-traversed passages in the new current area. If there are, it will repeat steps S21 and S22, continuously penetrate into more unscanned adjacent areas, and try to explore along a path as far as possible, constantly exploring unknown areas.

[0093] When the robot discovers that there is only one traversed passage in the current area, it indicates that it is no longer possible to continue exploring new areas along the current path, and a backtracking operation is required. At this time, the robot will accurately determine the position of the previous area based on the information recorded in the travel path, and then the robot will control itself to return to the previous area along the original path.

[0094] After the robot returns to the previous area, it will immediately check the passage conditions of this area. If it is found that there are un-traversed passages in the previous area, then the robot will perform the operation of traveling through this un-traversed passage to an unscanned adjacent area again, that is, repeat step S21 and subsequent processes to continue exploring new areas. Similar to when returning to the previous fork in the road and finding that there is still a road that has not been taken before, so continue to move forward along this road to explore.

[0095] If the robot discovers that there are also no un-traversed passages in the previous area, it means that there are no new exploration areas on this branch path and further backtracking is required. The robot will continue to obtain information about the area one level above from the travel path and control itself to return again. This process will be repeated continuously until the robot finds an area with un-traversed passages. Through such a continuous backtracking mechanism, the robot can comprehensively and systematically explore each area in the entire multi-area environment, avoiding missing any possible paths.

[0096] In this application, the key role of the depth-first strategy is that once a channel is selected, the robot will continuously explore deeply along this channel until it reaches an area where there are no new channels connected, or all channels in this area have been traversed. This strategy ensures that during the exploration process, the robot can penetrate as deeply as possible in one direction instead of just skimming the surface or frequently switching the exploration direction, avoiding disorderly and chaotic exploration situations that would lead to reduced efficiency. In addition, the backtracking mechanism is an important part of the depth-first search strategy, which ensures that the robot does not miss any area during the exploration process. When the exploration of a certain area is temporarily blocked, by backtracking to the previous area, the robot can continue to explore other unscanned areas, ensuring the integrity of the area connection relationship graph.

[0097] The robot can reasonably plan the exploration path through the depth-first strategy and the corresponding backtracking operations, complete the exploration tasks of multiple areas, facilitate the establishment of the connection relationship between multiple areas, and provide a data basis for subsequent multi-area task planning.

[0098] Exemplarily, there is a working environment containing Area 1, Area 2, Area 3, Area 4, and Area 5, and the robot is initially located in Area 5.

[0099] 1. As Figure 3 shown, the robot scans along the edge in Area 5 and discovers two channel labels. One channel label is numbered 5-1 (indicating the passage from Area 5 to Area 1), and the other is numbered 5-2 (indicating the passage from Area 5 to Area 2). The robot records the connection relationship between Area 5 and Area 1 and Area 2, and constructs a preliminary connection relationship.

[0100] 2. As Figure 4 shown, since there are un-traversed channels in Area 5, the robot randomly selects the channel leading to Area 2 and advances to Area 2 using the depth-first strategy. After reaching Area 2, Area 2 becomes the current area. The robot scans along the edge in Area 2 and discovers the channel labels 2-5 (already traversed, ignored), 2-3, and 2-4. The robot constructs the connection relationship between Area 2 and Area 3 and Area 4.

[0101] 3. As Figure 5 shown, since there are un-traversed channels in Area 2, the robot randomly selects the channel leading to Area 4 and advances to Area 4. Area 4 becomes the current area. The robot scans the channel labels in Area 4 and discovers that there is only the already traversed channel 4-2 (already traversed, ignored) in Area 4, and the robot returns to Area 2.

[0102] 4. As Figure 6As shown, in Area 2, only the un-traversed passage 2-3 exists. The robot selects the passage leading to Area 3 and travels to Area 3. Area 3 becomes the current area. The robot scans the passage tags in Area 3 and discovers 3-2 (already traversed, ignored), 3-1, and constructs the connection relationship between Area 3 and Area 1.

[0103] 5. As Figure 7 shown, since only the un-traversed passage 3-1 exists in Area 3, the robot selects the passage leading to Area 1. The robot scans the passage tags in Area 1 and discovers 1-3 (already traversed, ignored), 1-5. The robot selects the passage leading to Area 5 and travels to Area 5. There are no adjacent areas in Area 5 that have not been scanned. At this time, all areas have been traversed.

[0104] 6. During the traversal process, the robot records the time taken to walk through each passage. Based on these times, the robot generates a weighted graph, as Figure 8 shown, where Area 1, Area 2, Area 3, Area 4, and Area 5 are vertices, the passages connecting them are edges, and the weights of the edges are the time taken to walk through the corresponding passages. The weighted graph is analyzed using the greedy algorithm to plan the optimal mowing order, such as Area 5 -> Area 2 -> Area 4 -> Area 3 -> Area 1, completing the multi-area task planning.

[0105] Optionally, in Figure 5 if the robot preferentially travels from Area 2 to Area 3, then the path is Area 2 -> Area 3 -> Area 1 -> Area 5. At this time, there are no un-traversed passages in Area 5. At this time, there is still Area 4 that has not been scanned. At this time, the robot will reach Area 4 through Area 5 -> Area 1 -> Area 3 -> Area 2 -> Area 4 to complete the exploration of Area 4. At this time, the exploration of all areas has been completed.

[0106] As an optional implementation manner, determining that there is at least one un-traversed passage in the current area based on the connection relationship includes the following:

[0107] Step S21: Determine all the passages connected to the current area based on the connection relationship;

[0108] Step S22: Determine the traversed passages according to the travel path in the database;

[0109] Step S23: Compare all the passages connected to the current area with the traversed passages to determine at least one un-traversed passage existing in the current area.

[0110] The robot utilizes the constructed regional connection relationships to identify all the channels connecting the current region to other surrounding regions. This step provides a comprehensive data basis for subsequent screening of unvisited channels, ensuring that no possible channels are overlooked. Additionally, during the task execution, the robot records the information of the channels it has passed through in real-time into the travel path. By analyzing the travel path stored in the database, the robot can accurately determine which channels have been traversed, that is, record the visited channels. The robot compares the information obtained in the previous two steps, checking each of the channels connected to the current region against the visited channels one by one. If it is found that a certain channel does not exist in the list of visited channels, then that channel is an unvisited channel. This comparison mechanism can precisely screen out unvisited channels, providing a basis for the robot's subsequent planning, enabling it to targetedly move to new regions for scanning, avoiding passing through the already traveled channels repeatedly, and improving work efficiency.

[0111] As an optional implementation, determining at least one unvisited channel existing in the current region by comparing all the channels connected to the current region with the visited channels includes the following.

[0112] Step S31: Determine the channel identifiers of all the channels connected to the current region by identifying the number information of all the channel tags in the current region, where the number information is used to indicate the channel identifier between two regions;

[0113] Step S32: Determine the set of channel identifiers of the visited channels by analyzing the travel path;

[0114] Step S33: Compare all the channel identifiers corresponding to the current region with the set of channel identifiers of the visited channels;

[0115] Step S34: If there is a set channel identifier in the channel identifiers of the current region that does not appear in the set of channel identifiers, then regard the channel corresponding to the set channel identifier as an unvisited channel.

[0116] The robot uses the vision recognition system carried by itself to scan the channel labels at all channel ports in the current area. By identifying and analyzing these numbered information, it can determine the channel identifiers of all channels connected in the current area. The robot's database also details the travel paths during its task execution. Through in-depth analysis of these travel path data, the robot can extract the channel identifiers of the traversed channels. For example, in the travel path record, the robot travels from area 2 through the channel numbered "2-4" to area 4, then the channel identifier "2-4" is included in the set of channel identifiers of the traversed channels. As the robot continues to travel, the set of channel identifiers of the traversed channels is constantly updated and expanded, providing accurate data support for subsequent comparison. The robot compares all the channel identifiers corresponding to the current area with the set of channel identifiers of the traversed channels one by one. If a certain channel identifier cannot find a matching item in the set of channel identifiers of the traversed channels during the comparison process, then the channel corresponding to this channel identifier is an untraversed channel. For example, the channel identifier "1-6" of the current area does not appear in the set of channel identifiers of the traversed channels, then the channel from area 1 to area 6 is determined to be an untraversed channel.

[0117] In this application, by analyzing the travel paths stored in the database, the set of channel identifiers of the traversed channels is obtained and compared with all the corresponding channel identifiers in the current area, so as to accurately screen out the untraversed channels. After identifying the untraversed channels, the robot can explore them specifically without wasting time and resources on the channels that have already been walked.

[0118] As an optional implementation manner, recording the time taken to walk through the target channel includes the following contents:

[0119] Step S41: When the robot switches areas, record the first moment when the channel label at the first port of the target channel is scanned;

[0120] Step S42: When the robot travels through the target channel and reaches the next adjacent area that has not been scanned, record the second moment when the channel label at the second port of the target channel is scanned;

[0121] Step S43: Take the time difference between the second moment and the first moment as the time taken for the robot to pass through the target channel.

[0122] When the robot is about to switch regions, when its vision recognition system scans the channel label at the first port of the target channel, the high-precision timing module inside the robot will quickly capture this moment and accurately record the time at this moment as the first moment. The robot then starts the movement program and travels along the selected target channel towards the adjacent un-scanned area. During the movement, when the robot successfully passes through the target channel, reaches the next un-scanned adjacent area, and scans the channel label at the second port of the target channel, the timing module will start again and record the time at this moment as the second moment. The operation module of the robot will call the stored first moment and second moment data, subtract the second moment from the first moment, and thus obtain the time taken for the robot to walk in this target channel.

[0123] Optionally, a detection distance threshold can be set in the robot's vision detection system. When the detected distance from the channel label is greater than the set threshold, it is not used as the starting time recording point. Instead, wait for the robot to approach within the threshold range and then record the starting time. For example, if the set threshold is 0.5 meters, when the robot detects the channel label within 0.5 meters of the channel label, it starts to record the time, which can reduce the error caused by long-distance detection.

[0124] For example, in a multi-region mowing scenario, when the robot switches from region A to region B, when it reaches the starting end of the channel connecting region A and region B and scans the channel label at this port, assuming the time is 10:00, the timing module will accurately store this moment in its data storage unit. When the robot reaches region B and scans the channel label at the other end of the channel, the time is 10:01. Subtracting 10:00 from 10:01, it is obtained that the time taken for the robot to walk in this channel is 1 minute.

[0125] This application provides key quantitative data for generating a weighted graph by accurately recording the time taken for the robot to walk in the target channel. In the weighted graph, the weight of an edge represents the time cost for the robot to pass through the corresponding channel. The accurate time data enables the weighted graph to more realistically reflect the difficulty of passage and time consumption between regions.

[0126] Among them, for the same channel, the order of the number information contained in the channel labels at the two ports of this channel is swapped. For example, Figure 9In it, the channel label number information at one end of the channel is 1-2, indicating that one can go from area 1 to area 2 through this channel; while the channel label number information at the other end of the channel is 2-1, that is, one can reach area 1 from area 2 through this channel. Although the order of the number information is different, they correspond to the same time-consuming data for the robot to walk back and forth on this channel. This means that whether the robot enters the channel from area 1 to go to area 2 or enters the channel from area 2 to go to area 1, the recorded time-consuming for walking on this channel is the same. In this way, the robot does not need to separately record and maintain the time-consuming for going back and forth on the same channel, reducing the data storage volume.

[0127] In the embodiments of the present application, the channel label is placed at the edge position of the area, which can ensure that the robot can stably see the surface of the channel label during the process of scanning along the edge in the current area and walking between areas through the channel. The channel label can be set as a triangular double-sided structure to ensure that the robot can stably see the vertical surface of the triangular channel label, or can also be set as a trapezoidal double-sided structure, a cuboid structure, a cylinder structure, etc. The present application does not limit the shape of the channel label.

[0128] To ensure that the lawn mower can efficiently and accurately identify area information in a complex working environment, at least two identical number information are provided on the channel label. One of the number information faces the moving direction of the robot at the edge where the channel label is located. When the lawn mower moves along the edge of the area, it can always ensure that one number information is facing its moving direction. Without additional adjustment of the perspective, it can easily read the number information and can also record the first moment when the channel label is scanned in a timely manner. The other number information is set to face the channel direction. When the robot passes through the channel and reaches another area, it can immediately record the second moment when it reaches this area.

[0129] Optionally, the embodiments of the present application can analyze and calculate the weighted graph by using the algorithm module built in the robot (such as the greedy algorithm, the dynamic programming algorithm or the brute-force search algorithm, etc.) to plan the multi-area task. The analysis processes of the greedy algorithm and the brute-force search algorithm are as follows.

[0130] The process of planning the multi-area task by using the greedy algorithm is as follows: 1. Start from any area, and this area is used as the current area. 2. Check the weights of the connecting edges between the current area and all adjacent areas, and select the area connected by the edge with the smallest weight as the next area to go to. This is the core of the greedy algorithm, and the optimal solution in the current state (that is, the path with the smallest cost) is selected each time. 3. Update the selected next area as the current area and mark that this area has been visited. 4. Repeat steps 2 and 3 until all areas have been visited. 5. According to the order of the visited areas, form the final multi-area task planning path.

[0131] The process of planning a multi-region task using the brute-force search algorithm is as follows: 1. Determine the total number of regions in the multi-region task, as well as the connection relationships between regions and the weights of edges in the weighted graph. 2. Generate all possible region access sequences by permuting the numbers assigned to multiple vertices in the weighted graph. A region access sequence is a set of numbered sequences that indicate the order of accessing regions. Assuming there are n regions, there are n! possible access orders. For example, if there are 3 regions A, B, and C, the possible orders are A-B-C, A-C-B, B-A-C, B-C-A, C-A-B, and C-B-A. 3. For each generated region access order, accumulate the weights of the corresponding edges in the weighted graph to calculate the total weight of accessing all regions in that order. For example, for the access order A-B-C, if the weight from A to B is m and the weight from B to C is n, the total weight is m + n. 4. Compare all the calculated total weights to find the access order with the minimum (or maximum according to the task requirements) total weight. This optimal access order is the execution order of the multi-region task planned using the brute-force search algorithm. 5. Output the obtained optimal region access order as the planning result of the multi-region task.

[0132] Planning a multi-region task using the brute-force search algorithm includes: permuting the numbers assigned to multiple vertices in the weighted graph to generate multiple region access sequences, where a region access sequence is a set of numbered sequences that indicate the order of accessing regions; accumulating the weights of the corresponding edges in the weighted graph according to the region access order in the region access sequence to obtain the total weight of the edges of the region access sequence; comparing the total weights of the multiple region access sequences to determine the access order corresponding to the extreme total weight, and planning the multi-region task according to the access order, where the extreme total weight is the maximum total weight or the minimum total weight.

[0133] This application provides an overall process for planning a multi-region task, including the following steps.

[0134] 1. Robot deployment and system initialization: Place the robot in one of the working areas, start its vision system, and initialize the data structure for storing region connection relationships, such as creating an empty list to prepare for recording channel information.

[0135] 2. Edge scanning along the current region: The robot scans along the edge of the current region, and through the vision system, identifies the channel labels at the edge and the numbered information in each channel label.

[0136] 3. Completing the construction of the connection relationship of the current region: The robot continuously scans along the edge of the current region until all channel labels of the current region are identified, constructs a connection relationship graph between the current region and at least one adjacent region, clarifies the channel connection situation between regions, and determines the number of the current region.

[0137] 4. Determine un-traversed channels in the current area: The robot compares all the channels in the current area with the traversed channels based on the channel identifiers to find the un-traversed channels in the current area.

[0138] 5. Select a target channel and move to a new area: The robot randomly or proximately selects one of the un-traversed channels as the target channel, then enters the next un-scanned adjacent area through the target channel and records the time taken to pass through the target channel.

[0139] 6. Continuously traverse the new area: In the newly entered un-scanned area, the robot repeats operations such as scanning along the edge, constructing the area connection relationship, and determining un-traversed channels.

[0140] 7. Return to the previous area and continue exploration: If there are only traversed channels in the newly entered current area, the robot returns to the previous area, checks if there are any un-scanned adjacent areas in the returned area. If there are, it performs the step of selecting a target channel and moving to the un-scanned adjacent area; if not, it continues to backtrack to the previous area. Through continuous exploration, all areas are traversed.

[0141] 8. Generate a weighted graph: After traversing all areas, the robot assigns weights to each channel between areas based on the time taken to walk through each channel, and then generates a weighted graph with areas as vertices and channels as edges according to the area connection relationship. The vertex numbers in the weighted graph are the same as the corresponding area numbers.

[0142] 9. Plan multi-area tasks: The robot uses built-in algorithms such as the greedy algorithm and dynamic programming algorithm to analyze the weighted graph and determine the optimal execution path and order of multi-area tasks in combination with task requirements.

[0143] Based on the same technical concept, this application provides a planning device for multi-area tasks, as Figure 10 shown. The device includes:

[0144] A scanning module 1001, configured to control the robot to scan all channel labels along the edge in the current area, where the channel labels are set at the channel ports between every two adjacent areas;

[0145] A construction and determination module 1002, configured to construct the connection relationship between the current area and adjacent areas by reading the number information in the channel labels and determine the number of the current area, where the number information consists of the current area number and the number of the adjacent area to which it leads;

[0146] A travel and recording module 1003, configured to control the robot to travel through the target channel to an un-scanned adjacent area based on the connection relationship and record the time taken to pass through the target channel;

[0147] A generation and planning module 1004, configured to generate a weighted graph based on the connection relationships between regions and the time taken to pass through channels after traversing all regions, and plan multi-region tasks according to the weighted graph, where each vertex of the weighted graph represents a region, the number of the vertex is the number of the corresponding region, and the weight of each edge represents the time taken to pass through the channel.

[0148] Optionally, the movement and recording module 1003 is configured to:

[0149] After determining the target channel based on the connection relationship and the depth-first strategy, control the robot to move through the target channel to an un-scanned adjacent region according to the depth-first strategy.

[0150] Optionally, the movement and recording module 1003 is specifically configured to:

[0151] When it is determined that there is at least one un-traversed channel in the current region based on the connection relationship, control the robot to randomly or proximally select a target channel from at least one un-traversed channel, and move through the target channel to an un-scanned adjacent region;

[0152] Take the un-scanned adjacent region as the current region to construct the region connection relationship, and repeat the above operations to continuously penetrate into the un-scanned adjacent region, where the movement path of the robot is stored in the database;

[0153] Wherein, if there is only one traversed channel in the current region, return to the previous region according to the stored movement path;

[0154] If there is an un-traversed channel in the previous region, execute the step of selecting and moving through the target channel in the un-traversed channel to an un-scanned adjacent region;

[0155] If there is no un-traversed channel in the previous region, perform region backtracking according to the stored movement path until a region with an un-traversed channel is found.

[0156] Optionally, the movement and recording module 1003 is specifically configured to:

[0157] Determine all channels connected to the current region based on the connection relationship;

[0158] Determine the traversed channels according to the movement path in the database;

[0159] Compare all channels connected to the current region with the traversed channels to determine at least one un-traversed channel existing in the current region.

[0160] Optionally, the movement and recording module 1003 is specifically configured to:

[0161] By identifying the number information of all channel tags in the current area, determine the channel identifiers of all channels connected to the current area, where the number information is used to indicate the channel identifiers of the channels between two areas;

[0162] By analyzing the travel path, determine the set of channel identifiers of the traversed channels;

[0163] Compare all the channel identifiers corresponding to the current area with the set of channel identifiers of the traversed channels;

[0164] If there is a set channel identifier in the channel identifiers of the current area that does not appear in the set of channel identifiers, then regard the channel corresponding to the set channel identifier as an untraversed channel.

[0165] Optionally, the travel and recording module 1003 is further configured to:

[0166] When the robot switches areas, record the first moment when the channel tag at the first port of the target channel is scanned;

[0167] When the robot travels through the target channel to the next unscanned adjacent area, record the second moment when the channel tag at the second port of the target channel is scanned;

[0168] Take the time difference between the second moment and the first moment as the time taken for the robot to pass through the target channel.

[0169] Optionally, the generation and planning module 1004 is configured to:

[0170] Arrange and combine the numbers of multiple vertices in the weighted graph to generate multiple area access sequences, where the area access sequence is used to indicate the set of number sequences for accessing areas in order;

[0171] Accumulate the weights of the corresponding edges in the weighted graph according to the area access order in the area access sequence to obtain the total weight of the edges of the area access sequence;

[0172] Compare the total weights of the multiple area access sequences, determine the access order corresponding to the extreme total weight, and plan the multi-area task according to the access order, where the extreme total weight is the maximum total weight or the minimum total weight.

[0173] As Figure 11 shown, an embodiment of the present application provides an electronic device, including a processor 1101, a communication interface 1102, a memory 1103, and a communication bus 1104, where the processor 1101, the communication interface 1102, and the memory 1103 communicate with each other through the communication bus 1104.

[0174] The memory 1103 is used to store a computer program.

[0175] In one embodiment of the present application, when the processor 1101 executes the program stored in the memory 1103, it implements the multi-region task planning method provided in any of the foregoing method embodiments.

[0176] The electronic device provided in the embodiments of the present application may specifically be a monocular vision robot capable of performing edge scanning and task planning.

[0177] The embodiments of the present application also provide a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the steps of the multi-region task planning method provided in any of the foregoing method embodiments.

[0178] 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 may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0179] 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 part that contributes to the related technologies can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disc, etc., including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0180] It should be understood that the terms used herein are only for the purpose of describing specific example embodiments and are not intended to be limiting. Unless otherwise clearly specified in the context, the singular forms "a", "an", and "the" used herein may also include the plural forms. The terms "include", "comprise", "contain", and "have" are inclusive and thus specify the presence of the stated features, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, elements, components, and / or their combinations. The method steps, processes, and operations described herein are not to be construed as necessarily requiring them to be executed in the specific order described or illustrated, unless the execution order is clearly indicated. It should also be understood that additional or alternative steps may be used.

[0181] The above are only specific embodiments of the present application, enabling those skilled in the art to understand or implement the present application. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to these embodiments shown herein, but rather to the broadest scope consistent with the principles and novel features claimed herein.

Claims

1. A multi-region task planning method, characterized in that: The method comprises: Control the robot to scan all channel labels in the current area along the edge, wherein the channel labels are set at the channel ports between every two adjacent areas; Building a connection relationship between the current area and the adjacent area by reading the number information in the channel tag, and determining the number of the current area, wherein the number information consists of the current area number and the adjacent area number to which the current area leads; After determining a target channel based on the connection relationship, controlling the robot to move through the target channel to an unscanned adjacent area, and recording the time taken to pass through the target channel; After traversing all areas, a weighted graph is generated based on the connection relationship between areas and the time consumed to pass through the channel, and multi-area tasks are planned according to the weighted graph, where each vertex of the weighted graph represents an area, the vertex number is the number of the corresponding area, and the weight of each edge represents the time consumed to pass through the channel.

2. The method according to claim 1, characterized in that After determining the target channel based on the connection relationship, controlling the robot to travel through the target channel to an unscanned adjacent area includes: After determining the target channel based on the connection relationship and the depth-first strategy, the robot is controlled to move through the target channel to an unscanned adjacent area according to the depth-first strategy.

3. The method according to claim 2, characterized in that After determining the target channel based on the connection relationship and the depth-first strategy, controlling the robot to travel through the target channel to an unscanned adjacent area according to the depth-first strategy includes: In a case where it is determined based on the connection relationship that there is at least one untraversed channel in the current area, controlling the robot to randomly or nearest select a target channel from the at least one untraversed channel, and travel to an unscanned adjacent area through the target channel; Taking the unscanned adjacent area as the current area to construct an area connection relationship, repeating the above operation to continuously go deeper into the unscanned adjacent area, wherein the travel path of the robot is stored in a database; If there is only one traversed channel in the current area, return to the previous area according to the stored travel path; If there is an untraversed channel in the previous area, performing the steps of selecting and traveling to an unscanned adjacent area through a target channel among the untraversed channels; If there is no untraversed passage in the previous area, the area is backtracked according to the stored travel path until an area with an untraversed passage is found.

4. The method according to claim 3, characterized in that Determining that there is at least one untraversed channel in the current area based on the connection relationship includes: Determine all channels connected to the current area based on the connection relationship; determining a traversed channel based on the travel path in the database; All channels connected by the current region are compared with the traversed channels to determine at least one untraversed channel existing in the current region.

5. The method according to claim 4, characterized in that Comparing all channels connected by the current area with the traversed channels to determine at least one untraversed channel in the current area includes: Determine the channel identifiers of all channels connected to the current area by identifying the numbering information of all channel labels in the current area, wherein the numbering information is used to indicate the channel identifiers of channels between two areas; Determining a set of channel identifiers of traversed channels by analyzing the travel path; Comparing all channel identifiers corresponding to the current area with the channel identifier set of the traversed channels; If there is a set channel identifier in the channel identifiers of the current area that does not appear in the channel identifier set, the channel corresponding to the set channel identifier is regarded as an untraversed channel.

6. The method according to claim 1, characterized in that The time taken to record the target channel includes: When the robot switches regions, recording the first moment when the channel label at the first port of the target channel is scanned; When the robot passes through the target channel and moves to the next unscanned adjacent area, recording a second moment when the channel label at the second port of the target channel is scanned; The time difference between the second moment and the first moment is used as the time taken for the robot to pass through the target channel.

7. The method according to claim 1, characterized in that Planning a multi-region task according to the weighted graph includes: Arrange and combine a plurality of vertex numbers in the weighted graph to generate a plurality of region access sequences, wherein the region access sequences are used to indicate a set of number sequences for sequentially accessing regions; Accumulating the weights of corresponding edges in the weighted graph according to the region access order in the region access sequence to obtain a total weight of the edges in the region access sequence; The total weights of the plurality of region access sequences are compared to determine an access order corresponding to an extreme total weight, and a multi-region task is planned according to the access order, wherein the extreme total weight is the maximum total weight or the minimum total weight.

8. A multi-region task planning device, characterized in that: The device comprises: A scanning module, used to control the robot to scan all channel labels in the current area along the edge, wherein the channel labels are set at the channel ports between every two adjacent areas; a construction and determination module, configured to construct a connection relationship between the current area and the adjacent area by reading the number information in the channel tag, and determine the number of the current area, wherein the number information consists of the current area number and the adjacent area number to which the current area leads; a traveling and recording module, configured to, after determining a target channel based on the connection relationship, control the robot to travel through the target channel to an unscanned adjacent area, and record the time taken to pass through the target channel; The generation and planning module is used to generate a weighted graph based on the connection relationship between regions and the time consumed to pass through the channel after traversing all regions, and plan multi-region tasks according to the weighted graph, wherein each vertex of the weighted graph represents a region, the vertex number is the number of the corresponding region, and the weight of each edge represents the time consumed to pass through the channel.

9. An electronic device, characterized in that: It includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus; Memory for storing computer programs; A processor, configured to implement the method according to any one of claims 1 to 7 when executing a program stored in a memory.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images