Intelligent room cleaning method and device based on local community discovery
A cleaning robot with a camera and mechanical arm uses local community discovery to classify and prioritize objects for storage or reset, improving the efficiency and intelligence of room cleaning tasks.
Patent Information
- Application Number
- CN202411445987.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-16
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2044-10-16
AI Technical Summary
Existing local community discovery methods are too complicated and have not been effectively applied to room cleaning, as they are too complicated and have not been effectively applied to solve the complexity of room cleaning tasks.
A method and device using a movable cleaning robot with a camera and mechanical arm that employs local community discovery to identify and process objects in a room by classifying them into nodes to be stored or reset, constructing local communities based on priority, and optimizing the reset process.
This approach enhances the efficiency and intelligence of room cleaning by effectively applying local community discovery, ensuring all objects are stored or reset efficiently.
Smart Images

Figure CN119055153A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of smart home appliances, and more specifically, to: 1. a room smart cleaning method based on local community discovery; 2. a room cleaning device based on local community discovery. Background Art
[0002] With the development of science and technology, the work of room cleaning has shown a trend towards intelligent development, such as sweeping robots. There are also some devices for intelligent room cleaning, such as mobile cleaning robots with mechanical arms, but most of these robots rely on manual control and their functions are not perfect.
[0003] Among them, local community discovery is an important technology aimed at discovering community structures in complex networks such as social networks, transportation networks, citation networks and biological networks, which is very important for understanding the characteristics of these complex networks.
[0004] For example, the Chinese invention patent with patent number CN116720975A discloses a local community discovery method and system based on structural similarity. The inventor considered applying local community discovery to room cleaning work, but found that the method of the above patent was too complicated and could not be well applied to room cleaning work. Therefore, the inventor proposed a room intelligent cleaning method and device based on local community discovery. Summary of the invention
[0005] Based on this, it is necessary to provide a room intelligent cleaning method and device based on local community discovery to address the problem that the existing local community discovery method is too complicated and cannot be well applied to room cleaning work.
[0006] The present invention is implemented by the following technical solutions:
[0007] In a first aspect, the present invention discloses an intelligent room cleaning method based on local community discovery, in which a target room is cleaned by a room cleaning device, wherein the room cleaning device is a movable cleaning robot with a camera and a mechanical arm.
[0008] The room intelligent cleaning method based on local community discovery includes the following steps:
[0009] Step 1: Use the camera to obtain the current image in the target room and compare it with the initial image in the target room to determine whether there are M objects G1 to G2 to be processed. M ; The initial image in the target room corresponds to the state where all items in the target room are cleaned up;
[0010] If there are G1~G M , then proceed to step 2; otherwise, end;
[0011] Step 2: G1~G M Convert to node representation to get M given nodes O1~O M , and O1~O M Classified into Y nodes to be stored Red1~Red Y 、N nodes to be reset Yellow1~Yellow N ; Y+N=M;
[0012] If N>0, proceed to step 3; otherwise, proceed to step 6;
[0013] Step 3: Based on Red1~Red Y Construct A candidate local communities CR1~CR A ;
[0014] Among them, step three includes:
[0015] S31, from Red1 to Red Y Find A seed nodes s1~s A ;
[0016] S32, the maximum operating distance of the robot arm Max As the threshold, take the a-th seed node S a Construct the ath candidate local community CR for the center a ; a∈[1,A];
[0017] S33, traverse s1~s A , we get CR1~CR A ;
[0018] Step 4: Calculate CR1~CR A Priority;
[0019] The candidate collection local community with the highest priority is used as the current collection local community, and the items to be processed in it are collected;
[0020] Step 5: After the current local community is stored, return to step 1;
[0021] Step 6: Yellow1~Yellow N Perform optimality calculation; take the node to be reset with the smallest optimality as the current reset node, and reset the corresponding item to be reset;
[0022] Step 7: After the current reset node is reset, return to step 1.
[0023] This intelligent room cleaning method based on local community discovery implements the method or process according to an embodiment of the present disclosure.
[0024] In a second aspect, the present invention discloses a room cleaning device based on local community discovery, which is the room cleaning device in the intelligent room cleaning method based on local community discovery disclosed in the first aspect.
[0025] In a third aspect, the present invention discloses a computer program product, including a computer program. When the computer program is executed by a processor, the steps of the room intelligent cleaning method based on local community discovery disclosed in the first aspect are implemented.
[0026] Compared with the prior art, the present invention has the following beneficial effects:
[0027] The present invention uses a mobile cleaning robot with a camera and a mechanical arm as a room cleaning device, first scans the target room to obtain the current image, then compares it with the initial image of the target room, and distinguishes a number of items to be processed; then converts the several items to be processed according to the required operation and classifies them into a number of nodes to be stored and nodes to be reset; then first uses local community discovery to process the nodes to be stored, constructs a local community, and stores the local community with the highest priority, and then restarts until all the items to be stored are stored; then calculates the optimal degree of the nodes to be reset, and resets the nodes to be reset with the smallest optimal degree, and then restarts until all the items to be reset are reset. The present invention realizes the good application of local community discovery in room cleaning work, and effectively improves the efficiency and intelligence level of room cleaning. BRIEF DESCRIPTION OF THE DRAWINGS
[0028] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.
[0029] Figure 1 A flow chart of a room intelligent cleaning method based on local community discovery provided in Example 1 of the present invention;
[0030] Figure 2 This is the flowchart of step three;
[0031] Figure 3 This is the flowchart for step six. DETAILED DESCRIPTION
[0032] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0033] It should be noted that when a component is referred to as being "mounted on" another component, it may be directly on the other component or there may be a central component. When a component is considered to be "set on" another component, it may be directly set on the other component or there may be a central component at the same time. When a component is considered to be "fixed to" another component, it may be directly fixed on the other component or there may be a central component at the same time.
[0034] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which the present invention belongs. The terms used herein in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "or / and" used herein includes any and all combinations of one or more of the related listed items.
[0035] Example 1
[0036] This embodiment 1 discloses a room cleaning device based on local community discovery, which is a movable cleaning robot with a camera and a mechanical arm.
[0037] The room cleaning device is equipped with controllable wheels at the bottom, such as Mecanum wheels, so that it can move freely in the target room. The camera is used to take pictures of the situation in the target room; the robotic arm has multiple degrees of freedom, which is used to perform corresponding operations on the items to be processed, such as moving or storing them.
[0038] This embodiment 1 simultaneously discloses a room intelligent cleaning method based on local community discovery, that is, cleaning the target room by using the above-mentioned room cleaning device.
[0039] See also Figure 1 ,The room intelligent cleaning method based on local community discovery includes the following steps:
[0040] Step 1: Use the camera to obtain the current image in the target room and compare it with the initial image in the target room to determine whether there are M objects G1 to G2 to be processed. M .
[0041] The initial image in the target room corresponds to a state where all items in the target room are cleaned up.
[0042] Specifically, step one includes:
[0043] S11, pre-scanning the cleaned target room to obtain an initial image of the target room;
[0044] S12, using a target detection algorithm (such as YoloV8 algorithm) to process the initial image to obtain all object information in the initial image, such as the type of object, the original position of the object, etc.;
[0045] S13, scanning the target room through a camera from the current position of the room cleaning device to obtain a current image of the target room;
[0046] S14, using a target detection algorithm (such as YoloV8 algorithm) to process the current image in the target room to obtain all object information in the current image, including the type of object, the current location of the object, etc.;
[0047] S15, compare the current image in the target room with all the object information in the initial image to determine whether G1~G M ;
[0048] Among them, if an object to be processed only needs to be moved and does not need to be stored, then the object to be processed is an object to be reset;
[0049] If an item to be processed needs to be moved and stored, then the item to be processed is an item to be stored.
[0050] It should be noted that if there are G1~G M , then proceed to step 2; otherwise end.
[0051] Step 2: G1~G M Convert to node representation to get M given nodes O1~O M , and O1~O M Classified into Y nodes to be stored Red1~Red Y 、N nodes to be reset Yellow1~Yellow N ;Y+N=M.
[0052] Specifically, step 2 includes:
[0053] S21, constructing a three-dimensional rectangular coordinate system using the current position of the room cleaning device as a reference origin;
[0054] S22, obtain the point cloud data of the current image and process it into a three-dimensional model to obtain G1~G MThe volume and distance from the room cleaning device are matched to the three-dimensional rectangular coordinate system to form O1~O M ;
[0055] S23, according to the type of the object to be processed, M Divided into Red1~Red Y 、Yellow1~Yellow N ;
[0056] Wherein, if a certain item to be processed is an item to be stored, then the given node corresponding to the item to be processed is a node to be stored;
[0057] If a to-be-processed item is an to-be-reset item, then the given node corresponding to the to-be-processed item is a to-be-reset node.
[0058] It should be noted that if N>0, proceed to step 3; otherwise, proceed to step 6.
[0059] That is to say, if there are items to be stored, the items to be stored are processed first; otherwise, the items to be restored are processed first.
[0060] Step 3: Based on Red1~Red Y Construct A candidate local communities CR1~CR A .
[0061] See also Figure 2 , step three includes:
[0062] S31, from Red1 to Red Y Find A seed nodes s1~s A ;
[0063] Specifically, S31 includes:
[0064] S311, calculate Red1~Red Y Bias Add_Bias1~Add_Bias Y ;
[0065] in,
[0066] Where Add_Bias y Red y Bias of; y∈[1,Y]; Red y Indicates the yth node to be stored; Volume_red y Red y The volume of the corresponding items to be processed;
[0067] S312, according to Add_Bias1~Add_BiasY Calculate Red1~Red Y ReLU activation function value ReLU1~ReLU Y ;
[0068] in,
[0069] In the formula, ReLU y Red y The ReLU activation function value of
[0070] S313, traverse Red1~Red Y , we get s1~s A , and form a seed node set S.
[0071] Among them, if the ReLU activation function value of a node to be accepted is non-zero, the node to be accepted is used as a seed node.
[0072] S32, the maximum operating distance of the robot arm Max As the threshold, take the a-th seed node S a Construct the ath candidate local community CR for the center a ; a∈[1,A].
[0073] Specifically, S32 includes:
[0074] S321, based on O1~O M Construct an undirected weighted graph;
[0075] Where Graph = {V, E}; V represents the node set in Graph, V = [O1,…,O M ]; E represents the edge set of Graph, which is initially empty;
[0076] S322, assign the given nodes other than S in V to the non-seed node set H; H = [h1,…,h C ]; A+C=M;
[0077] Calculate a to h c Distance Among them, h c represents the cth non-S given node in H; c∈[1,C];
[0078] S323, if Then in s a 、h c Create an edge between a,c ;
[0079] Traverse s1~sA 、h1~h C , get several edges and update them to E;
[0080] S324, keep S a unchanged, with S a Centered on all e a,c 、h c Form candidate local community CR a .
[0081] S33, traverse s1~s A , we get CR1~CR A .
[0082] Step 4: Calculate CR1~CR A Priority;
[0083] The candidate storage local community with the highest priority is used as the current storage local community, and the items to be processed in the local community are stored by the robot arm.
[0084] Among them, CR a The priority σ a for:
[0085]
[0086] In the formula, Num(CR a ∩S) represents the number of seed nodes in CRa; Represents the sum of the volumes of items to be processed corresponding to all nodes in CRa; Represents the sum of the volumes of items to be processed corresponding to all nodes in V; Distance_S represents the sum of the distances between the items to be processed and the room cleaning device corresponding to all nodes in V; a Indicates S a The corresponding distance between the items to be processed and the room cleaning device.
[0087] Step 5: After the current local community is stored, return to step 1;
[0088] It should be noted that, considering that there may be obstructed items to be processed in the previous scan, this step must return to step 1 and start again until all the items to be stored are stored.
[0089] Step 6: Yellow1~Yellow N Perform optimality calculation; take the node to be reset with the smallest optimality as the current reset node, and reset the corresponding object to be reset through the robotic arm.
[0090] See also Figure 3, step six includes:
[0091] S61, obtain Yellow1~Yellow N The corresponding N destination positions form R destination nodes Destination1 to Destination N ;
[0092] Among them, the nth node to be reset is Yellow n Corresponding to the nth destination node Destination n ; n∈[1,N];
[0093] S62, calculate Destination n To Yellow n Distance
[0094] like Then Destination n As the first destination; otherwise, Destination n As a secondary endpoint;
[0095] S63, traverse Destination1~Destination N 、Yellow1~Yellow N , get P first-level endpoints i1~i P , Q secondary endpoints k1~k Q ; P + Q = N;
[0096] If P≠0, then proceed to S64; if P=0, then proceed to S65;
[0097] S64, calculate i1~i P The optimal degrees of the corresponding P nodes to be reset, and the node to be reset with the smallest optimal degree is taken as the current reset node;
[0098] Among them, the pth first-level terminal i p The corresponding node to be reset is I p ;
[0099] I p The optimal degree ε p for:
[0100]
[0101] In the formula, Indicates room cleaning equipment to I p The distance of the corresponding items to be processed; Indicates I pThe corresponding items to be processed go to i p The distance of the corresponding end point position;
[0102] S65, calculate k1~k Q The optimal degrees of the corresponding Q nodes to be reset, and the node to be reset with the smallest optimal degree is taken as the current reset node;
[0103] Among them, the qth secondary endpoint k q The corresponding node to be reset is K q ;
[0104] K q The optimal degree λ p for:
[0105]
[0106] In the formula, Indicates room cleaning equipment to K q The distance of the corresponding items to be processed; K q The corresponding items to be processed are k q The distance to the corresponding end point.
[0107] Step 7: After the current reset node is reset, return to step 1.
[0108] Similar to step five, considering that there may be obstructed objects to be processed in the previous scan, this step needs to return to step one and start again until all the objects to be reset are reset.
[0109] Through the above operations, the room cleaning device can realize intelligent cleaning of the target room.
[0110] Example 2
[0111] This embodiment 2 discloses a computer device, including a memory and a processor. The memory stores a computer program. When the processor executes the computer program, the steps of the room intelligent cleaning method based on local community discovery disclosed in embodiment 1 are implemented.
[0112] This embodiment 2 also discloses a readable storage medium, in which computer program instructions are stored. When the computer program instructions are read and executed by a processor, the steps of the room intelligent cleaning method based on local community discovery disclosed in embodiment 1 are executed.
[0113] This embodiment 2 also discloses a computer program product, including a computer program. When the computer program is executed by a processor, the steps of the room intelligent cleaning method based on local community discovery disclosed in embodiment 1 are implemented.
[0114] The above-mentioned embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.
Claims
1. A room intelligent cleaning method based on local community discovery, which cleans the target room through a room cleaning device, characterized in that: The room cleaning device is a movable cleaning robot with a camera and a mechanical arm; The room intelligent cleaning method based on local community discovery includes the following steps: Step 1: Use the camera to obtain the current image in the target room and compare it with the initial image in the target room to determine whether there are M objects G1 to G2 to be processed. M ; The initial image in the target room corresponds to the state where all items in the target room are cleaned up; If there are G1~G M , then proceed to step 2; otherwise, end; Step 2: G1~G M Convert to node representation to get M given nodes O1~O M , and O1~O M Classified into Y nodes to be stored Red1~Red Y 、N nodes to be reset Yellow1~Yellow N ; Y+N=M; If N>0, proceed to step 3; otherwise, proceed to step 6; Step 3: Based on Red1~Red Y Construct A candidate local communities CR1~CR A ; Among them, step three includes: S31, from Red1 to Red Y Find A seed nodes s1~s A ; S32, the maximum operating distance of the robot arm Max As the threshold, take the a-th seed node S a Construct the ath candidate local community CR for the center a ; a∈[1,A]; S33, traverse s1~s A , we get CR1~CR A ; Step 4: Calculate CR1~CR A Priority; The candidate storage local community with the highest priority is used as the current storage local community, and the items to be processed in it are stored by the robotic arm; Step 5: After the current local community is stored, return to step 1; Step 6: Yellow1~Yellow N Perform optimality calculation; take the node to be reset with the smallest optimality as the current reset node, and reset the corresponding object to be reset by the robotic arm; Step 7: After the current reset node is reset, return to step 1.
2. The room intelligent cleaning method based on local community discovery according to claim 1 is characterized in that: Step one includes: S11, pre-scanning the cleaned target room to obtain an initial image of the target room; S12, processing the initial image using a target detection algorithm to obtain information about all objects in the initial image; S13, scanning the target room through a camera from the current position of the room cleaning device to obtain a current image of the target room; S14, using a target detection algorithm to process the current image in the target room to obtain information about all objects in the current image; S15, compare the current image in the target room with all the object information in the initial image to determine whether G1~G M ; Among them, if an object to be processed only needs to be moved and does not need to be stored, then the object to be processed is an object to be reset; If an item to be processed not only needs to be moved but also needs to be stored, then the item to be processed is an item to be stored.
3. The room intelligent cleaning method based on local community discovery according to claim 1 is characterized in that: Step 2 includes: S21, constructing a three-dimensional rectangular coordinate system using the current position of the room cleaning device as a reference origin; S22, obtain the point cloud data of the current image and process it into a three-dimensional model to obtain G1~G M The volume and distance from the room cleaning device are matched to the three-dimensional rectangular coordinate system to form O1~O M ; S23, according to the type of the object to be processed, M Divided into Red1~Red Y 、Yellow1~Yellow N ; Wherein, if a certain item to be processed is an item to be stored, then the given node corresponding to the item to be processed is a node to be stored; If a to-be-processed item is an to-be-reset item, then the given node corresponding to the to-be-processed item is a to-be-reset node.
4. The room intelligent cleaning method based on local community discovery according to claim 3 is characterized in that: S31 includes: S311, calculate Red1~Red Y Bias Add_Bias1~Add_Bias Y ; in, Where Add_Bias y Indicates Red y Bias of; y∈[1,Y]; Red y Indicates the yth node to be stored; Volume_red y Indicates Red y The volume of the corresponding items to be processed; S312, according to Add_Bias1~Add_Bias Y Calculate Red1~Red Y ReLU activation function value ReLU1~ReLU Y ; in, In the formula, ReLU y Indicates Red y The ReLU activation function value of S313, traverse Red1~Red Y , we get s1~s A , and form a seed node set S; Among them, if the ReLU activation function value of a node to be accepted is non-zero, the node to be accepted is used as a seed node.
5. The intelligent room cleaning method based on local community discovery according to claim 4 is characterized in that: S32 includes: S321, based on O1~O M Construct an undirected weighted graph; Where Graph = {V, E}; V represents the node set in Graph, V = [O1,…,O M ]; E represents the edge set of Graph, which is initially empty; S322, assign the given nodes other than S in V to the non-seed node set H; H = [h1,…,h C ]; A+C=M; Calculate a to h c Distance Among them, h c represents the cth non-S given node in H; c∈[1,C]; S323, if Then in s a 、h c Create an edge between a,c ; Traverse s1~s A 、h1~h C , get several edges and update them to E; S324, keep S a unchanged, with S a Centered on all e a,c 、h c Form candidate local community CR a .
6. The room intelligent cleaning method based on local community discovery according to claim 1 is characterized in that: In step 4, CR a The priority σ a for: In the formula, Num(CR a ∩S) represents the number of seed nodes in CRa; Represents the sum of the volumes of items to be processed corresponding to all nodes in CRa; Represents the sum of the volumes of items to be processed corresponding to all nodes in V; Distance_S represents the sum of the distances between the items to be processed and the room cleaning device corresponding to all nodes in V; a Indicates S a The corresponding distance between the items to be processed and the room cleaning device.
7. The room intelligent cleaning method based on local community discovery according to claim 1 is characterized in that: Step six includes: S61, obtain Yellow1~Yellow N The corresponding N destination positions form R destination nodes Destination1 to Destination N ; Among them, the nth node to be reset is Yellow n Corresponding to the nth destination node Destination n ; n∈[1,N]; S62, calculate Destination n To Yellow n Distance like Then Destination n As the first destination; otherwise, Destination n As a secondary endpoint; S63, traverse Destination1~Destination N 、Yellow1~Yellow N , get P first-level endpoints i1~i P , Q secondary endpoints k1~k Q ; P + Q = N; If P≠0, then proceed to S64; if P=0, then proceed to S65; S64, calculate i1~i P The optimal degrees of the corresponding P nodes to be reset, and the node to be reset with the smallest optimal degree is taken as the current reset node; S65, calculate k1~k Q The optimal degrees of the corresponding Q nodes to be reset are calculated, and the node to be reset with the smallest optimal degree is taken as the current reset node.
8. The intelligent room cleaning method based on local community discovery according to claim 7 is characterized in that: In S64, the pth first-level terminal i p The corresponding node to be reset is I p ; I p The optimal degree ε p for: In the formula, Indicates room cleaning equipment to I p The distance of the corresponding items to be processed; Indicates I p The corresponding items to be processed go to i p The distance of the corresponding end point position; In S65, the qth secondary endpoint k q The corresponding node to be reset is K q ; K q The optimal degree λ p for: In the formula, Indicates room cleaning equipment to K q The distance of the corresponding items to be processed; K q The corresponding items to be processed are k q The distance to the corresponding end point.
9. A room cleaning device based on local community discovery, characterized in that: It is a room cleaning device in the intelligent room cleaning method based on local community discovery as described in any one of claims 1-8.
10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the intelligent room cleaning method based on local community discovery are implemented as described in any one of claims 1 to 8.
Citation Information
Patent Citations
Local community discovery method and system based on structural similarity
CN116720975A
Intelligent cleaning method of robot, device thereof and robot
CN109664309A
Dynamic article searching method and device in home environment
CN110110245A
Garbage classification and pickup robot based on deep learning
CN112149573A
Method for realizing real-time monitoring of state of sky-near-air ground vehicle network node
CN112333049A
Cited By
Robot control method, cleaning robot, cleaning system, and storage medium
CN122807956A
