Control method of cleaning equipment and cleaning equipment

Abstract

The embodiment of the invention provides a control method of cleaning equipment and the cleaning equipment, and relates to the technical field of intelligent control. The method comprises the steps that accessory bin state identity information of an accessory bin is collected, and the accessory bin is used for storing cleaning accessories for assisting cleaning equipment in cleaning; detecting whether the accessory bin is at a preset position based on the accessory bin state identity information, and obtaining an in-place detection result; and generating a cleaning strategy based on the in-place detection result, and executing a cleaning task according to the cleaning strategy. According to the method, high-reliability accessory bin in-place detection can be achieved by collecting and using the state identity information of the accessory bin, then the cleaning strategy beneficial to assisting the cleaning equipment in cleaning can be generated according to the in-place detection result, and the cleanliness and the cleaning efficiency can be improved when the cleaning equipment executes the cleaning task.

Description

Technical Field

[0001] This application relates to the field of intelligent control technology, and in particular to a control method for cleaning equipment and a cleaning equipment. Background Technology

[0002] With the rapid development of smart home technology, cleaning equipment such as robotic vacuum cleaners and smart vacuum cleaners have been widely used in homes, offices and industrial settings, becoming important tools for improving quality of life and work efficiency.

[0003] In complex and ever-changing cleaning environments, users have placed higher demands on the intelligence level of cleaning equipment. They not only require cleaning equipment to have basic autonomous navigation, obstacle avoidance, and cleaning functions, but also expect it to be able to dynamically adapt to the needs of different cleaning tasks. For example, deep cleaning of special areas such as floor crevices, high surfaces (such as desktops and cabinet doors), or narrow corners (such as wall corners and furniture gaps).

[0004] Taking robotic vacuum cleaners as an example, existing robotic vacuum cleaners do not have dedicated cleaning accessories and can only rely on the vacuum cleaner at the bottom of the robot to clean the floor. These cleaning accessories include, for example, brushes, portable vacuum cleaners, and suction hoses.

[0005] Since existing technologies do not equip cleaning equipment with cleaning accessories, they also lack the control logic and algorithms for detecting, identifying, grasping, and using these accessories. Therefore, existing cleaning equipment struggles to accurately detect whether the accessory storage compartment is in the preset location, and it is also difficult to control the cleaning equipment to perform cleaning tasks using the accessories. Summary of the Invention

[0006] This application provides a control method and a cleaning device for a cleaning equipment, which is used to control the cleaning equipment to perform more accurate on-site detection of the accessory compartment storing cleaning accessories, and generate and execute a cleaning strategy based on the on-site detection results to improve cleanliness and cleaning efficiency.

[0007] In a first aspect, embodiments of this application provide a control method for a cleaning device. The method includes: collecting the status and identity information of a parts compartment, the parts compartment being used to store cleaning parts that assist the cleaning device in cleaning; detecting whether the parts compartment is in a preset position based on the status and identity information of the parts compartment, and obtaining an on-site detection result; generating a cleaning strategy based on the on-site detection result, and executing a cleaning task according to the cleaning strategy.

[0008] In one possible implementation, detecting whether a parts warehouse is in a preset location based on its status and identity information, and obtaining a presence detection result, includes: matching the parts warehouse status and identity information with a preset parts warehouse feature model, wherein the parts warehouse feature model is a model constructed based on features extracted when the parts warehouse is in the preset location; if the match is successful, obtaining a presence detection result for the parts warehouse being in the preset location; if the match fails, obtaining a presence detection result for the parts warehouse not being in the preset location.

[0009] In one possible implementation, the parts warehouse status identity information includes a unique identifier for the parts warehouse, and the parts warehouse feature model includes identifier features extracted from the unique identifier for the parts warehouse; matching the parts warehouse status identity information with the preset parts warehouse feature model includes: matching the unique identifier for the parts warehouse with the identifier features of the parts warehouse feature model.

[0010] In one possible implementation, the parts compartment status identity information includes the spatial pose of the parts compartment, and the parts compartment feature model includes the in-situ spatial pose range of the parts compartment; matching the parts compartment status identity information with the preset parts compartment feature model includes: matching the spatial pose of the parts compartment with the in-situ spatial pose range of the parts compartment feature model.

[0011] In one possible implementation, generating a cleaning strategy based on in-situ detection results includes: generating a cleaning strategy based on the cleaning accessories in the accessories compartment when the in-situ detection results indicate that the accessories compartment is in a preset position.

[0012] In one possible implementation, generating a cleaning strategy based on cleaning accessories in the accessory compartment includes: identifying at least one cleaning accessory in the accessory compartment, determining the type of the at least one cleaning accessory, and generating a cleaning strategy based on cleaning task requirements and the type of the at least one cleaning accessory.

[0013] In one possible implementation, identifying at least one cleaning accessory in the accessory compartment and determining the type of the at least one cleaning accessory includes: identifying a unique accessory identifier for at least one cleaning accessory in the accessory compartment and determining the type of the at least one cleaning accessory based on the unique accessory identifier.

[0014] In one possible implementation, identifying at least one cleaning accessory in the accessory compartment and determining the type of at least one cleaning accessory includes: scanning a near-field communication tag affixed to at least one cleaning accessory in the accessory compartment using a near-field communication reader mounted on the cleaning device; identifying at least one cleaning accessory in the accessory compartment and determining the type of at least one cleaning accessory based on the data interaction identification result between the near-field communication reader and the near-field communication tag.

[0015] In one possible implementation, the cleaning strategy includes an accessory grasping parameter set, which includes at least one grasping action parameter for grasping a cleaning accessory; generating the cleaning strategy based on the cleaning task requirements and the type of at least one cleaning accessory includes: determining a target cleaning accessory that meets the cleaning task requirements based on the cleaning task requirements and the type of at least one cleaning accessory; calculating the grasping action parameters for grasping the target cleaning accessory based on the relative pose between the cleaning device and the target cleaning accessory, thereby generating the accessory grasping parameter set of the cleaning strategy.

[0016] In one possible implementation, the method further includes: grabbing a target cleaning accessory using a gripping device of a cleaning device according to an accessory gripping parameter set; detecting the gripping opening degree of the gripping device using a position encoder of the gripping device during the gripping process; and determining whether the gripping device has grabbed the target cleaning accessory based on the gripping opening degree.

[0017] In one possible implementation, generating a cleaning strategy based on the in-situ detection results includes: when the in-situ detection results indicate that the accessory compartment is not in a preset position, collecting environmental perception data to allow the cleaning equipment to perceive its surroundings; based on the environmental perception data, identifying whether cleaning tools exist in the environment through a visual model, wherein the cleaning tools are tools used to assist in cleaning; and generating a cleaning strategy based on the cleaning tools identified in the environment.

[0018] In one possible implementation, generating a cleaning strategy based on cleaning tools identified in the environment includes: determining the tool type of the cleaning tools identified in the environment using a visual model; and generating a cleaning strategy based on cleaning task requirements and tool type.

[0019] In one possible implementation, a cleaning strategy is generated based on the cleaning task requirements and the tool type, including: if the cleaning task requirements include cleaning blind spots, determining whether the tool type meets the cleaning task requirements; if the tool type meets the cleaning task requirements, generating a cleaning strategy that includes cleaning blind spots based on the tool type.

[0020] In one possible implementation, the cleaning strategy includes a tool grasping parameter set, which includes at least one grasping action parameter for grasping a cleaning tool; generating the cleaning strategy based on cleaning task requirements and tool types includes: when multiple cleaning tools of various tool types are identified, determining a target cleaning tool that meets the cleaning task requirements based on the cleaning task requirements and the tool types of the multiple cleaning tools; calculating the grasping action parameters for grasping the target cleaning tool based on the relative pose between the cleaning device and the target cleaning tool, and generating the tool grasping parameter set of the cleaning strategy.

[0021] In one possible implementation, the method further includes: grasping a target cleaning tool using a grasping device of a cleaning device according to a tool grasping parameter set; detecting the grasping opening degree of the grasping device using a position encoder of the grasping device during the grasping process; and determining whether the grasping device has grasped the target cleaning tool based on the grasping opening degree.

[0022] In one possible implementation, generating a cleaning strategy based on the in-situ detection results includes: when the in-situ detection results indicate that the accessory compartment is not in a preset location, calling the environmental map of the cleaning equipment, the environmental map being obtained by the cleaning equipment performing a spatial scan and map construction of its environment, the environmental map including the annotation information of obstacles; determining at least one cleaning tool among at least one obstacle based on the annotation information of at least one obstacle in the environmental map, the cleaning tool being a tool used to assist in cleaning; and generating a cleaning strategy based on at least one cleaning tool.

[0023] In one possible implementation, the method further includes: collecting environmental perception data of the environment when the cleaning equipment catches at least one cleaning tool; updating the environmental map based on the environmental perception data to obtain an updated environmental map.

[0024] Secondly, embodiments of this application provide a control device for a cleaning equipment. The control device includes: a data acquisition module for acquiring the status and identity information of a parts compartment, the parts compartment being used to store cleaning accessories that assist the cleaning equipment in cleaning; a detection module for detecting whether the parts compartment is in a preset position based on the status and identity information of the parts compartment, and obtaining an on-site detection result; and a generation module for generating a cleaning strategy based on the on-site detection result, and executing a cleaning task according to the cleaning strategy.

[0025] In one possible implementation, the detection module is specifically used to: match the parts warehouse status identity information with a preset parts warehouse feature model, which is a model constructed based on the features extracted when the parts warehouse is in a preset location; if the match is successful, obtain the in-situ detection result of the parts warehouse in the preset location; if the match fails, obtain the in-situ detection result of the parts warehouse not being in the preset location.

[0026] In one possible implementation, the parts warehouse status identity information includes a unique identifier for the parts warehouse, and the parts warehouse feature model includes identifier features extracted from the unique identifier of the parts warehouse; the detection module is specifically used to match the unique identifier of the parts warehouse with the identifier features of the parts warehouse feature model.

[0027] In one possible implementation, the parts storage status identity information includes the spatial pose of the parts storage, and the parts storage feature model includes the in-situ spatial pose range of the parts storage; the detection module is specifically used to match the spatial pose of the parts storage with the in-situ spatial pose range of the parts storage feature model.

[0028] In one possible implementation, the generation module is specifically used to: generate a cleaning strategy based on the cleaning parts in the parts compartment, given that the in-situ detection results indicate that the parts compartment is in a preset position.

[0029] In one possible implementation, the generation module is specifically used to: identify at least one cleaning accessory in the accessory compartment, determine the type of the at least one cleaning accessory, and generate a cleaning strategy based on the cleaning task requirements and the type of the at least one cleaning accessory.

[0030] In one possible implementation, the generation module is specifically used to: identify the accessory unique identifier of at least one cleaning accessory in the accessory compartment, and determine the type of at least one cleaning accessory based on the accessory unique identifier.

[0031] In one possible implementation, the generation module is specifically used to: scan a near-field communication tag affixed to at least one cleaning accessory in the accessory compartment using a near-field communication reader installed on the cleaning device; identify at least one cleaning accessory in the accessory compartment based on the data interaction identification result between the near-field communication reader and the near-field communication tag, and determine the type of at least one cleaning accessory.

[0032] In one possible implementation, the cleaning strategy includes an accessory grasping parameter set, which includes at least one grasping action parameter for grasping a cleaning accessory; the generation module is specifically used to: determine a target cleaning accessory that meets the cleaning task requirements based on the cleaning task requirements and the type of at least one cleaning accessory; calculate the grasping action parameters for grasping the target cleaning accessory based on the relative pose between the cleaning equipment and the target cleaning accessory, and generate the accessory grasping parameter set of the cleaning strategy.

[0033] In one possible implementation, the control device further includes a gripping module, which is used to: grip a target cleaning accessory using a gripping device of the cleaning equipment according to an accessory gripping parameter set; detect the gripping opening degree of the gripping device using a position encoder during the gripping process; and determine whether the gripping device has gripped the target cleaning accessory based on the gripping opening degree.

[0034] In one possible implementation, the generation module is specifically used to: collect environmental perception data of the environment in which the cleaning equipment perceives the environment when the in-situ detection result indicates that the accessory compartment is not in the preset position; based on the environmental perception data, identify whether there are cleaning tools in the environment through a visual model, wherein the cleaning tools are tools used to assist cleaning; and generate a cleaning strategy based on the cleaning tools identified in the environment.

[0035] In one possible implementation, the generation module is specifically used to: determine the tool type of the cleaning tools identified in the environment through a visual model; and generate a cleaning strategy based on the cleaning task requirements and the tool type.

[0036] In one possible implementation, the generation module is specifically used to: determine whether the tool type meets the cleaning task requirements when the cleaning task requirements include cleaning blind spots; if the tool type meets the cleaning task requirements, generate a cleaning strategy that includes cleaning blind spots based on the tool type.

[0037] In one possible implementation, the cleaning strategy includes a tool grasping parameter set, which includes at least one grasping action parameter for grasping a cleaning tool; the generation module is specifically used to: when multiple cleaning tools of various tool types are identified, determine a target cleaning tool that meets the cleaning task requirements based on the cleaning task requirements and the tool types of the multiple cleaning tools; calculate the grasping action parameters for grasping the target cleaning tool based on the relative pose between the cleaning equipment and the target cleaning tool, and generate the tool grasping parameter set of the cleaning strategy.

[0038] In one possible implementation, the control device further includes a gripping module, which is configured to: grip a target cleaning tool using a gripping device of the cleaning equipment according to a tool gripping parameter set; detect the gripping opening degree of the gripping device using a position encoder during the gripping process; and determine whether the gripping device has gripped the target cleaning tool based on the gripping opening degree.

[0039] In one possible implementation, the generation module is specifically used to: when the in-situ detection result indicates that the accessory compartment is not in a preset position, call the environmental map of the cleaning equipment, the environmental map is obtained by the cleaning equipment after spatial scanning and map construction of its environment, and the environmental map includes the annotation information of obstacles; based on the annotation information of at least one obstacle in the environmental map, determine at least one cleaning tool among at least one obstacle, the cleaning tool is a tool used to assist cleaning; and generate a cleaning strategy based on at least one cleaning tool.

[0040] In one possible implementation, the control device further includes an update module, which is used to: collect environmental perception data of the environment when the cleaning equipment picks up at least one cleaning tool; and update the environmental map based on the environmental perception data to obtain an updated environmental map.

[0041] Thirdly, embodiments of this application provide a cleaning device, which includes a gripping device for performing gripping actions, and the cleaning device is used to implement the first aspect and / or various possible implementations of the first aspect.

[0042] Fourthly, embodiments of this application provide an electronic device, including: a memory and a processor; the memory stores computer-executable instructions; the processor executes the computer-executable instructions stored in the memory, causing the processor to perform the first aspect and / or various possible implementations of the first aspect as described above.

[0043] Fifthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the first aspect and / or various possible implementations of the first aspect.

[0044] Sixthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the first aspect and / or various possible implementations of the first aspect.

[0045] The control method and cleaning equipment provided in this application embodiment collect the status and identity information of the parts compartment and detect whether the parts compartment is in a preset position based on the status and identity information to obtain an on-site detection result. Furthermore, a cleaning strategy can be generated based on the on-site detection result, and a cleaning task can be executed according to the cleaning strategy. In this way, by collecting and using the status and identity information of the parts compartment, a highly reliable on-site detection of the parts compartment can be achieved. Furthermore, a cleaning strategy that assists the cleaning equipment in cleaning can be generated based on the on-site detection result, thereby improving the cleanliness and efficiency of the cleaning equipment when performing cleaning tasks. Attached Figure Description

[0046] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0047] Figure 1 A flowchart illustrating the control method for the cleaning equipment provided in this application embodiment;

[0048] Figure 2 This is a schematic diagram of the control system of the cleaning equipment provided in the embodiments of this application;

[0049] Figure 3 This is a schematic diagram of the control logic of the cleaning equipment provided in the embodiments of this application;

[0050] Figure 4 This is a schematic diagram of the structure of the control device for the cleaning equipment provided in the embodiments of this application;

[0051] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0052] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0053] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0054] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of the relevant data all comply with the relevant laws, regulations, and standards of the relevant countries and regions, have taken necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation access points for users to choose to authorize or refuse.

[0055] Furthermore, the technical solution involved in this application, which involves big data analysis of user information (including but not limited to personal biometrics, identity data, consumption data, asset data, electronic terminal operation data, etc.) and the use of artificial intelligence technology for automated decision-making, and makes decisions that have a significant impact on personal rights based on the results of automated decision-making, provides users with corresponding operation entry points for users to choose to agree to or reject the results of automated decision-making; if the user chooses to reject, the process will proceed to the expert decision-making process.

[0056] In this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" or "for example" in this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0057] In the embodiments of this application, the use of terms such as "first" and "second" is to distinguish between identical or similar items that have essentially the same function and effect. For example, "first electronic device" and "second electronic device" are merely used to distinguish different electronic devices and do not limit their order of execution. Those skilled in the art will understand that the terms "first" and "second" do not limit the quantity or execution order, and that "first" and "second" do not necessarily imply that they are different.

[0058] In this application embodiment, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the relationship between associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following associated objects have an "or" relationship.

[0059] For example, existing intelligent cleaning devices such as robotic vacuum cleaners have many technical pain points. For instance, robotic vacuum cleaners cannot enter narrow spaces (such as the narrow gap between a sofa and a wall), making it difficult to effectively clean them using the vacuum cleaner at the bottom of the robot. Also, it is difficult to effectively clean non-floor areas (such as tabletops and cabinet doors) using the vacuum cleaner at the bottom.

[0060] For example, while some existing robotic vacuum cleaners are equipped with robotic arms, they don't fully utilize the advantages of these arms. For instance, they can only use the robotic arm to pick up small pieces of trash like crumpled paper and paper cups and dispose of them in the trash can, without considering using the robotic arm to grab cleaning accessories or tools for deep cleaning in narrow spaces, off-floor areas, corners, and other special areas. Therefore, existing cleaning equipment cannot perform comprehensive cleaning when controlled, resulting in lower cleaning effectiveness and efficiency.

[0061] The inventor conceived this idea while researching pain points in existing scenarios. If a robotic vacuum cleaner is delivered to the user equipped with specialized cleaning accessories, it can effectively clean hard-to-reach areas such as floor crevices, raised surfaces, or narrow corners, increasing cleaning coverage and capability, thereby improving cleanliness and efficiency. Furthermore, even without optional cleaning accessories, if the cleaning device can autonomously locate cleaning tools to clean specific areas during operation, it can also improve cleanliness and efficiency.

[0062] For example, when a robot vacuum cleaner recognizes a brush that the user has authorized to use in the environment, it can grab the brush with its robotic arm and clean the gaps between furniture or cabinet doors, increasing the cleaned area and improving cleanliness and efficiency.

[0063] Taking robotic vacuum cleaners as an example, since existing robotic vacuum cleaners are not equipped with cleaning accessories specifically for assisting cleaning, they do not have preset algorithmic logic for finding and using cleaning accessories. Therefore, there is no need or method for detecting whether the accessory compartment storing cleaning accessories is in a preset location, and consequently, there is no specific implementation for generating and executing cleaning strategies based on the on-site detection results of whether the accessory compartment is in a preset location.

[0064] In view of this, embodiments of this application provide a control method for cleaning equipment. This method collects the status and identity information of a parts compartment and detects whether the parts compartment is in a preset position based on this information to obtain an on-site detection result. Furthermore, a cleaning strategy can be generated based on the on-site detection result, and a cleaning task can be executed according to the strategy. Thus, by collecting and using the parts compartment status and identity information, highly reliable on-site detection of the parts compartment can be achieved. Furthermore, a cleaning strategy that assists the cleaning equipment in cleaning can be generated based on the on-site detection result, thereby improving the cleanliness and efficiency of the cleaning equipment when performing cleaning tasks.

[0065] The technical solutions of this application will be described in detail below with reference to specific embodiments. The specific embodiments described below can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of this application will be described below with reference to the accompanying drawings.

[0066] Figure 1 This is a flowchart illustrating the control method for cleaning equipment provided in an embodiment of this application. The executing entity of this method can be an electronic device with corresponding data storage and computing capabilities. This electronic device can be understood as a cleaning device, such as a robotic vacuum cleaner; it can also be understood as a processor, processing unit, or control chip within the cleaning device. Figure 1 As shown, the method includes:

[0067] S101, collect the status and identity information of the parts warehouse. The parts warehouse is used to store cleaning parts that assist in cleaning equipment cleaning.

[0068] For example, cleaning equipment can be electronic devices used for cleaning and hygiene, such as robotic vacuum cleaners and robot vacuums. Cleaning accessories can be understood as devices that are used in conjunction with cleaning equipment to assist the cleaning equipment in cleaning and hygiene. For example, they can be special cleaning equipment designed and configured by the provider of the cleaning equipment, such as brushes, portable vacuum cleaners, vacuum hoses, and vacuum heads.

[0069] The accessory compartment can be understood as a storage device used to store cleaning accessories. For example, the accessory compartment is equipped with multiple slots, and each slot can be used to lock and fix a cleaning accessory for storage.

[0070] The accessory compartment status and identity information can be understood as multi-dimensional information used to characterize the physical state and identity of the accessory compartment, including but not limited to the accessory compartment's unique identifier and spatial pose. The unique identifier of the accessory compartment can be, for example, a QR code, serial number, predefined code, barcode, icon, and / or Near Field Communication (NFC) tag. A unique identifier can be understood as an identifier that is globally unique within the cleaning equipment. The spatial pose of the accessory compartment includes, for example, its spatial position coordinates and attitude parameters. These attitude parameters are, for example, rotation angles within a spatial coordinate system (X-axis, Y-axis, Z-axis), such as roll angle, yaw angle, and pitch angle.

[0071] When collecting the status and identification information of the parts compartment, the cleaning equipment can scan the QR code, serial number, custom code, and / or icon of the parts compartment using its vision device; it can also collect the spatial pose of the parts compartment within the cleaning equipment's coordinate system using its vision device or positioning device. The vision device includes, for example, a camera, a camera, and / or a scanner. The positioning device can include, for example, a positioning radar, a laser rangefinder, and / or a depth camera.

[0072] For example, an inertial measurement unit (IMU) sensor can also be installed in the parts compartment to determine the attitude parameters of the parts compartment itself. The cleaning equipment can read the attitude parameters of the parts compartment through data communication to obtain the status and identity information of the parts compartment.

[0073] S102, based on the status and identity information of the parts warehouse, detect whether the parts warehouse is in the preset location and obtain the on-site detection result.

[0074] For example, a preset location can be understood as a standard installation location for the accessory compartment, which can be determined based on the device's mechanical structure or a physical space range defined by the user's configuration. For instance, the reserved area around the charging compartment of a robotic vacuum cleaner can be pre-defined using the robotic vacuum cleaner's map.

[0075] The presence detection result can be understood as a logical judgment result obtained after verifying or matching the status and identity information of the parts compartment. It can be used to indicate whether the parts compartment is in a preset location. For example, if the cleaning equipment matches the unique identifier of the parts compartment within the preset range, it means that the match is successful, and the presence detection result indicating that the parts compartment is in the preset location can be obtained, i.e., "in place"; if the cleaning equipment cannot match the unique identifier of the parts compartment within the preset range, it means that the match fails, and the presence detection result indicating that the parts compartment is not in the preset location can be obtained, i.e., "out of place" or "not in place".

[0076] S103, generate a cleaning strategy based on the in-situ detection results, and execute the cleaning task according to the cleaning strategy.

[0077] For example, after obtaining the presence detection result, the cleaning equipment knows whether the accessory compartment is in place or not. If it is in place, the cleaning equipment can perform a comprehensive analysis based on the available cleaning accessories in the accessory compartment and the items that need to be cleaned, generate a cleaning strategy, and execute the cleaning task according to the cleaning strategy.

[0078] The cleaning strategy can be understood as a set of control instructions for the cleaning equipment generated based on in-situ detection results, used to guide the cleaning equipment to perform cleaning tasks. The cleaning strategy may include, for example, cleaning paths, suction power, and cleaning head rotation speed.

[0079] For example, when the accessory compartment is detected to be in place, a cleaning path planning instruction to call up the brush accessory can be generated; when the accessory compartment is detected to be out of place, a cleaning strategy to use the cleaning device itself for cleaning can be generated. Alternatively, when the accessory compartment is detected to be out of place, a visual model can be used to identify whether there are usable cleaning tools in the environment, and if cleaning tools are present, a cleaning path planning instruction based on the cleaning tools can be generated. Here, cleaning tools can be understood as tools other than cleaning accessories that can be used to assist in cleaning.

[0080] For example, to achieve the above steps S101 to S103, firstly, the status and identity information of the parts warehouse is collected. This status and identity information of the parts warehouse can be obtained through multimodal sensors (such as vision devices and near-field communication card readers), including unique identifiers and spatial poses.

[0081] Secondly, by comparing the collected status and identity information of the parts compartment with preset detection standards for whether the parts compartment is in a preset location, a presence detection result (either in-situ or out-of-situ) can be obtained. For example, if the NFC tag of the parts compartment is read by the NFC reader in the cleaning equipment within the preset location, a presence detection result (in-situ) can be obtained; if the NFC tag of the parts compartment cannot be read by the NFC reader in the cleaning equipment within the preset location, an out-of-situ presence detection result (out-of-situ) can be obtained. As another example, by identifying the parts compartment and locating its spatial pose using a vision device and a positioning device, a presence detection result (in-situ) can be obtained when the spatial pose of the parts compartment matches a preset spatial pose; an out-of-situ presence detection result (out-of-situ) can be obtained when the spatial pose of the parts compartment does not match the preset spatial pose.

[0082] Finally, a cleaning strategy can be generated based on the on-site detection results. For example, if the equipment is in place, the cleaning accessory information in the accessory compartment is used to generate a cleaning path and cleaning parameters (such as travel speed, operating frequency, power, etc.); if the equipment is not in place, environmental sensing is triggered to identify whether there are available cleaning tools in the environment. The entire process described above can process and analyze sensor data through a preset algorithm to achieve real-time detection of the accessory compartment status and real-time generation of cleaning strategies, thus realizing closed-loop control of the cleaning equipment.

[0083] For example, most existing cleaning equipment does not have an accessory compartment, let alone a requirement to detect whether the accessory compartment is in place. However, the cleaning accessories in the accessory compartment are an important aid to improving the cleaning capabilities of the cleaning equipment. Therefore, the method of this application embodiment, based on detecting whether the accessory compartment is in a preset position, can provide a basis for expanding the cleaning capabilities of the cleaning equipment and can generate cleaning strategies that are more conducive to improving cleanliness and cleaning efficiency.

[0084] The control method for cleaning equipment provided in this application collects the status and identity information of the parts compartment and detects whether the parts compartment is in a preset position based on the status and identity information to obtain an on-site detection result. Furthermore, a cleaning strategy can be generated based on the on-site detection result, and a cleaning task can be executed according to the cleaning strategy. In this way, by collecting and using the status and identity information of the parts compartment, highly reliable on-site detection of the parts compartment can be achieved. Furthermore, a cleaning strategy that assists the cleaning equipment in cleaning can be generated based on the on-site detection result, thereby improving the cleanliness and efficiency of the cleaning equipment when performing cleaning tasks.

[0085] It should be noted that in some application scenarios, before the cleaning equipment performs in-situ testing of the parts compartment, the space environment may have already been cleaned globally or partially, or it may not have been cleaned globally or partially.

[0086] For example, when a robot vacuum starts a cleaning task, it first performs a general cleaning of the room. During cleaning, it marks some areas that cannot be thoroughly cleaned, such as furniture crevices, tabletops, or corners of interior walls. In this case, the robot vacuum can return to a preset position to detect the presence of the accessory compartment. If the accessory compartment is detected, it can generate an appropriate cleaning strategy based on the cleaning accessories in the compartment and the characteristics of the remaining area (e.g., narrow crevices are long and narrow, requiring a long-bristled brush) to thoroughly clean the remaining area.

[0087] In other application scenarios, the space environment has not been cleaned globally or locally before the cleaning equipment performs in-situ testing of the parts compartment.

[0088] For example, when a robot vacuum starts a cleaning task, it first checks if the accessory compartment is in place. If it does, it can select the last cleaning accessory used during the last cleaning to generate a cleaning strategy; or it can select the cleaning accessory with the highest historical usage frequency. Then, the robot vacuum begins cleaning the space, either globally or locally. Because the robot vacuum already has a cleaning accessory with it when it starts cleaning, when it encounters hard-to-clean areas that can be thoroughly cleaned with that accessory, such as crevices, tabletops, or corners, it can clean them directly without returning to the preset position for on-site detection and regenerating a cleaning strategy. This improves cleaning efficiency.

[0089] Of course, the description of using one cleaning accessory at a time in the various embodiments of this application is only for illustrative purposes and understanding. If the cleaning device has the ability to use multiple cleaning accessories simultaneously, a cleaning strategy can also be generated based on multiple cleaning accessories, and multiple cleaning accessories can be used simultaneously to perform cleaning tasks. This application does not limit this. For example, a robotic vacuum cleaner has multiple robotic arms that can simultaneously grasp multiple cleaning accessories. A cleaning strategy can be generated based on multiple cleaning accessories, and cleaning tasks can be performed based on these multiple cleaning accessories, which can greatly improve cleaning efficiency.

[0090] In one possible implementation, detecting whether a parts warehouse is in a preset location based on its status and identity information, and obtaining a presence detection result, includes: matching the parts warehouse status and identity information with a preset parts warehouse feature model, wherein the parts warehouse feature model is a model constructed based on features extracted when the parts warehouse is in the preset location; if the match is successful, obtaining a presence detection result for the parts warehouse being in the preset location; if the match fails, obtaining a presence detection result for the parts warehouse not being in the preset location.

[0091] For example, the preset accessory compartment feature model can be understood as a model constructed based on the features extracted from the accessory compartment in a preset location. For example, it includes a unique identifier feature of the preset location and / or a model with a spatial pose range, which can be used to match the accessory compartment's status and identity information. For example, the unique identifier feature of the preset location is the hash value of the QR code, and the spatial pose range is the spatial position coordinates (x±5cm, y±5cm, z±2cm) and the rotation angle (yaw angle ±10°).

[0092] Matching can be understood as comparing the status and identity information of the parts warehouse with a pre-defined parts warehouse feature model using rule-based matching or machine learning algorithms. For example, hash comparison can be used to verify unique identifiers, and thresholds can be used to determine whether the deviation of spatial pose is within an acceptable range. Matching can be combined with location information. For instance, matching within a pre-defined range of a pre-defined location using the parts warehouse feature model can improve the accuracy of in-situ detection.

[0093] For example, the collected status and identity information of the parts warehouse can be verified and matched against a preset feature model to determine whether the parts warehouse is in place. The verification and matching process can be divided into two branches: if the unique identifier (such as a QR code) matches the identifier features in the preset parts warehouse feature model, and the spatial pose (such as IMU data) is within a preset spatial range, then it is determined to be in place, and an in-place detection result is obtained; otherwise, it is determined to be out of place, and an out-of-place in-place detection result is obtained. The verification and matching logic can be implemented through preset rules (such as hash comparison, threshold judgment) or machine learning models (such as classifiers) to ensure the reliability of the in-place detection results.

[0094] In this embodiment, since the parts storage feature model is constructed based on features extracted from the parts storage in a preset location, the parts storage status and identity information can characterize the status and identity of the parts storage, ensuring the uniqueness of the parts storage's identity and reducing the probability of misidentification. Logical verification and matching using the preset parts storage feature model can accurately determine whether the parts storage is in place, improving the accuracy of presence detection. By matching the parts storage status and identity information with the preset parts storage feature model, presence can be detected conveniently and accurately, balancing detection efficiency and accuracy.

[0095] For example, the unique identifier or spatial pose of the parts compartment can be matched with the feature model of the parts compartment to obtain the in-situ detection result.

[0096] In one possible implementation, the parts warehouse status identity information includes a unique identifier for the parts warehouse, and the parts warehouse feature model includes identifier features extracted from the unique identifier for the parts warehouse; matching the parts warehouse status identity information with the preset parts warehouse feature model includes: matching the unique identifier for the parts warehouse with the identifier features of the parts warehouse feature model.

[0097] For example, the identifier features of the parts warehouse feature model can be understood as feature data extracted from the unique identifier of the parts warehouse, which can be used to compare with the unique identifier of the parts warehouse. For example, the hash value of the unique identifier or the feature vector after dimensionality reduction by a feature extraction algorithm. Among them, the feature extraction algorithm is, for example, Principal Component Analysis (PCA) algorithm.

[0098] The matching logic can be further refined by matching unique identifiers with identifier features. The unique identifier of the parts warehouse (such as a QR code) can be obtained through image recognition or scanning and compared with the identifier features (such as the encoding features of the QR code) in the preset parts warehouse feature model. If the match is successful, it is determined that the parts are in place; otherwise, it is determined that the parts are not in place.

[0099] In this embodiment, the unique identifier is a globally unique identifier for the accessory compartment within the entire cleaning equipment. Matching the unique identifier refines the matching logic for accessory compartment presence detection. The global uniqueness of the unique identifier ensures accurate identification of the accessory compartment, avoiding misjudgments due to duplicate labels or environmental interference, and improving the robustness of the detection. Based on this, the accuracy of accessory compartment presence detection can be enhanced, providing a more reliable foundation for the generation of subsequent cleaning strategies.

[0100] In one possible implementation, the parts compartment status identity information includes the spatial pose of the parts compartment, and the parts compartment feature model includes the in-situ spatial pose range of the parts compartment; matching the parts compartment status identity information with the preset parts compartment feature model includes: matching the spatial pose of the parts compartment with the in-situ spatial pose range of the parts compartment feature model.

[0101] For example, the in-situ spatial pose range of the accessory compartment can be understood as a preset pose and its range pre-defined for the accessory compartment. For instance, a user or cleaning device can set the in-situ pose range of the accessory compartment in the environmental map of the cleaning device through an application program (APP).

[0102] For example, in the environmental map constructed after the cleaning equipment scans the environment, the preset position and position range can be set by defining the center coordinates and boundary coordinates of the preset location of the parts compartment; the preset attitude can be set by inputting attitude parameters (such as the yaw angle), and the attitude range can be set by inputting the range of attitude parameters (such as yaw angle ±5°). Based on the preset position and position range, as well as the preset attitude and attitude range, the in-situ spatial pose range of the parts compartment can be set.

[0103] During in-situ detection, the current spatial position (x, y, z) and attitude (roll angle, yaw angle, and pitch angle) of the parts compartment are obtained, thus yielding the spatial pose of the parts compartment. This spatial pose is then matched against the in-situ spatial pose range of the parts compartment's feature model. If the spatial pose falls within this range, the match is considered successful, resulting in an in-situ detection result indicating the parts compartment is at a preset position. Conversely, if the spatial pose does not fall within this range, the match is considered unsuccessful, resulting in an in-situ detection result indicating the parts compartment is not at a preset position.

[0104] In this embodiment, the spatial pose of the parts compartment is matched with the in-situ spatial pose range of the parts compartment feature model. By using the in-situ spatial pose range as a threshold for judging the spatial pose, it is possible to effectively verify whether the parts compartment is in a preset position, avoiding detection failures caused by loose installation or displacement. The combination of the two, through rule-based or model-based matching algorithms, can improve the accuracy of in-situ detection results and provide a reliable basis for the generation of subsequent cleaning strategies.

[0105] In summary, by collecting the status and identity information of the parts warehouse and matching it with a preset parts warehouse feature model, the accuracy of parts warehouse presence detection can be improved.

[0106] For example, matching unique identifiers (such as QR codes) ensures the uniqueness of the parts compartment's identity, reducing the probability of misidentification; matching spatial pose (such as spatial coordinates and attitude parameters) verifies whether the parts compartment is in a preset position, accurately determining the physical space where the parts compartment is located and avoiding detection failures due to loose installation or displacement. By combining unique identifiers and spatial poses with parts compartment feature model matching algorithms (such as rule-based threshold judgment or machine learning classification), a logical closed loop is achieved, improving the reliability of in-situ detection results.

[0107] Ultimately, the on-site detection results can drive the generation of cleaning strategies. For example, when the accessory compartment is in place, the type information of the cleaning accessories can be prioritized, or when they are not in place, the identification process of environmental tools can be triggered, thereby enabling the cleaning equipment to adaptively respond to the status of the cleaning accessories. Based on this, the method of this application embodiment can improve the accuracy of accessory compartment on-site detection and the scenario adaptability of cleaning strategy generation, reduce the need for manual intervention by users, and enhance the intelligence level of the cleaning equipment.

[0108] For example, the in-situ detection results include two types: in-situ and out-of-situ. Different cleaning strategies can be generated for these two different in-situ detection results.

[0109] In one possible implementation, generating a cleaning strategy based on in-situ detection results includes: generating a cleaning strategy based on the cleaning accessories in the accessories compartment when the in-situ detection results indicate that the accessories compartment is in a preset position.

[0110] For example, when the in-situ detection result indicates that the accessory compartment is in a preset position, the cleaning equipment knows that the accessory compartment exists, and can then combine the cleaning accessories in the accessory compartment to generate a cleaning strategy that is more conducive to completing the cleaning task.

[0111] For example, when generating a cleaning strategy based on cleaning accessories in the accessory warehouse, a cleaning accessory could be randomly selected to generate the corresponding cleaning strategy. Alternatively, the required cleaning accessories could be determined first based on the cleaning task's requirements. Then, the accessory warehouse could be checked to see if the required cleaning accessory exists; if it does, a corresponding cleaning strategy can be generated based on that accessory.

[0112] For example, one could first iterate through all the cleaning accessories currently stored in the accessory warehouse, analyze all the accessories based on the cleaning task requirements, and weigh one or more accessories that are most conducive to efficiently completing the cleaning task. When generating and executing a cleaning strategy, multiple cleaning accessories can be used simultaneously, or one cleaning accessory can be used for cleaning followed by another, and so on, until the cleaning task is completed.

[0113] When generating a cleaning strategy based on the cleaning accessories in the accessory warehouse, the cleaning accessories in the accessory warehouse can be flexibly matched, and optimization and iterative analysis can be carried out with high cleanliness and cleaning efficiency as constraints, so that the cleaning task can be completed efficiently.

[0114] In this embodiment of the application, when the in-situ detection result indicates that the accessory compartment is in a preset position, a cleaning strategy is generated based on the cleaning accessories in the accessory compartment. This achieves the purpose of using the cleaning accessories stored in the accessory compartment to generate a targeted cleaning strategy. By using cleaning accessories, the cleaning capabilities of the cleaning equipment can be expanded, the cleaning completion rate can be improved, and the user experience can be enhanced.

[0115] In one possible implementation, generating a cleaning strategy based on cleaning accessories in the accessory compartment includes: identifying at least one cleaning accessory in the accessory compartment, determining the type of the at least one cleaning accessory, and generating a cleaning strategy based on cleaning task requirements and the type of the at least one cleaning accessory.

[0116] For example, when identifying at least one cleaning accessory in the accessory compartment, the identification can be performed according to a preset identification quantity. For instance, if the execution logic is preset to identify N cleaning accessories, then during execution, N cleaning accessories can be identified accordingly, where N can be any value greater than zero and less than the total number of cleaning accessories. Alternatively, when identifying at least one cleaning accessory in the accessory compartment, all cleaning accessories in the accessory compartment can be identified.

[0117] After identifying at least one cleaning accessory, the type of each cleaning accessory can be determined based on its accessory information or its characteristics. For example, the accessory information may include information indicating the type of cleaning accessory, such as information indicating that a brush is a sweeping type accessory or that a crevice tool is a vacuuming type accessory.

[0118] Alternatively, by analyzing the size, shape, edges, texture and / or material of at least one cleaning accessory through image recognition, the cleaning purpose of the cleaning accessory can be determined, and thus the type of cleaning accessory can be determined.

[0119] Cleaning task requirements can be understood as cleaning goals defined by the user or cleaning equipment, such as cleaning high surfaces, cleaning crevices, or cleaning corners and bends in walls. For example, a user-defined task requirement to "clean the desktop".

[0120] When generating a cleaning strategy based on the cleaning task requirements and the type of at least one cleaning accessory, for example, by analyzing the necessary characteristics required to meet the cleaning task requirements and filtering the types of at least one cleaning accessory according to the necessary characteristics, a cleaning accessory that meets the necessary characteristics is determined. This cleaning accessory can be understood as the target cleaning accessory. A corresponding cleaning strategy can be generated based on the target cleaning accessory.

[0121] For example, by analyzing the cleaning task requirements for cleaning a desktop, it can be determined that the necessary feature to meet these requirements is the ability to clean non-floor spaces. The brushes stored in the accessory compartment can be used to clean non-floor spaces, therefore it can be determined that the brushes meet the cleaning task requirements, and a cleaning strategy can be generated based on the brushes.

[0122] In this embodiment, when generating a cleaning strategy based on cleaning accessories in the accessory compartment, the type of at least one cleaning accessory is determined by identifying at least one cleaning accessory in the accessory compartment; a cleaning strategy is then generated based on the cleaning task requirements and the type of at least one cleaning accessory. In this way, by combining both the cleaning task requirements and the type of cleaning accessory, targeted analysis can be conducted to generate personalized cleaning strategies that are adapted to the current task, thereby improving the degree to which the cleaning strategy meets the cleaning task requirements.

[0123] In one possible implementation, identifying at least one cleaning accessory in the accessory compartment and determining the type of the at least one cleaning accessory includes: identifying a unique accessory identifier for at least one cleaning accessory in the accessory compartment and determining the type of the at least one cleaning accessory based on the unique accessory identifier.

[0124] For example, the type of cleaning accessory can be determined by identifying its unique identifier. Similar to the unique identifier of the accessory compartment, the unique identifier is a unique identifier for the cleaning accessory, possessing global uniqueness within the entire cleaning equipment and serving as its identification function. For instance, the unique identifier can be a QR code, serial number, predefined code, barcode, icon, and / or near-field communication tag, etc.

[0125] For example, by scanning the QR code on the cleaning accessory itself, the accessory information can be determined, such as the type of the cleaning accessory (brush, vacuum cleaner, wiping cloth, etc.). It should be understood that the unique identifier of the cleaning accessory (such as a QR code) can be set on the cleaning accessory itself or on the accessory compartment. For example, the accessory compartment has multiple storage slots, each slot can hold one cleaning accessory, and the unique identifier of the corresponding stored cleaning accessory can be marked around each storage slot. Alternatively, the unique identifier of the cleaning accessory can be displayed in other ways, which is not limited in this embodiment.

[0126] In this embodiment of the application, by identifying the unique identifier of at least one cleaning accessory in the accessory compartment and determining the type of at least one cleaning accessory based on the unique identifier, the purpose of quickly and conveniently determining the type of cleaning accessory can be achieved, thereby improving the accuracy and speed of type determination.

[0127] In one possible implementation, identifying at least one cleaning accessory in the accessory compartment and determining the type of at least one cleaning accessory includes: scanning a near-field communication tag affixed to at least one cleaning accessory in the accessory compartment using a near-field communication reader mounted on the cleaning device; identifying at least one cleaning accessory in the accessory compartment and determining the type of at least one cleaning accessory based on the data interaction identification result between the near-field communication reader and the near-field communication tag.

[0128] For example, a near-field communication (NFC) reader can be understood as a dedicated device based on NFC technology that can read or write data in carriers such as NFC tags and NFC cards within a short distance (e.g., within 10cm). NFC readers can be installed on cleaning equipment.

[0129] Near Field Communication (NFC) tags can be understood as passive electronic tags with a built-in NFC chip and antenna, requiring no power supply, and capable of being sensed by NFC readers within a short distance to complete data storage or interaction. NFC tags can be attached to cleaning accessories.

[0130] By scanning the near-field communication (NFC) tag on at least one cleaning accessory in the accessory compartment using a NFC reader installed on the cleaning equipment, the cleaning equipment can obtain accessory information through data communication between the two, thus determining the accessory's type. For example, if the cleaning equipment has a built-in accessory information mapping table, the type of each cleaning accessory can be determined by matching the mapping relationships in the table.

[0131] By setting a near-field communication tag on each cleaning accessory, the probability of misidentification can be reduced when the accessory's unique identifier is not on the cleaning accessory itself (for example, the accessory's unique identifier is on the accessory compartment itself).

[0132] For example, multiple unique identifiers for accessories are displayed on the surface of the accessory compartment, arranged from left to right as follows: accessory A above the first slot, accessory B above the second slot, and accessory C above the third slot. However, currently, cleaning accessory A, corresponding to accessory A, is stored in the second slot; cleaning accessory B, corresponding to accessory B, is stored in the third slot; and cleaning accessory C, corresponding to accessory C, is stored in the first slot. Therefore, if the type of cleaning accessory is still identified by its unique identifier, misidentification will occur.

[0133] In this embodiment of the application, by setting the near-field communication tag on the cleaning accessory body, the accessory information is embedded and bound to the corresponding cleaning accessory, which will not lead to misidentification due to reasons such as blurred unique identifier, falling off, or misplaced storage, thus improving the reliability of identification.

[0134] In one possible implementation, the cleaning strategy includes an accessory grasping parameter set, which includes at least one grasping action parameter for grasping a cleaning accessory; generating the cleaning strategy based on the cleaning task requirements and the type of at least one cleaning accessory includes: determining a target cleaning accessory that meets the cleaning task requirements based on the cleaning task requirements and the type of at least one cleaning accessory; calculating the grasping action parameters for grasping the target cleaning accessory based on the relative pose between the cleaning device and the target cleaning accessory, thereby generating the accessory grasping parameter set of the cleaning strategy.

[0135] For example, grasping action parameters can be understood as parameters that control the execution of grasping actions. These parameters can be of various types, such as grasping angle parameters, force parameters, and speed parameters. For instance, when controlling a robotic arm with a gripper at its end to perform a grasping action, the grasping parameter set may include the gripper's opening and closing degree, the rotation angle of the robotic arm joints, the spatial position of each joint of the robotic arm, and the robotic arm's movement speed.

[0136] When identifying cleaning accessories in the accessory compartment, there may be multiple cleaning accessories stored instead of just one. Therefore, it is necessary to determine the target cleaning accessory that is most beneficial to completing the cleaning task based on the cleaning task requirements.

[0137] For example, absorbent cloths and brushes are identified in the accessory compartment. For the cleaning task of cleaning water accumulation on a desktop, by analyzing the cleaning characteristics of the water (easily flowing), the absorbent cloth can be identified as the more advantageous cleaning accessory for completing the cleaning task. Therefore, the absorbent cloth can be determined as the target cleaning accessory to meet the cleaning task requirements.

[0138] When grasping a target cleaning accessory to perform a cleaning task, in order to grasp the target cleaning accessory more accurately and maintain a more favorable gripping posture for cleaning, the grasping action parameters of grasping the target cleaning accessory can be calculated based on the relative pose between the cleaning equipment and the target cleaning accessory, and the accessory grasping parameter set in the cleaning strategy can be generated.

[0139] For example, the spatial position and attitude parameters of the cleaning equipment body (or the gripping device of the cleaning equipment) and the target cleaning accessory can be obtained respectively. By calculating the difference in spatial position and attitude parameters between the cleaning equipment body (or the gripping device of the cleaning equipment) and the target cleaning accessory, the relative pose of the cleaning equipment body (or the gripping device of the cleaning equipment) and the target cleaning accessory can be obtained. This relative pose can also be understood as the pose difference. Using the relative pose as the constraint condition for inversely solving the gripping device (such as a robotic arm and its joints), at least one gripping action parameter of the gripping device can be solved inversely, and then a set of gripping parameters can be obtained. This set of gripping parameters can be understood as part of the cleaning strategy.

[0140] For example, after determining the target cleaning accessory that meets the cleaning task requirements based on the cleaning task requirements and the accessory types of multiple cleaning accessories, the relative pose between the target cleaning accessory and the robotic arm of the cleaning equipment can be determined and calculated by a device for detecting pose, such as a depth camera.

[0141] Based on the relative pose, one or more sets of robotic arm joint solutions can be obtained through optimization and iterative calculation or prediction using a pre-trained model. Each set of robotic arm joint solutions can include the spatial coordinates of the robotic arm's movement from its current position to the target position (the position where it grasps the target cleaning accessory) and key rotation angles. By controlling the motion of the robotic arm based on the optimal solution, the robotic arm can grasp the target cleaning accessory.

[0142] In this embodiment, based on the cleaning task requirements and the type of at least one cleaning accessory, a target cleaning accessory that meets the cleaning task requirements can be determined. This target cleaning accessory helps to complete the cleaning task with high cleanliness and high cleaning efficiency. Based on the relative pose between the cleaning device and the target cleaning accessory, the grasping action parameters for grasping the target cleaning accessory are calculated, thus generating an accessory grasping parameter set for the cleaning strategy. Therefore, based on the cleaning strategy including the accessory grasping parameter set, executing the cleaning strategy can successfully grasp the target cleaning accessory, facilitating the execution of the corresponding cleaning task.

[0143] In one possible implementation, the method further includes: grabbing a target cleaning accessory using a gripping device of a cleaning device according to an accessory gripping parameter set; detecting the gripping opening degree of the gripping device using a position encoder of the gripping device during the gripping process; and determining whether the gripping device has grabbed the target cleaning accessory based on the gripping opening degree.

[0144] For example, a gripping device can be understood as a mechanical structure in a cleaning device used to grip cleaning accessories or tools, such as a robotic arm mounted on the cleaning device and its end gripper. For instance, a robotic arm equipped with grippers can be used to grip cleaning accessories such as brushes or absorbent cloths. A position encoder can be understood as a sensor that measures the displacement or angle of the gripping mechanical components of the gripping device, and can be used to provide real-time feedback on the gripping status. For example, an absolute encoder can monitor the closing angle of the grippers.

[0145] After the cleaning strategy is generated, the gripping device can perform the gripping action, and the gripping opening degree can be monitored in real time by the position encoder. For example, when the gripper closes, the position encoder detects whether the gripping opening degree (the opening degree of the gripper) is less than the width of the holding part of the target cleaning accessory. If it is less, it is determined that the gripping was unsuccessful and the retry logic is triggered; if the gripping opening degree is greater than or equal to the width of the holding part of the target cleaning accessory, it can be determined that the gripping was successful. Alternatively, it can be combined with the force detection and weight detection after gripping to comprehensively determine whether the target cleaning accessory has been gripped, so as to improve the accuracy of the judgment result.

[0146] In this embodiment, the closed-loop feedback control of the gripping device significantly improves the reliability of the gripping action and increases the success rate of gripping the target cleaning accessory. For example, when gripping the target cleaning accessory, the opening and closing of the grippers and the force are adjusted in real time by adjusting the gripping opening degree, which can reduce failures or damage to the target cleaning accessory caused by fixed parameters. Since the method of this embodiment combines the closed-loop feedback process of target cleaning accessory detection, cleaning strategy generation and execution, it can improve the adaptive cleaning capability of the cleaning equipment in complex scenarios.

[0147] For example, if the cleaning equipment is not in the accessory compartment, it lacks the ability to actively identify cleaning tools in the environment, which will prevent the cleaning task from making full use of scene resources and will not be conducive to improving cleaning degree and cleaning efficiency. To this end, the method provided in this application embodiment can analyze environmental perception data through a visual model (such as an image recognition model based on deep learning), identify cleaning tools (such as mops and long-handled brushes) in the environment, and generate an appropriate cleaning strategy in combination with the cleaning task requirements.

[0148] In one possible implementation, generating a cleaning strategy based on the in-situ detection results includes: when the in-situ detection results indicate that the accessory compartment is not in a preset position, collecting environmental perception data to allow the cleaning equipment to perceive its surroundings; based on the environmental perception data, identifying whether cleaning tools exist in the environment through a visual model, wherein the cleaning tools are tools used to assist in cleaning; and generating a cleaning strategy based on the cleaning tools identified in the environment.

[0149] For example, environmental perception data can be understood as environmental information collected by cleaning equipment through sensors to identify available cleaning tools. These sensors can be, for example, depth cameras, cameras, LiDAR, millimeter-wave radar, and / or Time-of-Flight (TOF) radar. For instance, color images and depth information collected by a Red Green Blue and Depth (RGBD) camera are both environmental perception data.

[0150] Cleaning tools can be understood as auxiliary cleaning tools used in addition to cleaning accessories. These can be tools found in the environment that are not included with the manufacturer's standard equipment, such as user-provided or user-permitted brushes, small vacuum cleaners, rags, and mops. Visual models can be understood as image recognition models based on deep learning, used to detect the environment and cleaning tools within it. Examples include U-Nets. For instance, a pre-trained UNet model can be used to identify cleaning tools such as mops and long-handled brushes.

[0151] When the accessory compartment is not in place, the system can identify available cleaning tools in the environment by combining environmental perception data with a visual model. For example, the cleaning equipment captures environmental images through a camera, and the visual model analyzes the object features in the captured images (such as the planar shape of the mop) and generates a cleaning strategy (such as grabbing the mop for cleaning non-floor spaces). This process leverages the generalization ability of machine learning models to adapt to tools of different shapes and materials, expanding the coverage of cleaning scenarios.

[0152] In this embodiment, the active identification of available cleaning tools in the environment is achieved through the collection of environmental perception data and its combination with a visual model. This solution solves the problem of difficulty in generating cleaning strategies when the accessory compartment is not in place. For example, when a user places a rag on a table, the cleaning device can automatically identify and plan a cleaning path for a higher position. Through the generalization ability of the machine learning model, it adapts to the identification needs of different tools, significantly improving the flexibility and completion rate of cleaning tasks.

[0153] By introducing a visual model, the dynamic nature and scene adaptability issues of cleaning tool recognition are addressed. The visual model, based on a pre-trained feature library of cleaning tools (such as the planar shape of a mop and the rod-like structure of a long-handled brush), extracts the visual features of target objects through real-time image input (such as RGB-D camera data) and matches them with the feature library. The matching results can be further combined with cleaning task requirements (such as cleaning high surfaces or crevices) to generate strategies; for example, prioritizing the planning of high-level cleaning paths when a mop is detected. This solution leverages the generalization ability of machine learning models to adapt to cleaning tools of different shapes and materials, while avoiding the limitations of traditional rule-based matching (such as color thresholds), significantly improving the accuracy and scene coverage of environmental tool recognition.

[0154] In one possible implementation, generating a cleaning strategy based on cleaning tools identified in the environment includes: determining the tool type of the cleaning tools identified in the environment using a visual model; and generating a cleaning strategy based on cleaning task requirements and tool type.

[0155] For example, tool type can be understood as a category of cleaning tools, such as mops, long-handled brushes, telescopic poles, etc. For instance, a visual model can be used to identify the planar shape of a mop and classify it as a "planar cleaning tool." Cleaning task requirements are similar to those described in the above embodiments and can be understood as cleaning goals defined by the user or the cleaning equipment.

[0156] The system uses a visual model to categorize identified cleaning tools and generates strategies based on cleaning task requirements. For example, if the task requires cleaning high surfaces and a mop is identified, a cleaning path planning instruction to invoke the mop is generated. This process ensures the strategy matches the actual needs by matching task requirements with tool types.

[0157] In this embodiment, dynamic adaptation of cleaning strategies is achieved by matching tool types with task requirements. For example, when cleaning crevices, if a brush is detected, a strategy to invoke the brush is generated; if a mop is detected, it is ignored. This refinement step significantly improves the targeting of strategy generation, reduces resource waste, and increases cleaning efficiency.

[0158] For example, after the cleaning equipment recognizes the cleaning tool, it can use the cleaning tool to clean non-dead-angle areas or clean dead-angle areas.

[0159] In one possible implementation, a cleaning strategy is generated based on the cleaning task requirements and the tool type, including: if the cleaning task requirements include cleaning blind spots, determining whether the tool type meets the cleaning task requirements; if the tool type meets the cleaning task requirements, generating a cleaning strategy that includes cleaning blind spots based on the tool type.

[0160] For example, cleaning dead zones can be understood as areas that are difficult for the cleaning device itself to clean directly. For instance, if the cleaning device uses a combination of a bottom vacuum and a rotating brush to clean the floor, then the cleaning device cannot directly clean non-floor areas or inaccessible crevices. Therefore, cleaning dead zones include non-floor areas and inaccessible crevices.

[0161] Non-ground areas can be understood as any spatial area not on the ground, such as multiple non-ground spaces in a vertical space perpendicular to the ground, such as desktops, cabinet doors, the inside of cabinets, and walls. Gap spaces can be gaps formed between any one or more objects, such as gaps between the floor and the bed, gaps between floorboards, gaps in cable trays, or gaps between the side wall of a sofa and the wall.

[0162] Furthermore, for cleaning devices with certain shapes, such as circular disc-shaped robotic vacuum cleaners, when cleaning the inner corners of walls, obstacles such as walls may cause cleaning dead zones near the corners. Using the method described in this application, the robotic vacuum cleaner, by using its robotic arm to grab a mop, brush, etc., can perform a more thorough deep cleaning of these dead zones, improving the overall cleanliness.

[0163] Cleaning task requirements can be user-defined or generated by the cleaning equipment itself. For example, if the user issues a whole-house cleaning instruction, this instruction includes the need to clean hard-to-reach areas. The cleaning equipment can analyze the instruction and determine that these hard-to-reach areas should be cleaned, and then generate a cleaning strategy based on this requirement.

[0164] Furthermore, in order to effectively clean hard-to-reach areas, the cleaning equipment can first determine whether the type of the identified cleaning tool meets the cleaning task requirements. For example, if the cleaning task requires cleaning narrow gaps between the sofa and the wall that the robot vacuum cannot access, the robot vacuum can analyze the characteristics of the gap (long and narrow, unable to enter) and combine this with the analysis of long-bristled brush tools (tools that are longer and can reach into narrow gaps). This analysis can determine that the long-bristled brush cleaning tool meets the cleaning task requirements, and a cleaning strategy can then be generated based on this long-bristled brush.

[0165] Generating cleaning strategies based on tool type, including those for cleaning hard-to-reach areas, can be understood as generating effective cleaning strategies for these areas based on the type of cleaning tool. For example, when generating a cleaning strategy for cleaning the narrow gap between a sofa and a wall based on the type of long-bristled brush cleaning tool, this strategy includes the gripping angle of the robotic arm to allow the long-bristled brush to reach into the gap; the strategy may also include the brushing angle, speed, and number of strokes when brushing the gap; and it may further include the planned movement path of the robot vacuum and the joint angles of the robotic arm to complete the cleaning task.

[0166] In this embodiment, when the cleaning task includes the need to clean hard-to-reach areas, it can be first determined whether the cleaning tools can be used to clean these areas. If so, a corresponding cleaning strategy is generated. This allows the cleaning equipment to utilize cleaning tools found in the environment to clean hard-to-reach areas, reducing the area of ​​cleaning residue, improving cleaning completion, and enhancing the user experience.

[0167] In one possible implementation, the cleaning strategy includes a tool grasping parameter set, which includes at least one grasping action parameter for grasping a cleaning tool; generating the cleaning strategy based on cleaning task requirements and tool types includes: when multiple cleaning tools of various tool types are identified, determining a target cleaning tool that meets the cleaning task requirements based on the cleaning task requirements and the tool types of the multiple cleaning tools; calculating the grasping action parameters for grasping the target cleaning tool based on the relative pose between the cleaning device and the target cleaning tool, and generating the tool grasping parameter set of the cleaning strategy.

[0168] For example, grasping action parameters can be understood as parameters that control the execution of grasping actions. These parameters can be of various types, such as grasping angle parameters, force parameters, and speed parameters. For instance, when controlling a robotic arm with a gripper at its end to perform a grasping action, the grasping parameter set may include the gripper's opening and closing degree, the rotation angle of the robotic arm joints, the spatial position of each joint of the robotic arm, and the robotic arm's movement speed.

[0169] When identifying cleaning tools in the environment, more than one tool may be identified. Therefore, it is necessary to determine the target cleaning tool that is most beneficial to completing the cleaning task based on the requirements of the cleaning task.

[0170] For example, long-handled brushes, short-handled brushes, and toothbrushes that can be used for cleaning are identified in the environment. For cleaning tasks requiring narrow, elongated spaces, by analyzing the spatial characteristics (length) of these spaces, long-handled brushes are identified as the most advantageous cleaning tool for completing the task, among long-handled brushes, short-handled brushes, and toothbrushes that can be used for cleaning. Therefore, long-handled brushes can be identified as the target cleaning tool to meet the cleaning task requirements.

[0171] When grasping a target cleaning tool to perform a cleaning task, in order to grasp the target cleaning tool more accurately and maintain a gripping posture that is easy to clean, the grasping action parameters of grasping the target cleaning tool can be calculated based on the relative pose between the cleaning equipment and the target cleaning tool, and a tool grasping parameter set for the cleaning strategy can be generated.

[0172] For example, the spatial position and attitude parameters of the cleaning equipment body (or the gripping device of the cleaning equipment) and the target cleaning tool can be obtained respectively. By calculating the difference in spatial position and attitude parameters between the cleaning equipment body (or the gripping device of the cleaning equipment) and the target cleaning tool, the relative pose of the cleaning equipment body (or the gripping device of the cleaning equipment) and the target cleaning tool can be obtained. This relative pose can also be understood as the pose difference. Using the relative pose as the constraint condition for inversely solving the gripping device (such as a robotic arm and its joints), at least one gripping action parameter of the gripping device can be solved inversely, and then the gripping parameter set can be obtained. This gripping parameter set can be understood as part of the cleaning strategy.

[0173] For example, after determining the target cleaning tool that meets the cleaning task requirements based on the needs of the cleaning task and the types of multiple cleaning tools, the relative pose between the target cleaning tool and the robotic arm of the cleaning equipment can be determined and calculated using devices such as radar to detect the pose. Based on this relative pose, one or more sets of robotic arm joint solutions can be obtained through optimization iterative calculations or prediction using pre-trained models. Each set of robotic arm joint solutions can include the spatial coordinates of the robotic arm's movement from its current position to the target position (the position where it grasps the target cleaning tool) and key rotation angles. By controlling the motion of the robotic arm based on the optimal robotic arm solution, the robotic arm can grasp the target cleaning tool.

[0174] In this embodiment, when multiple cleaning tools of various types are identified, a target cleaning tool that meets the cleaning task requirements can be determined based on the cleaning task needs and the tool types of the multiple cleaning tools. This target cleaning tool helps to complete the cleaning task with high cleanliness and high cleaning efficiency. Based on the relative pose between the cleaning device and the target cleaning tool, the grasping action parameters for grasping the target cleaning tool are calculated, thus generating a tool grasping parameter set for the cleaning strategy. Therefore, based on the cleaning strategy including the tool grasping parameter set, executing the cleaning strategy can successfully grasp the target cleaning tool, facilitating the execution of the corresponding cleaning task.

[0175] In one possible implementation, the method further includes: grasping a target cleaning tool using a grasping device of a cleaning device according to a tool grasping parameter set; detecting the grasping opening degree of the grasping device using a position encoder of the grasping device during the grasping process; and determining whether the grasping device has grasped the target cleaning tool based on the grasping opening degree.

[0176] For example, a gripping device can be understood as a mechanical structure in a cleaning device used to grip cleaning accessories or tools, such as a robotic arm mounted on the cleaning device and its end grippers. For instance, a robotic arm equipped with grippers can be used to grip a brush or mop. A position encoder can be understood as a sensor that measures the displacement or angle of a mechanical component and can be used to provide real-time feedback on the gripping status. For example, an absolute encoder can monitor the closing angle of the grippers.

[0177] After the cleaning strategy is generated, the gripping device can perform the gripping action, and the gripping opening degree can be monitored in real time by the position encoder. For example, when the gripper closes, the position encoder detects whether the gripping opening degree (the opening degree of the gripper) is less than the width of the holding part of the target cleaning tool. If it is less, it is determined that the gripping was unsuccessful and the retry logic is triggered; if the gripping opening degree is greater than or equal to the width of the holding part of the target cleaning tool, it can be determined that the gripping was successful. Alternatively, it can be combined with the force detection and weight detection after the gripping to comprehensively determine whether the target cleaning tool has been gripped, so as to improve the accuracy of the judgment result.

[0178] In this embodiment, the reliability of the gripping action is significantly improved through closed-loop feedback control of the gripping device. For example, when gripping a brush, the opening and closing of the grippers and the force are adjusted in real time by adjusting the gripping opening and closing degree, which can reduce failures or damage to the target cleaning tool caused by fixed parameters. This solution combines a feedback closed loop of cleaning tool detection, cleaning strategy generation and execution processes, realizing the adaptive cleaning capability of the cleaning equipment in complex scenarios.

[0179] In one possible implementation, generating a cleaning strategy based on the in-situ detection results includes: when the in-situ detection results indicate that the accessory compartment is not in a preset location, calling the environmental map of the cleaning equipment, the environmental map being obtained by the cleaning equipment performing a spatial scan and map construction of its environment, the environmental map including the annotation information of obstacles; determining at least one cleaning tool among at least one obstacle based on the annotation information of at least one obstacle in the environmental map, the cleaning tool being a tool used to assist in cleaning; and generating a cleaning strategy based on at least one cleaning tool.

[0180] For example, the environment of the task space can be scanned and mapped when the cleaning equipment performs a cleaning task for the first time and on each subsequent task. For instance, the environment map can be constructed using a Simultaneous Localization and Mapping (SLAM) algorithm. During the construction of the environment map, identified items can also be labeled.

[0181] For example, semantic and location annotations can be performed on all obstacles. For instance, when constructing an environment map based on the SLAM algorithm, obstacles in space (such as table legs, walls, long-handled brushes, etc.) are identified through a visual model, and the semantics (item type) and location coordinates of each obstacle in the environment map are identified through semantic classification. Both semantics and location coordinates can serve as annotation information for obstacles.

[0182] It should be understood that cleaning equipment may interpret any object in the environment that blocks its path as an obstacle. Therefore, some cleaning tools that can be used to assist in cleaning may be interpreted as obstacles before they are identified and their types are determined.

[0183] Furthermore, objects like table legs and walls are fixed obstacles and cannot be used to assist cleaning, while long-handled brushes are obstacles that can be picked up by cleaning equipment and can be used to assist cleaning. Therefore, they can be added as cleaning tools in the labeling information. Alternatively, the tool type and location coordinates of the cleaning tool can also be added to the labeling information.

[0184] When a cleaning strategy needs to be generated and executed using cleaning tools according to the requirements of a cleaning task, the cleaning equipment can first analyze the tool type by checking the annotation information of one or more obstacles already marked on the environmental map to determine whether there are cleaning tools (obstacles) that meet the requirements of the cleaning task. If so, the equipment can move to the cleaning tool based on the location coordinates in the annotation information, using those coordinates as the destination, to grab the tool and execute the cleaning task.

[0185] As can be understood from the above embodiments, cleaning equipment employs at least two decision-making strategies when cleaning tools are required. One is to perceive the environment and identify and analyze the target cleaning tool based on the environmental perception data. The other is to access a pre-built environmental map and query whether the target cleaning tool exists based on the marking information of at least one obstacle in the map. These two decision-making strategies can be executed individually or sequentially.

[0186] In this embodiment, by calling the environmental map of the cleaning equipment and identifying at least one cleaning tool based on the obstacle markings in the environmental map, a cleaning strategy can be generated based on one or more cleaning tools. This reduces the energy and computing power costs associated with environmental perception and cleaning tool identification, and also allows for more convenient and faster acquisition of cleaning tools that can help complete the cleaning task, thereby improving cleaning speed and efficiency.

[0187] In one possible implementation, the method further includes: collecting environmental perception data of the environment when the cleaning equipment catches at least one cleaning tool; updating the environmental map based on the environmental perception data to obtain an updated environmental map.

[0188] For example, after the cleaning equipment captures at least one cleaning tool, the spatial layout and labeling information of obstacles in the environmental map can be updated. This updated environmental map can be used for subsequent navigation path planning and auxiliary cleaning strategy generation.

[0189] For example, after capturing a long-handled brush, the area occupied by the long-handled brush can be updated into a passable free space in the environment map, where waypoint search and path planning can be performed.

[0190] In this embodiment of the application, the environmental map is updated after the cleaning tool is grabbed. The obstacle marking information in the environmental map can be updated in real time. For example, after the cleaning tool is grabbed, several passable paths can be cleared. This helps to plan and generate more convenient navigation paths, reduce the situation where the cleaning equipment detours, and improve cleaning efficiency.

[0191] For example, based on the above embodiments, the method of this application embodiment can also be optimized based on multi-sensor fusion for in-situ detection of parts storage compartments.

[0192] For example, during the acquisition of the accessory compartment's status and identity information, multi-sensor fusion technology is introduced. This combines data from an NFC reader, an IMU (Inertial Measurement Unit), and a visual sensor (such as an RGBD camera). A weighted fusion algorithm is used to cross-verify the accessory compartment's unique identifier and spatial pose. For instance, if an NFC read fails, a visual sensor is used to identify a QR code or predefined pattern on the accessory compartment's surface, and IMU data is used to correct any deviations in the spatial pose.

[0193] Therefore, multi-sensor fusion can significantly improve the robustness of parts warehouse status detection. In complex scenarios (such as strong light interference or NFC tag damage), visual sensors and IMUs can serve as redundant backups, avoiding detection failures caused by the failure of a single sensor. Furthermore, by dynamically adjusting the confidence levels of each sensor through a weighted algorithm, it can adapt to detection needs in different environments; for example, it prioritizes IMU data in vibration environments and visual data in well-lit environments. Ultimately, this solution achieves high reliability and environmental adaptability in parts warehouse status detection, reducing the false positive rate.

[0194] For example, based on the above embodiments, the method of this application embodiment can also accelerate the identification of cleaning accessory types based on edge computing.

[0195] For example, upon receiving a location detection result indicating that the accessory compartment is in place, the cleaning device uses its built-in edge computing module (such as an embedded AI chip) to locally process the accessory's unique identifier, rather than relying on cloud computing. For instance, after an NFC reader scans the accessory's unique identifier, the edge computing module directly calls a pre-trained lightweight machine learning model (such as a lightweight convolutional neural network) for type recognition, avoiding delays in communication with the cloud.

[0196] Therefore, edge computing can significantly shorten the response time for accessory type recognition and improve the real-time performance of cleaning strategy generation. In scenarios where users frequently change accessories (such as home users switching between nozzles and brushes based on cleaning needs), localized processing can reduce communication latency and ensure that the device can quickly adjust its cleaning actions. Furthermore, the low-power design of the edge computing module can extend the battery life of cleaning devices while reducing reliance on network connectivity, adapting to usage needs in environments with offline or weak network connections.

[0197] For example, based on the above embodiments, the method of this application embodiment can also realize environmental tool recognition enhancement based on multimodal data fusion.

[0198] For example, when a location detection result indicating that the accessory compartment is not in place is obtained, multimodal data fusion technology is used to combine visual models and LiDAR data to improve the accuracy of cleaning tool identification in the environment. For instance, after the visual model detects an object that is suspected to be a cleaning tool, the LiDAR verifies its spatial shape (such as the planar features of a mop or the rod-like structure of a long-handled brush) using point cloud data, and performs secondary confirmation by combining it with a pre-set tool type knowledge base.

[0199] Therefore, multimodal data fusion can effectively solve the problem of misidentification by visual models in complex lighting or occlusion scenarios. For example, in strong backlighting, the visual model may mistakenly identify shadowed areas as cleaning tools, while the point cloud data from LiDAR can provide three-dimensional morphological information to help the system eliminate misjudgments. In addition, this solution can identify more types of cleaning tools (such as non-standard shaped homemade tools), expand the environmental adaptability of the equipment, and improve the flexibility and completion rate of cleaning tasks.

[0200] For example, based on the above embodiments, the method of this application embodiment can also realize adaptive optimization of grasping action parameters based on reinforcement learning.

[0201] For example, in the stage of generating grasping motion parameters, a reinforcement learning model (such as a deep Q-network, DQN) is introduced. The model is trained using historical successful / failed grasping data, and the grasping parameters (such as the joint angle of the robotic arm and the opening and closing degree of the gripper) are dynamically adjusted. For instance, the model generates the optimal grasping strategy based on the type and location of the current accessory / tool ​​and environmental interference factors (such as ground friction), and the results are fed back through a position encoder after execution to optimize the model parameters.

[0202] Based on this, reinforcement learning enables gripping strategies to be adaptive, adapting to the physical characteristics of different accessories / tools (such as the softness of brushes and the rigidity of mops) and environmental changes (such as the slipperiness of the ground). For example, when gripping a slippery brush, the model can automatically adjust the gripper opening to increase friction; when gripping a deformable mop, the model can optimize the robotic arm path to avoid deformation. Ultimately, this approach can significantly improve the gripping success rate, reduce failures caused by fixed parameters, and decrease the reliance on manual adjustments to the equipment.

[0203] For example, based on the above embodiments, the method of this application embodiment can also realize a real-time capture feedback closed loop based on a high-precision encoder.

[0204] For example, a high-precision absolute encoder can be integrated into the gripping device to monitor the rotation angle of each joint of the robotic arm and the opening and closing degree of the gripper in real time, and to dynamically analyze the feedback data through a sliding time window algorithm. For instance, if the encoder detects a sudden drop in the rate of change of the gripper opening degree during the gripper closing process, the system can determine that it has contacted the target part / tool ​​and trigger a fine-tuning action to optimize the gripping force.

[0205] Based on this, the combination of a high-precision encoder and a sliding time window algorithm enables millisecond-level response to gripping feedback, significantly reducing the latency of closed-loop control. For example, when gripping fragile parts, the system can sense the contact force of the grippers in real time and automatically adjust it to avoid damage to the parts due to excessive gripping force. In addition, this solution can adapt to the gripping needs of different parts / tools (such as brushes requiring gentle gripping and mops requiring stable gripping), improving the precision of gripping actions and reducing the risk of equipment damage.

[0206] In existing technologies, there is a lack of solutions for whether robotic vacuum cleaners equipped with robotic arms can grasp cleaning tools or detect whether cleaning accessories are in the accessory compartment. There are also no solutions for how to utilize cleaning equipment to identify cleaning tools, actively grasp them, and dynamically expand the robot's degrees of freedom based on different cleaning tools. The method provided in this application aims to address these technological gaps.

[0207] For robotic vacuum cleaners equipped with robotic arms and accessory compartments, it's necessary to check if the accessory compartment is in place before picking up cleaning accessories; otherwise, it will affect subsequent cleaning tasks. After confirming the accessory compartment is in place, it's also necessary to check if the cleaning accessories are in place and if they are the target cleaning accessories that meet the cleaning task requirements.

[0208] The following is through Figure 2 and Figure 3 The control method for the cleaning equipment provided in the embodiments of this application will be further described. Figure 2 This is a schematic diagram of the control system of the cleaning equipment provided in an embodiment of this application. Figure 3 This is a schematic diagram of the control logic of the cleaning equipment provided in an embodiment of this application. Figure 2 The control system of the cleaning equipment shown can be used to execute any of the methods provided in the embodiments of this application. The control system of the cleaning equipment can be understood as a system including electronic equipment and accessory compartments, etc. The control system of the cleaning equipment may include software modules and / or hardware modules.

[0209] like Figure 2 As shown, the control system of the cleaning equipment includes a multimodal sensing module, an accessory storage module, a grasping executability analysis module, an accessory and tool usage decision module, and a robotic arm control module.

[0210] Multimodal perception module: Equipped with a depth camera and robotic arm docking signal sensor, it collects real-time 3D environmental information and obstacle attributes (size, material). Based on a visual artificial intelligence (AI) model, it identifies graspable objects (cleaning accessories or cleaning tools) and calculates their relative pose to the robotic arm.

[0211] For example, cleaning accessories can refer to the accessories included in the accessory compartment officially provided with the robot vacuum cleaner. Cleaning tools can refer to small brushes that users can provide for the robot vacuum cleaner to use if they do not purchase the accessory compartment. The visual AI model can identify the small brushes as tools based on algorithms and training. The robot vacuum cleaner's robotic arm can then grasp and use these cleaning tools.

[0212] Accessory compartment module: Equipped with an accessory module that can communicate with the robot vacuum cleaner, providing real-time signal assistance to the robotic arm in grasping accessories. The accessory compartment and the accessories' QR codes help the machine identify whether the compartment and accessories are in place. For example, after successful docking with the robotic arm via QR code or signal docking module (such as TOF radar), the robot arm can be controlled to dock with the accessories.

[0213] The gripping feasibility analysis module is equipped with a position encoder to determine whether the gripper has opened to the appropriate position. This can be used to detect whether the gripper has grasped the desired item. It ensures that the appropriate tool or item has been grasped based on similar sizes. For example, the degree of gripper opening differs between successfully grasping a cleaning accessory and failing to do so; this difference can be used to determine whether the cleaning accessory or tool has been successfully grasped.

[0214] Equipped with a gripper camera, it can identify whether the accessory compartment is in place. For example, the gripper camera can identify whether a QR code for a cleaning accessory is present. Alternatively, it can determine whether the accessory compartment is in place based on depth.

[0215] Accessory and Tool Usage Decision Module: Based on visual AI model recognition of the grasping tools or accessories, this module determines different cleaning modes to achieve cleaning of non-floor spaces and narrow spaces. For example, it includes modes suitable for cleaning baseboards. This includes, but is not limited to, cabinet door cleaning modes, such as the robot vacuum opening a cabinet to clean it, or opening the door to clean behind the door panel. The decision is based on the mapping relationship between cleaning tools or accessories and various preset cleaning modes and scenarios.

[0216] Robotic arm control module: Based on the type of cleaning accessories and tools, it invokes pre-trained grasping strategies and plans obstacle avoidance trajectories. For example, it moves obstacles to a temporary safe area and retracts the robotic arm, or chooses to temporarily maintain the current grasping state (e.g., when there is no safe area around).

[0217] like Figure 3As shown, the control logic during cleaning is as follows: After starting, the cleaning task begins. It determines whether the user has selected the robotic arm-assisted cleaning mode. If not, the current cleaning task is completed; if so, it checks for a signal from the accessory compartment. If not, during cleaning, obstacles are identified and their semantics are marked. It then determines whether a grabbable cleaning tool / accessory for assisting cleaning is detected. If not, the current cleaning task is completed; if so, the cleaning accessory / tool ​​is grabbed.

[0218] Next, it can be determined whether the grab was successful. If not, the grabbing continues; if successful, the map is updated, and the cleaning equipment navigates to the corresponding vertical cleaning space based on the type of cleaning accessory / tool ​​grabbed. It then determines whether the grabbed cleaning accessory is from the accessory compartment. If so, the cleaning accessory is remotely opened, and the robotic arm holds the accessory and cleans the items in the vertical cleaning space. If not, the robotic arm holds the cleaning tool and cleans the items in the vertical cleaning space. Afterward, the environmental map can be updated.

[0219] Next, it can be determined whether there are any uncleaned vertical spaces. If so, the process returns to the step of grabbing cleaning accessories / tools; otherwise, the cleaning accessories are returned. If cleaning tools were grabbed, they are placed back in the user-marked cleaning tool area or back to the position where they were grabbed, and the cleaning equipment is recharged. After that, the process ends.

[0220] For example, the workflow might involve the user selecting a robotic arm-assisted cleaning mode. After cleaning begins, a localized environmental scan is initiated to identify and match cleaning accessory signals and recognize the accessory compartment's QR code. If no accessory signal or QR code is detected, the robot vacuum can identify usable cleaning tools during cleaning and perform robotic arm-assisted cleaning. Accessory signals assist the robotic arm in docking and gripping cleaning accessories, or AI can be used to identify cleaning tools. The degree of gripper opening helps determine if the corresponding tool has been gripped. A gripper camera can be used to identify whether the correct tool has been gripped. The robotic arm performs gripping and movement operations, simultaneously updating the environmental map, which may include sub-maps such as obstacle maps, narrow space maps, and cleanable non-ground areas. The robotic arm then performs the assisted cleaning operation.

[0221] The method provided in this application overcomes the limitation of traditional robotic vacuum cleaners that can only clean the ground surface, enabling the robotic arm to assist in cleaning a wider range of spaces, including non-ground areas. Based on multi-system collaborative optimization, the robotic arm control, object recognition, and path planning algorithms are deeply integrated, achieving a synergistic effect greater than the sum of its parts. Furthermore, it enhances scene adaptability, handling cleaning tasks in complex areas such as non-ground spaces and narrow areas, reducing the need for manual pre-cleaning. This method can also reuse existing robotic arm hardware resources and expand functionality through software upgrades, offering cost-effectiveness advantages.

[0222] Figure 4 This is a schematic diagram of the structure of the control device for the cleaning equipment provided in the embodiments of this application, as shown below. Figure 4 As shown in the figure, this application embodiment provides a control device for a cleaning equipment, the control device including:

[0223] The acquisition module 401 is used to acquire the status and identity information of the parts storage compartment, which is used to store cleaning parts for auxiliary cleaning equipment; the detection module 402 is used to detect whether the parts storage compartment is in a preset position based on the status and identity information of the parts storage compartment, and obtain the on-site detection result; the generation module 403 is used to generate a cleaning strategy based on the on-site detection result, and execute the cleaning task according to the cleaning strategy.

[0224] In one possible implementation, the detection module 402 is specifically used to: match the parts warehouse status identity information with a preset parts warehouse feature model, the parts warehouse feature model being a model constructed based on features extracted from the parts warehouse in a preset location; if the match is successful, obtain the on-site detection result of the parts warehouse in the preset location; if the match fails, obtain the on-site detection result of the parts warehouse not being in the preset location.

[0225] In one possible implementation, the parts warehouse status identity information includes a unique identifier for the parts warehouse, and the parts warehouse feature model includes identifier features extracted from the unique identifier of the parts warehouse; the detection module 402 is specifically used to match the unique identifier of the parts warehouse with the identifier features of the parts warehouse feature model.

[0226] In one possible implementation, the parts storage status identity information includes the spatial pose of the parts storage, and the parts storage feature model includes the in-situ spatial pose range of the parts storage; the detection module 402 is specifically used to: match the spatial pose of the parts storage with the in-situ spatial pose range of the parts storage feature model.

[0227] In one possible implementation, the generation module 403 is specifically used to: generate a cleaning strategy based on the cleaning parts in the parts compartment when the in-situ detection result indicates that the parts compartment is in a preset position.

[0228] In one possible implementation, the generation module 403 is specifically used to: identify at least one cleaning accessory in the accessory compartment, determine the type of at least one cleaning accessory, and generate a cleaning strategy based on the cleaning task requirements and the type of at least one cleaning accessory.

[0229] In one possible implementation, the generation module 403 is specifically used to: identify the accessory unique identifier of at least one cleaning accessory in the accessory compartment, and determine the type of at least one cleaning accessory based on the accessory unique identifier.

[0230] In one possible implementation, the generation module 403 is specifically used to: scan a near-field communication tag affixed to at least one cleaning accessory in the accessory compartment using a near-field communication reader installed on the cleaning device; identify at least one cleaning accessory in the accessory compartment based on the data interaction identification result between the near-field communication reader and the near-field communication tag, and determine the type of at least one cleaning accessory.

[0231] In one possible implementation, the cleaning strategy includes an accessory grasping parameter set, which includes at least one grasping action parameter for grasping a cleaning accessory; the generation module 403 is specifically used to: determine a target cleaning accessory that meets the cleaning task requirements based on the cleaning task requirements and the type of at least one cleaning accessory; calculate the grasping action parameters for grasping the target cleaning accessory based on the relative pose between the cleaning equipment and the target cleaning accessory, and generate the accessory grasping parameter set of the cleaning strategy.

[0232] In one possible implementation, the control device further includes a gripping module, which is used to: grip a target cleaning accessory using a gripping device of the cleaning equipment according to an accessory gripping parameter set; detect the gripping opening degree of the gripping device using a position encoder during the gripping process; and determine whether the gripping device has gripped the target cleaning accessory based on the gripping opening degree.

[0233] In one possible implementation, the generation module 403 is specifically used to: collect environmental perception data of the environment in which the cleaning equipment perceives the environment when the in-situ detection result indicates that the accessory compartment is not in the preset position; based on the environmental perception data, identify whether there are cleaning tools in the environment through a visual model, wherein the cleaning tools are tools used to assist cleaning; and generate a cleaning strategy based on the cleaning tools identified in the environment.

[0234] In one possible implementation, the generation module 403 is specifically used to: determine the tool type of the cleaning tools identified in the environment through a visual model; and generate a cleaning strategy based on the cleaning task requirements and the tool type.

[0235] In one possible implementation, the generation module 403 is specifically used to: determine whether the tool type meets the cleaning task requirements when the cleaning task requirements include cleaning blind spots; if the tool type meets the cleaning task requirements, generate a cleaning strategy that includes cleaning blind spots based on the tool type.

[0236] In one possible implementation, the cleaning strategy includes a tool grasping parameter set, which includes at least one grasping action parameter for grasping a cleaning tool; the generation module 403 is specifically configured to: when multiple cleaning tools of various tool types are identified, determine a target cleaning tool that meets the cleaning task requirements based on the cleaning task requirements and the tool types of the multiple cleaning tools; calculate the grasping action parameters for grasping the target cleaning tool based on the relative pose between the cleaning equipment and the target cleaning tool, and generate the tool grasping parameter set of the cleaning strategy.

[0237] In one possible implementation, the control device further includes a gripping module, which is configured to: grip a target cleaning tool using a gripping device of the cleaning equipment according to a tool gripping parameter set; detect the gripping opening degree of the gripping device using a position encoder during the gripping process; and determine whether the gripping device has gripped the target cleaning tool based on the gripping opening degree.

[0238] In one possible implementation, the generation module 403 is specifically used to: when the in-situ detection result indicates that the accessory compartment is not in a preset position, call the environmental map of the cleaning equipment, the environmental map is obtained by the cleaning equipment after spatial scanning and map construction of its environment, and the environmental map includes the annotation information of obstacles; based on the annotation information of at least one obstacle in the environmental map, determine at least one cleaning tool among at least one obstacle, the cleaning tool is a tool used to assist cleaning; and generate a cleaning strategy based on at least one cleaning tool.

[0239] In one possible implementation, the control device further includes an update module, which is used to: collect environmental perception data of the environment when the cleaning equipment picks up at least one cleaning tool; and update the environmental map based on the environmental perception data to obtain an updated environmental map.

[0240] The control device for cleaning equipment provided in this application can be used to execute the technical solution of the control method for cleaning equipment in any of the above embodiments of this application. Its implementation principle and technical effect are similar, and will not be described again here.

[0241] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application, such as... Figure 5 As shown, the electronic device of this embodiment may include: at least one processor 501; and a memory 502 communicatively connected to the at least one processor; wherein the memory 502 stores instructions executable by the at least one processor 501, the instructions being executed by the at least one processor 501 to cause the electronic device to perform the method as described in any of the above embodiments.

[0242] Optionally, the memory 502 can be either standalone or integrated with the processor 501.

[0243] The implementation principle and technical effects of the electronic device provided in this embodiment can be found in the foregoing embodiments, and will not be repeated here.

[0244] This application also provides a cleaning device, which includes a gripping device for performing gripping actions, and the cleaning device is used to implement the method of any of the foregoing embodiments.

[0245] For example, cleaning equipment could be a robotic vacuum cleaner, a smart vacuum cleaner, etc.

[0246] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the method of any of the foregoing embodiments.

[0247] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the method of any of the foregoing embodiments.

[0248] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For instance, the division of modules is only a logical functional division, and there may be other division methods in actual implementation. For example, multiple modules may be combined or integrated into another system, or some features may be ignored or not executed.

[0249] The integrated modules described above, implemented as software functional modules, can be stored in a computer-readable storage medium. These software functional modules, stored in a storage medium, include several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute some steps of the methods of the various embodiments of this application.

[0250] It should be understood that the aforementioned processor can be a Central Processing Unit (CPU) or other general-purpose processors. The processor can also be a Digital Signal Processor (DSP) or an Application Specific Integrated Circuit (ASIC), etc. A general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in the application can be directly manifested as being executed by a hardware processor, or executed by a combination of hardware and software modules within the processor.

[0251] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device, and may also be various media that can store program code, such as USB flash drives, portable hard drives, read-only memory (ROM), disks or optical discs.

[0252] The aforementioned storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof. Examples of storage media include Static Random-Access Memory (SRAM) or Electrically Erasable Programmable Read Only Memory (EEPROM).

[0253] Storage media can be, for example, erasable programmable read-only memory (EPROM) or programmable read-only memory (PROM). Storage media can also be read-only memory (ROM), magnetic storage, flash memory, magnetic disks, or optical disks. Storage media can be any available medium accessible to general-purpose or special-purpose computers.

[0254] An exemplary storage medium is coupled to a processor, enabling the processor to read information from and write information to the storage medium. Alternatively, the storage medium can be an integral part of the processor. The processor and storage medium can reside within an application-specific integrated circuit (ASIC). Alternatively, the processor and storage medium can exist as discrete components within an electronic device or host device.

[0255] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0256] The sequence numbers of the embodiments in this application are merely for description and do not represent the superiority or inferiority of the embodiments. Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method.

[0257] Based on this understanding, the technical solution of this application, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods of the various embodiments of this application.

[0258] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

[0259] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily essential to this application.

[0260] It should be further noted that although the steps in the flowchart are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise explicitly stated in this document, there is no strict order requirement for the execution of these steps, and they can be executed in other orders.

[0261] Furthermore, at least some steps in the flowchart may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but may be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but may be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0262] In the above embodiments, the descriptions of each embodiment have their own emphasis. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments. The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as the combination of these technical features does not contradict each other, it should be considered within the scope of this specification.

[0263] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this application are indicated by the following claims.

[0264] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

Claims

1. A control method for cleaning equipment, characterized in that, The method includes: Collect the status and identity information of the accessory storage compartment, which is used to store cleaning accessories that assist the cleaning equipment in cleaning; Based on the status and identity information of the parts storage compartment, detect whether the parts storage compartment is in a preset location and obtain an on-site detection result; A cleaning strategy is generated based on the in-situ detection results, and a cleaning task is executed according to the cleaning strategy.

2. The method according to claim 1, characterized in that, The step of detecting whether the parts warehouse is in a preset location based on the parts warehouse status and identity information, and obtaining an on-site detection result, includes: The status and identity information of the parts warehouse is matched with a preset parts warehouse feature model, which is a model constructed based on the features extracted from the parts warehouse in a preset location. If the match is successful, the in-situ detection result of the accessory compartment at the preset location is obtained; If the matching fails, the result of the on-site detection is obtained that the accessory compartment is not in the preset position.

3. The method according to claim 2, characterized in that, The spare parts warehouse status identity information includes the unique identifier of the spare parts warehouse, and the spare parts warehouse feature model includes identifier features extracted from the unique identifier of the spare parts warehouse; The step of matching the status and identity information of the parts warehouse with a preset parts warehouse feature model includes: The unique identifier of the parts warehouse is matched with the identifier features of the parts warehouse feature model.

4. The method according to claim 2, characterized in that, The status and identity information of the parts storage compartment includes the spatial pose of the parts storage compartment, and the feature model of the parts storage compartment includes the in-situ spatial pose range of the parts storage compartment; The step of matching the status and identity information of the parts warehouse with a preset parts warehouse feature model includes: The spatial pose of the parts compartment is matched with the in-situ spatial pose range of the parts compartment feature model.

5. The method according to any one of claims 1-4, characterized in that, The generation of a cleaning strategy based on the in-situ detection results includes: When the in-situ detection results indicate that the accessory compartment is in a preset position, a cleaning strategy is generated based on the cleaning accessories in the accessory compartment.

6. The method according to claim 5, characterized in that, The step of generating a cleaning strategy based on the cleaning accessories in the accessory compartment includes: Identify at least one cleaning accessory in the accessory compartment and determine the type of the at least one cleaning accessory; The cleaning strategy is generated based on the cleaning task requirements and the type of the at least one cleaning accessory.

7. The method according to claim 6, characterized in that, The step of identifying at least one cleaning accessory in the accessory compartment and determining the type of the at least one cleaning accessory includes: Identify the unique identifier of at least one cleaning accessory in the accessory compartment, and determine the type of the at least one cleaning accessory based on the unique identifier.

8. The method according to claim 6, characterized in that, The step of identifying at least one cleaning accessory in the accessory compartment and determining the type of the at least one cleaning accessory includes: Scan the near-field communication tag on at least one cleaning accessory in the accessory compartment using the near-field communication card reader installed on the cleaning equipment; Based on the data interaction identification results between the near-field communication reader and the near-field communication tag, at least one cleaning accessory in the accessory compartment is identified, and the type of the at least one cleaning accessory is determined.

9. The method according to claim 6, characterized in that, The cleaning strategy includes an accessory grasping parameter set, which includes at least one grasping action parameter for grasping the cleaning accessory; generating the cleaning strategy based on the cleaning task requirements and the type of the at least one cleaning accessory includes: Based on the cleaning task requirements and the type of the at least one cleaning accessory, determine the target cleaning accessory that meets the cleaning task requirements; Based on the relative pose between the cleaning device and the target cleaning accessory, the grasping action parameters for grasping the target cleaning accessory are calculated, and the accessory grasping parameter set of the cleaning strategy is generated.

10. The method according to claim 9, characterized in that, The method further includes: The target cleaning accessory is grasped by the grasping device of the cleaning equipment according to the accessory grasping parameter set. During the grasping process, the grasping opening degree of the grasping device is detected by the position encoder of the grasping device; Based on the gripping opening degree, it is determined whether the gripping device has gripped the target cleaning accessory.

11. The method according to any one of claims 1-4, characterized in that, The generation of a cleaning strategy based on the in-situ detection results includes: When the in-situ detection result indicates that the accessory compartment is not in the preset position, environmental perception data is collected to enable the cleaning equipment to perceive its surrounding environment. Based on the environmental perception data, a visual model is used to identify whether cleaning tools exist in the environment, and the cleaning tools are tools used to assist in cleaning. The cleaning strategy is generated based on the cleaning tools identified within the environment.

12. The method according to claim 11, characterized in that, The step of generating the cleaning strategy based on the cleaning tools identified in the environment includes: The visual model is used to determine the tool type of the cleaning tools identified in the environment; The cleaning strategy is generated based on the cleaning task requirements and the tool type.

13. The method according to claim 12, characterized in that, The step of generating the cleaning strategy based on the cleaning task requirements and the tool type includes: If the cleaning task requires cleaning hard-to-reach areas, determine whether the tool type meets the cleaning task requirements. If the tool type meets the cleaning task requirements, a cleaning strategy including cleaning blind spots is generated based on the tool type.

14. The method according to claim 12, characterized in that, The cleaning strategy includes a tool grasping parameter set, which includes at least one grasping action parameter for grasping the cleaning tool; generating the cleaning strategy based on the cleaning task requirements and the tool type includes: When multiple cleaning tools of various tool types are identified, a target cleaning tool that meets the cleaning task requirements is determined based on the cleaning task requirements and the tool types of the multiple cleaning tools. Based on the relative pose between the cleaning device and the target cleaning tool, the grasping action parameters for grasping the target cleaning tool are calculated, and the tool grasping parameter set of the cleaning strategy is generated.

15. The method according to claim 14, characterized in that, The method further includes: The target cleaning tool is grasped by the grasping device of the cleaning equipment according to the tool grasping parameter set; During the grasping process, the grasping opening degree of the grasping device is detected by the position encoder of the grasping device; Based on the gripping opening degree, it is determined whether the gripping device has gripped the target cleaning tool.

16. The method according to any one of claims 1-4, characterized in that, The generation of a cleaning strategy based on the in-situ detection results includes: If the in-situ detection result indicates that the accessory compartment is not in the preset position, the environmental map of the cleaning equipment is invoked. The environmental map is obtained by the cleaning equipment after spatial scanning and map construction of its environment, and the environmental map includes the annotation information of obstacles. Based on the labeling information of at least one obstacle in the environmental map, at least one cleaning tool is identified among the at least one obstacle, and the cleaning tool is a tool used to assist in cleaning; The cleaning strategy is generated based on the at least one cleaning tool.

17. The method according to claim 16, characterized in that, The method further includes: When the cleaning equipment catches at least one cleaning tool, environmental perception data of the environment is collected; The environmental map is updated based on the environmental perception data to obtain an updated environmental map.

18. A cleaning device, characterized in that, The cleaning equipment includes a gripping device for performing a gripping action, and the cleaning equipment is used to perform the method as described in any one of claims 1-17.

Images