Method for controlling cleaning robot, storage medium, and related device

Abstract

A method for controlling a cleaning robot, comprising: in response to a pet search instruction issued by a target user, acquiring activity heatmap information of a target pet, the activity heatmap information of the target pet being distribution information of activity frequency in different regions which is obtained on the basis of historical activity information of the target pet; and on the basis of the activity heatmap information of the target pet, controlling the cleaning robot to move so as to search for the target pet.

Description

A cleaning robot control method, storage medium, and related equipment Cross-references to related applications

[0001] This application claims priority to Chinese patent application No. 202510147823.5, filed on February 10, 2025, the entire contents of which are incorporated herein by reference. Technical Field

[0002] This disclosure relates to the field of cleaning robot control technology, and in particular to a cleaning robot control method, storage medium and related equipment. Background Technology

[0003] With the widespread adoption of smart home devices, robotic vacuum cleaners have become indispensable cleaning assistants for many families. In recent years, the functional scope of robotic vacuum cleaners has gradually expanded, extending from traditional floor cleaning to more complex home management tasks, such as security monitoring and air quality detection. Pet-finding functionality, as an emerging intelligent feature, is gradually gaining attention and application from users. This function allows the robotic vacuum cleaner to locate and track the position of pets in the home environment, helping users monitor and interact with them in real time.

[0004] In related technologies, pet-finding functions struggle to accurately identify and locate targets in complex home environments or under dynamic pet behavior conditions, and are time-consuming and inefficient.

[0005] Therefore, it is necessary to propose a control method for cleaning robots to improve the efficiency and accuracy of pet finding. Summary of the Invention

[0006] This disclosure aims to at least address the problems of poor accuracy and efficiency in pet-finding by cleaning robots in related technologies.

[0007] In view of this, in a first aspect, this disclosure proposes a cleaning robot control method, comprising: in response to a pet search command issued by a target user, acquiring target pet activity heat information; the target pet activity heat information being distribution information of activity frequency in different areas obtained based on the target pet's historical activity information; and controlling the cleaning robot to move according to the target pet activity heat information in order to search for the target pet.

[0008] In some implementations, controlling the cleaning robot to move based on the target pet's activity thermal information to locate the target pet includes:

[0009] The first planned path is generated based on the target pet activity heat map information and path planning algorithm.

[0010] The cleaning robot is controlled to move based on the first planned path to find the target pet.

[0011] In some implementations, controlling the cleaning robot to move based on the target pet's activity thermal information to locate the target pet includes:

[0012] The heat map information of the target pet's activities will be displayed on the control terminal of the target user.

[0013] Obtain the search area determined by the target user based on the aforementioned heatmap information of the target pet's activity;

[0014] Control the cleaning robot to move within the search area to locate the target pet.

[0015] In some implementations, the cleaning robot control method further includes:

[0016] Obtain manually added markers from the target user;

[0017] Based on the manual addition of markers, the cleaning robot is controlled to locate the target pet.

[0018] In some implementations, the above-mentioned response to a pet search command issued by a target user to obtain target pet activity heatmap information includes:

[0019] The target pet is determined based on the pet search command issued by the target user.

[0020] Match the target pet's activity heat information to the target pet in the heat information database.

[0021] In some implementations, the aforementioned heatmap database stores heatmap information on the activities of multiple target pets.

[0022] In some implementations, the pet search instruction includes an image matching instruction.

[0023] The process of determining the target pet based on the pet search command issued by the target user includes:

[0024] Retrieve the pet image information corresponding to the above image matching command;

[0025] The target pet was identified based on the pet image information and image recognition algorithm.

[0026] In some implementations, the pet search instructions include voice search instructions.

[0027] The process of determining the target pet based on the pet search command issued by the target user includes:

[0028] The above voice search command is subjected to speech recognition to obtain semantic information;

[0029] The target pet was determined based on the aforementioned semantic information and semantic recognition algorithm.

[0030] In some implementations, the pet search instruction includes a target pet selection instruction.

[0031] The process of determining the target pet based on the pet search command issued by the target user includes:

[0032] Based on the pet display information selected by the target user on the target user's control terminal, a pet search instruction is generated; the pet display information includes at least one of the pet icon, pet name, and pet thumbnail.

[0033] Determine the target pet based on the above pet search instructions.

[0034] In some implementations, controlling the cleaning robot to move based on the target pet's activity thermal information to locate the target pet includes:

[0035] Based on the target pet activity heat map information and preset thresholds, determine the heat map locations;

[0036] If there is only one heat source location, control the cleaning robot to move to the heat source location in order to find the target pet.

[0037] If there are at least two of the aforementioned heat points, a second planning path is generated based on all of the aforementioned heat points;

[0038] The cleaning robot is controlled to move based on the second planned path described above in order to find the target pet.

[0039] In some implementations, the pet search instruction includes information about the pet's location that the target user has guessed.

[0040] The above methods also include:

[0041] A third planned route is generated based on all the above-mentioned heat maps and the pet location information guessed by the target users.

[0042] The cleaning robot is controlled to move based on the third planned path described above in order to find the target pet.

[0043] In some implementations, the cleaning robot control method further includes:

[0044] Based on the cleaning task and / or pet search task, obtain the image information and pet location information of the target pet;

[0045] Based on the image information and location information of the target pet, a deep learning algorithm is used to determine the thermal information of the target pet's activity.

[0046] In some implementations, the aforementioned target pet activity thermal information also includes time information.

[0047] The above-mentioned response to the pet search command issued by the target user, obtaining the target pet's activity heatmap information, includes:

[0048] In response to a pet search command issued by the target user, determine the current time;

[0049] Based on the current time, obtain the heat map information of the target pet's activity that matches the current time.

[0050] Secondly, this disclosure also proposes a control device for a cleaning robot, comprising:

[0051] The acquisition module is configured to acquire the target pet's activity heatmap information in response to a pet search command issued by the target user; the target pet's activity heatmap information is the distribution information of activity intensity in different areas obtained based on the target pet's historical activity information;

[0052] The search module is configured to control the movement of the cleaning robot based on the target pet's activity thermal information in order to locate the target pet.

[0053] In some embodiments, the searching module includes: a first path generation unit and a first control unit.

[0054] The first path generation unit is configured to generate a first planned path based on the target pet activity heat map information and the path planning algorithm;

[0055] The first control unit is configured to control the movement of the cleaning robot based on the first planned path in order to find the target pet.

[0056] In some implementations, the searching module includes: an information display unit, an area determination unit, and a second control unit.

[0057] The information display unit is configured to display the target pet's activity heat map information on the target user's control terminal;

[0058] The area determination unit is configured to acquire the search area determined by the target user based on the target pet's activity heat map information;

[0059] The second control unit is configured to control the cleaning robot to move within the search area to find the target pet.

[0060] In some embodiments, the cleaning robot control device further includes: a marker acquisition module and a manual addition control module.

[0061] The marker acquisition module is configured to acquire manually added markers by the target user;

[0062] The manual addition control module is configured to control the cleaning robot to find the target pet based on the manually added marker points.

[0063] In some embodiments, the acquisition module includes: a first search unit and a first matching unit.

[0064] The first search unit is configured to determine the target pet based on the pet search command issued by the target user;

[0065] The first matching unit is configured to match the target pet's activity thermal information in the thermal information database.

[0066] In some implementations, the heat map database stores heat map information on the activities of multiple target pets.

[0067] In some implementations, the pet search instruction includes an image matching instruction.

[0068] The first search unit is configured to obtain pet image information corresponding to the image matching instruction;

[0069] The first search unit is also configured to determine the target pet based on the pet image information and the image recognition algorithm.

[0070] In some implementations, the pet search instructions include voice search instructions.

[0071] The first search unit is configured to perform speech recognition on the voice search command to obtain semantic information;

[0072] The first search unit is also configured to determine the target pet based on the semantic information and the semantic recognition algorithm.

[0073] In some implementations, the pet search instruction includes a target pet selection instruction.

[0074] The first search unit is configured to generate a pet search instruction based on the pet display information selected by the target user on the target user's control terminal; the pet display information includes at least one of a pet icon, a pet name, and a pet thumbnail.

[0075] The first search unit is also configured to determine the target pet according to the pet search instruction.

[0076] In some embodiments, the searching module includes: a first thermal point determination unit and a third control unit.

[0077] The first thermal point determination unit is configured to determine thermal points based on the thermal information of the target pet's activity and a preset threshold.

[0078] When there is only one heat source, the third control unit is configured to control the cleaning robot to move to the heat source in order to find the target pet.

[0079] In some embodiments, the searching module includes: a second thermal point determination unit, a second path generation unit, and a fourth control unit.

[0080] The second thermal point determination unit is configured to determine thermal points based on the thermal information of the target pet's activity and a preset threshold.

[0081] When there are at least two heat points, the second generation unit is configured to generate a second planned path based on all of the heat points;

[0082] The fourth control unit is configured to control the movement of the cleaning robot based on the second planned path in order to find the target pet.

[0083] In some implementations, the pet search instruction includes information about the pet's location that the target user has guessed.

[0084] The device further includes: a third path generation module and a third path control module.

[0085] The third path generation module is configured to generate a third planned path based on all the heat map locations and the pet location information guessed by the target user.

[0086] The third path control module is configured to control the movement of the cleaning robot based on the third planned path in order to find the target pet.

[0087] In some embodiments, the cleaning robot control device further includes:

[0088] The information acquisition module is configured to acquire image information and pet location information of the target pet based on cleaning tasks and / or pet search tasks;

[0089] The thermal information determination module is configured to determine the thermal information of the target pet's activity based on the target pet's image information and the pet's location information using a deep learning algorithm.

[0090] In some implementations, the target pet activity thermal information also includes time information.

[0091] The acquisition module includes: a time determination unit and a second matching unit.

[0092] The time determination unit is configured to determine the current time in response to a pet search command issued by the target user;

[0093] The second matching unit is configured to acquire heat information of target pet activity that matches the current time based on the current time.

[0094] Thirdly, this disclosure also proposes an electronic device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor is configured to implement the steps of the cleaning robot control method as described in any of the first aspects above when executing the computer program stored in the memory.

[0095] Fourthly, this disclosure also proposes a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the cleaning robot control method of any one of the first aspects.

[0096] Fifthly, this disclosure also proposes a cleaning robot, including the electronic device as described in claim 3. Attached Figure Description

[0097] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit this specification. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0098] Figure 1 is a schematic flowchart of a cleaning robot control method according to some embodiments of the present disclosure;

[0099] Figure 2 is a structural schematic diagram of a possible sweeping robot according to some embodiments of the present disclosure;

[0100] Figure 3 is a structural schematic diagram of another possible sweeping robot according to some embodiments of the present disclosure;

[0101] Figure 4 is a schematic diagram of a cleaning robot working scene according to some embodiments of the present disclosure;

[0102] Figure 5 is a structural schematic diagram of a cleaning robot control device according to some embodiments of the present disclosure;

[0103] Figure 6 is a structural schematic diagram of an electronic device according to some embodiments of the present disclosure; and

[0104] Figure 7 is a schematic diagram of a cleaning robot structure according to some embodiments of the present disclosure. Detailed Implementation

[0105] The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this disclosure are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus. The technical solutions of this disclosure will now be clearly and completely described in conjunction with the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this disclosure, and not all of them.

[0106] Please refer to Figure 1, which is a flowchart of a cleaning robot control method according to some embodiments of the present disclosure, and may include steps S110-S120.

[0107] In S110, in response to the pet search command issued by the target user, the target pet activity heat map information is obtained; the target pet activity heat map information is the distribution information of activity frequency in different areas obtained based on the target pet's historical activity information.

[0108] For example, a target user can issue a "find pet" command to the cleaning robot via a smart terminal, such as a mobile app or voice assistant. Upon receiving the command, the cleaning robot activates its pet search mode. The cleaning robot analyzes the target pet's historical activity data locally or in its corresponding cloud environment. This historical activity data can be collected through smart cameras installed in the home, pet wearable devices (such as tracking collars), or sensors built into the cleaning robot, including RGB cameras and LiDAR. Thermal information can be a distribution map obtained by statistically analyzing the frequency of pet activity in different areas. Thermal information can be generated by collecting pet movement data through smart collars or GPS trackers and calculating dwell time and visit frequency in different areas. Thermal information can also be obtained by deploying cameras in specific areas and using computer vision technology to identify pet activity locations and statistically analyze their distribution. Furthermore, thermal information can be obtained by using environmental sensors (such as pressure sensors and sound detectors) to detect pet behavior data, thereby determining the activity level of a particular area.

[0109] Heatmap data can include areas where pets are frequently active (such as pet beds and kitchens) as well as areas where they are infrequently active (such as corners and hard-to-reach areas). This distribution information is usually presented in the form of a heatmap, indicating the probability of pet activity in different areas.

[0110] In step S120, the cleaning robot is controlled to move according to the target pet's activity thermal information in order to find the target pet.

[0111] In some embodiments, the cleaning robot generates a path plan based on a pet activity heatmap, prioritizing navigation to areas where the pet is frequently active. During path planning, the cleaning robot combines built-in maps (such as Simultaneous Localization and Mapping (SLAM) technology) with heatmap information to avoid furniture obstacles and ensure efficient movement. If the cleaning robot fails to find the target pet in a high-probability area, the system can switch to a medium- or low-probability area to continue the search, based on a set threshold. During movement, real-time sensor data, such as camera recognition and microphone sound detection, can be used to further narrow down the target area. The cleaning robot's built-in visual module (such as a camera or thermal infrared sensor) or sound recognition module (such as detecting barks or collar signals) confirms the target pet's location. Once the target pet is found, the cleaning robot notifies the user, such as through an app push notification or voice announcement.

[0112] Figure 2 is a structural schematic diagram of a possible robotic vacuum cleaner according to some embodiments of the present disclosure. The processor of the cleaning robot body serves as the core control unit, executing logic related to cleaning tasks and pet location. The cleaning task plans a path based on stored map and heatmap information, directing the moving components to complete efficient cleaning. The pet location task combines deep learning algorithms and a vision module to process images, generating real-time heatmap information of the target pet's activity, and navigating the cleaning robot to high-probability areas. The processor can also dynamically receive location information from the charging station, ensuring automatic navigation back to the charging station when the battery is low. The communication module enables data interaction with the charging station or external devices (such as a mobile app). During pet location, it receives pet location instructions (such as "Find Xiaobai") from the user's smart terminal and can send the current task progress to the charging station or an external server. During cleaning tasks, it receives cleaning area instructions and provides feedback on the task status after completion. The moving components can support precise navigation based on heatmap information or path planning. In pet location tasks, the moving components prioritize high-probability areas on the heatmap (such as the pet's bed or kitchen). The vision module serves as input for executing deep learning algorithms. In pet-finding tasks, images of pets can be captured, their shape, color, and movements identified, and real-time heatmaps updated. In cleaning tasks, obstacle and furniture locations are identified to aid dynamic path planning. Upon user commands to find a pet, real-time pet activity information is captured to confirm its location. The memory stores historical activity heatmaps, map data, and task records. Heatmap information can be combined with a time dimension to record the distribution of the target pet's activity at different times. The memory also supports saving cleaning task progress and planning future tasks.

[0113] The charging station's processor manages communication with the robotic vacuum cleaner, provides location information, and supports automatic recharging. During pet-finding tasks, when the robot's battery level drops below a threshold, the communication module prompts it to return to the charging station. The communication module provides the robot with real-time location information, ensuring it accurately returns to the charging station for recharging.

[0114] In summary, this disclosure proposes a cleaning robot control method that utilizes multiple data sources to generate activity heatmap information of the target pet, comprehensively considering historical behavioral data and real-time environmental information, significantly improving the accuracy of pet identification and location. The target user can flexibly issue commands to find the pet via a smart terminal (such as a mobile app or voice assistant), and the cleaning robot automatically enters pet search mode upon receiving the command. The target pet activity heatmap information intuitively displays the probability distribution of pet activity to the user, who can view the robot's search process and progress in real time through an interactive interface. This disclosed control method enables the cleaning robot to collect pet information and locate the pet while performing cleaning tasks. Users do not need to purchase additional independent equipment to achieve multi-functional integration on the cleaning robot. This disclosed cleaning robot control method, through multi-modal data fusion, heatmap analysis, path planning optimization, and flexible user interaction design, effectively overcomes the problems of insufficient identification accuracy, low efficiency, and limited user experience in related technologies, significantly improving the practicality and intelligence level of the pet-finding function.

[0115] In some embodiments, controlling the movement of the cleaning robot based on the target pet's activity thermal information to locate the target pet includes:

[0116] Based on the aforementioned target pet activity heatmap information and path planning algorithm, a first planned path is generated; and

[0117] The cleaning robot is controlled to move based on the first planned path to find the target pet.

[0118] In some embodiments, the target pet activity thermal information is activity data of the target pet collected by the cleaning robot through multiple data sources. This data can be pet dynamic monitoring data provided by a home smart camera, location information transmitted by a pet wearable device (such as a positioning collar), or pet activity information recorded by the cleaning robot's own sensors (such as cameras and infrared sensors).

[0119] A preliminary planned path is generated based on the target pet's activity heatmap and a path planning algorithm. The path planning algorithm can be an A* algorithm, Dijkstra's algorithm, or similar algorithms. The preliminary planned path prioritizes high-probability areas and gradually covers medium- and low-probability areas. The path is also dynamically optimized considering factors such as the cleaning robot's battery level and obstacle avoidance requirements.

[0120] If the thermal information is updated in real time (e.g., sensors detect a new pet location), the path planning algorithm can dynamically adjust the initial planned path. The initial planned path can be adjusted based on sensor information to avoid dynamic obstacles such as furniture. When a pet is detected moving in a specific area, the algorithm directly replans the path to that area.

[0121] The cleaning robot moves sequentially according to the generated initial planned path, with the navigation module ensuring the accuracy of path execution. In high-probability areas, the cleaning robot employs a small-scale search pattern (such as a spiral path) to ensure that no key points are missed.

[0122] During movement, the cleaning robot can perceive the pet's location in real time through, but not limited to, the following methods: recognizing the pet's shape, color, and other characteristics through a visual module; capturing the pet's barks or collar signals through an audio module; and detecting the pet's body temperature through a thermal sensor module.

[0123] If the target pet is not found in a high-probability area, the cleaning robot will automatically adjust its path, reorder the remaining low-to-medium probability areas, and continue the search. If the real-time sensors detect the target pet, the system will directly generate a new path and head towards the target location.

[0124] After confirming the target using vision, sound, or infrared modules, the cleaning robot sends a notification to the user's device. The pet's location is marked in the app, and real-time video or voice announcements are provided. The user is given feedback on the covered area and suggestions on whether to expand the search area or regenerate the route.

[0125] This embodiment intelligently generates a first planned path and controls the cleaning robot to move efficiently by combining the target pet's activity heat map information with a path planning algorithm. This method significantly improves the efficiency and accuracy of pet-finding tasks, enhances the system's flexibility and user experience, and is suitable for diverse home scenarios and complex environments.

[0126] In embodiments of this disclosure, controlling the movement of the cleaning robot based on the target pet's activity thermal information to locate the target pet includes:

[0127] The heat map information of the target pet's activities will be displayed on the control terminal of the target user.

[0128] Obtain the search area determined by the target user based on the aforementioned heatmap information of the target pet's activity; and

[0129] Control the cleaning robot to move within the search area to locate the target pet.

[0130] In some embodiments, after acquiring the heat map information of the target pet's activity, the cleaning robot can display it as a visual heat map on the user's control terminal (such as a mobile app or tablet). The target pet's activity heat map information can be displayed using color gradients to represent the probability of pet activity in different areas; for example, red represents areas with high activity frequency, and yellow or green represents areas with medium to low activity frequency.

[0131] On the control terminal, target users can independently select or adjust the search area based on the target pet's activity heatmap information and their own needs. Area selection can be done manually or through other quick options. Manual selection allows users to directly circle specific areas on the heatmap. Quick options allow users to quickly confirm or adjust based on system-recommended high-probability areas. Furthermore, target users can select multiple areas to search, which is particularly useful for areas where pets may frequently be active, such as pet beds and balconies.

[0132] The cleaning robot generates a precise movement path based on the user-specified search area and path planning algorithms. Within the designated search area, the robot prioritizes navigation to areas with high activity according to probability distribution. Utilizing SLAM technology and sensors, it dynamically avoids obstacles such as furniture and walls, ensuring efficient movement. If the user selects multiple search areas, the robot automatically switches to the next target area after completing the current one. During the search, the robot uses multimodal perception modules (such as cameras, microphones, and infrared sensors) to detect the presence of the target pet in real time. Visual recognition involves capturing the pet's morphological features in real time using the camera. Sound localization uses the microphone to detect the pet's barks or signals emitted from its collar, further narrowing the search area. Thermal sensing detects the pet's body temperature to help identify where the pet is hiding (such as under a sofa or bed). If the cleaning robot does not find the target pet in the designated area, the system dynamically adjusts its path based on activity thermal information, prioritizing the inspection of other potential areas.

[0133] When the cleaning robot successfully locates the target pet, it notifies the user of the target location via the user terminal (such as an app push or voice assistant), or marks the pet's location on the terminal map and prompts "Pet found".

[0134] If the target pet is not found, the cleaning robot will provide feedback to the user on the search progress (such as covered areas and remaining uncovered areas). It will suggest whether to expand the search area or enable the whole-house search mode.

[0135] The method proposed in this embodiment incorporates the opinions of target users into the search process. By using heatmaps to assist users in selecting areas, it avoids the inefficiency of full-coverage searches, resulting in a more precise search range and significantly reduced search time. By providing heatmap visualization and flexible area selection capabilities, users can adjust their strategies in real time to meet personalized needs. Combining user-specified areas with optimized cleaning robot paths significantly increases the probability of successfully finding the pet within the target area. Through the method provided in this embodiment, the cleaning robot not only achieves significant improvements in efficiency and accuracy but also enhances practicality and user experience through intuitive user interaction design, offering a more flexible solution for the practical application of pet-finding functions.

[0136] In embodiments of this disclosure, the method further includes:

[0137] Obtain manually added markers from the target user; and

[0138] Based on the manual addition of markers, the cleaning robot is controlled to locate the target pet.

[0139] In some embodiments, the target user can manually add target markers by observing pet activity heatmaps using a control terminal (such as a mobile app, tablet, or voice assistant). These markers specify areas where the pet may be active or hiding, such as a room or near furniture. The markers supplement the system-generated heatmaps, covering areas the user deems potentially potential targets.

[0140] Marker points can be set using one or more combinations of methods, including direct selection, voice input, and text description. Direct selection involves the user clicking on the target location on the home map interface to generate a marker. Voice input allows the user to specify the location via voice commands, such as "Looking for my pet, it might be on the balcony." Text description allows the user to input a specific location description, and the system automatically locates the corresponding map position.

[0141] Users can set the search priority of markers and allow users to add multiple markers at the same time. Users can add, delete, or adjust the position and priority of markers at any time.

[0142] The cleaning robot combines user-marked points with heatmap information about pet activity, prioritizing both to generate a path. If a marker has higher priority than a heatmap area, the cleaning robot will prioritize visiting the marker. Path planning dynamically balances the coverage order between markers and heatmap areas to ensure maximum efficiency.

[0143] This disclosure significantly improves the flexibility, accuracy, and user experience of a cleaning robot's pet-finding function by combining target user-added markers with thermal information and path planning. Whether in single-marker or multi-marker scenarios, this method can efficiently complete target searches, meeting diverse household needs.

[0144] In embodiments of this disclosure, the above-mentioned response to a pet search command issued by a target user to obtain target pet activity heatmap information includes:

[0145] The target pet is determined based on the pet search command issued by the target user; and

[0146] Match the target pet activity heat information corresponding to the above target pet in the heat information database; the above heat map database stores the pet activity heat information of multiple target pets.

[0147] In some embodiments, the target user issues a "Find Pet" command via a smart terminal (such as a mobile app or voice assistant). The target user may have multiple target pets in their home, and a heatmap database stores heatmap information on the activity of these pets. The target user can select one or more target pets to search for. This selection can be made through a list, voice command, or by inputting images or features on the smart terminal. Using a list allows the target user to choose the pet on the smart terminal interface. Using a voice command allows the target user to directly say the pet's name, such as "Find Xiaobai." Inputting images or features allows the target user to upload a photo of the pet or select pre-defined pet features (such as color, size, etc.).

[0148] The system determines the target pet based on the user's input. If the user owns multiple pets, the system selects and locks the target pet. The system also supports searching for multiple pets at once. The system verifies whether the target pet is already linked to the system (e.g., identified by a collar, tag, etc.). If not linked, the system prompts the user to register or select another pet.

[0149] The heatmap database stores activity heatmap information for multiple target pets. Each pet's activity heatmap is generated based on its historical activity data, and includes, but is not limited to, the following: spatial distribution information, temporal dimension information, and specific behavioral habits. Spatial distribution information includes the frequency of the pet's activity in different areas of the home. Temporal dimension information includes the pet's activity patterns at different times of day. Specific behavioral habits include preferred hiding places, etc.

[0150] Pet activity heatmaps are collected through devices worn by pets, home cameras, or sensors on cleaning robots. The database is categorized by pet identity for easy and quick searching. The system matches the target pet's activity heatmap against the corresponding information in the heatmap database. Alternatively, the system searches the database based on the pet identifier entered by the user. It returns an activity heatmap of the target pet, containing spatial distribution information on activity probability. If multiple target pets are being searched, the system will match multiple heatmaps simultaneously and display the activity distribution for each pet separately. The system also supports dynamic updates to the heatmaps, continuously supplementing them with the latest pet activity data to ensure the heatmaps reflect current activity trends.

[0151] In embodiments of this disclosure, the pet search instruction includes an image matching instruction.

[0152] The process of determining the target pet based on the pet search command issued by the target user includes:

[0153] Retrieve the pet image information corresponding to the above image matching command; and

[0154] The target pet was identified based on the pet image information and image recognition algorithm.

[0155] In some embodiments, target users can upload pet photos of their target pets via smart devices (such as mobile apps or tablets). The pet photos can be close-ups of the pet, including facial features, body shape, and other information. The system supports various image upload formats, such as real-time shooting, album selection, or importing from social media.

[0156] The system uses image recognition algorithms (such as deep learning models: ResNet, YOLO, etc.) to analyze the pet's appearance features. Extracted features may include, but are not limited to, color features, morphological features, facial features, and feature markers. Color features include the distribution of the pet's coat color. Morphological features include the shape of the ears and tail, and the overall body shape. Facial features include the shape of the eyes and the position of the nose. Feature markers include unique markings, scars, or collars.

[0157] The image recognition algorithm compares the extracted features with pet feature templates in the heat map database. It then matches the extracted pet features against the pet information already linked in the heat map database. If a unique match is found, the system confirms the target pet's identity. If multiple pets are matched, the system lists possible options for the target user to confirm. If no target pet is matched, the system prompts the user to link a new pet or re-upload images. The target user can flexibly operate through image upload commands to adapt to diverse needs.

[0158] In embodiments of this disclosure, the pet search instructions include voice search instructions.

[0159] The process of determining the target pet based on the pet search command issued by the target user includes:

[0160] The above voice search command is subjected to speech recognition to obtain semantic information; and

[0161] The target pet was determined based on the aforementioned semantic information and semantic recognition algorithm.

[0162] In some embodiments, the target user issues voice commands to the system via a smart terminal. The content of the voice commands may include, but is not limited to: the pet's name, the pet's characteristics, location prompts, and action requests.

[0163] The system uses speech recognition algorithms to convert speech signals into text. During the speech recognition process, an adaptive model can be used to improve recognition accuracy for different language environments or target user accents.

[0164] The system analyzes the semantic information of voice commands based on semantic recognition algorithms, extracts key content, and matches this extracted semantic information with the pet's already bound information. The system searches a heatmap database for records matching the pet's name or characteristics. If a unique pet is matched, the target pet is confirmed. If multiple possible matches exist, the system requests more information from the target user to further confirm the target. If the target user is searching for multiple pets simultaneously, the system processes the semantic information separately, generating corresponding heatmaps and search plans. The system also supports combined commands, such as "Look for the white cat and the orange cat; the orange cat might be in the living room."

[0165] The method proposed in this disclosure, through speech recognition and semantic analysis combined with thermal information and path planning technology, enables cleaning robots to quickly respond to voice search commands. Its efficient semantic extraction, dynamic path adjustment, and real-time notification functions not only improve the accuracy and efficiency of pet retrieval but also significantly enhance the convenience and flexibility of the user experience.

[0166] In embodiments of this disclosure, the pet search instruction includes a target pet selection instruction.

[0167] The process of determining the target pet based on the pet search command issued by the target user includes:

[0168] Based on the pet display information selected by the target user on the target user's control terminal, a pet search instruction is generated; the pet display information includes at least one of a pet icon, a pet name, and a pet thumbnail; and

[0169] Determine the target pet based on the above pet search instructions.

[0170] In some embodiments, the target user views pet information bound to the system via a control terminal. This information includes one or a combination of pet icons, pet names, and pet thumbnails. For example, the user can view a pet icon and pet name, a pet name and pet thumbnail, a pet icon and pet thumbnail, or a pet icon, pet name, and pet thumbnail. A pet icon is an icon that visually identifies the pet, such as a graphic of a specific color or shape. The pet name is the name set by the target user for the pet. A pet thumbnail is a photo or characteristic image of the pet, such as a realistic photograph or an abstract cartoon representation.

[0171] The target user selects a target pet through interactive operations on the terminal interface. Clicking on a pet in the pet list generates a search command for that target pet. The target user can also select multiple pets, such as clicking on both "Xiaobai" and "Wangcai" simultaneously to generate a command to search for multiple target pets.

[0172] The method disclosed herein allows target users to intuitively select the pet they need to find by using the target pet selection command, and achieves accurate and efficient pet search function by generating search commands and matching them with the heat information bound to the target pet.

[0173] In embodiments of this disclosure, controlling the movement of the cleaning robot based on the target pet's activity thermal information to locate the target pet includes:

[0174] Based on the aforementioned heat map information of the target pet's activity and preset thresholds, heat map locations are determined; and

[0175] If there is only one heat source location, control the cleaning robot to move to the heat source location in order to find the target pet.

[0176] In some embodiments, heatmaps are regions or locations extracted based on the target pet's activity heatmap information and a preset threshold, representing areas where the target pet is highly likely to be active. The preset threshold is used to filter high-probability activity areas. For example, when the heatmap information shows that the activity probability of a certain area is greater than a certain threshold (e.g., 80%), that area is marked as a heatmap. The activity probability distribution of different areas is extracted from the target pet's activity heatmap. The activity probability value of each area is compared with the preset threshold. If the probability value is greater than the threshold, that area is marked as a heatmap.

[0177] When there is only one heat map location (e.g., the target pet's main activity area is clearly concentrated in one place), the system directly controls the cleaning robot to move to that heat map location. Within the heat map location area, the cleaning robot uses a local search mode (such as spiral search or area coverage). Multimodal perception modules, including vision, sound, and thermal sensing, are used to further locate the pet's position. This method is simple and efficient, avoiding unnecessary path planning and improving search efficiency.

[0178] In embodiments of this disclosure, controlling the movement of the cleaning robot based on the target pet's activity thermal information to locate the target pet includes:

[0179] Based on the target pet activity heat map information and preset thresholds, determine the heat map locations;

[0180] In some embodiments, when at least two heatmap locations exist (the target pet may be active in multiple high-probability areas), a second planned path is generated based on all heatmap locations. Algorithms such as A*, Dijkstra's algorithm, and the Traveling Salesman Problem Algorithm (TSP) are used to determine the optimal path connecting all heatmap locations. In this approach, several high-probability locations are selected based on the target pet's activity heatmap information, generating a search path. The search path can be optimized according to the proximity of the locations to save time. Heatmap information can be updated in real time during movement; if the probability of a location changes or the target pet is detected, the path will be dynamically adjusted.

[0181] By employing algorithms such as A*, Dijkstra's algorithm, or the Traveling Salesman Problem (TSP), the system can plan the optimal path between all heatmap locations, ensuring that the cleaning robot visits high-probability areas sequentially via the shortest path, avoiding redundant movements or inefficient searches. Through a secondary planned path generated from multiple heatmap locations, the cleaning robot can intelligently adapt to different pet activity patterns, optimize path selection, and dynamically adjust search strategies, ultimately improving the efficiency and accuracy of finding the target pet and providing users with a more intelligent and convenient pet-finding solution.

[0182] In the embodiments of this disclosure, the pet search instruction includes the pet's location information guessed by the target user.

[0183] The above methods also include:

[0184] A third planned path is generated based on all the aforementioned heatmap locations and the pet location information guessed by the target user; and

[0185] The cleaning robot is controlled to move based on the third planned path described above in order to find the target pet.

[0186] In some embodiments, the pet location information guessed by the target user is the pet's possible location that the target user infers based on experience or observation, and this pet location information guessed by the target user can be transmitted to the system through an interactive terminal.

[0187] Heatmaps are extracted from the target pet's activity heatmap data to represent areas where the pet is likely to be active. These heatmaps are determined based on the activity probability distribution, with higher-probability areas receiving higher priority. The system integrates the target user's guessed location information with the heatmaps to generate a more accurate search strategy. For example, the target user's guessed location information can be given higher weight, prioritizing areas entered by the target user. The heatmaps serve as a reference, covering other potential areas.

[0188] Based on the target user's guessed location information and heatmap points, a third planned path is generated. For example, the cleaning robot can prioritize covering the area guessed by the target user, and then sequentially cover the heatmap points after completing the guessed area. The cleaning robot can also comprehensively consider the user's guessed location and heatmap points, generating an efficient path based on distance or using path optimization algorithms to minimize travel distance and time. If the target pet is found at a certain stage, the system immediately terminates the subsequent path and notifies the target user. If the target pet is not found in the user's guessed area, the system continues to complete the path according to the heatmap points.

[0189] This disclosure significantly improves the efficiency and accuracy of cleaning robots in locating target pets by combining the target user's guessed information with the target pet's heatmap locations to generate a third-stage planning path. It prioritizes the information input by the target user and utilizes heatmap locations to supplement coverage of other potential areas, effectively avoiding omissions. This method excels in target user interactivity, path optimization, and dynamic adaptability, making it suitable for pet locating tasks in diverse household scenarios.

[0190] In embodiments of this disclosure, the cleaning robot control method further includes:

[0191] Based on cleaning and / or pet search tasks, obtain image information and location information of the target pet; and

[0192] Based on the image information and location information of the target pet, a deep learning algorithm is used to determine the thermal information of the target pet's activity.

[0193] In some embodiments, the generation of target pet activity thermal information relies on image information and pet location information collected by the cleaning robot during the task. The data source for target pet activity thermal information can be data collected during cleaning tasks. During routine cleaning, the cleaning robot acquires images of the home environment through its built-in camera. If the camera detects the shape, color, or characteristics of the target pet (such as movement or stillness), it records its location. Alternatively, the data source can be data collected during pet search tasks. After the user issues a pet search command, the cleaning robot activates search mode, collecting data on the target pet through multimodal sensors. The location of the target pet is recorded and uploaded each time it is detected for subsequent analysis of activity thermal information.

[0194] Deep learning models use collected data as input to analyze the activity areas of target pets. Algorithms such as the "You Only Look Once" (YOLO) algorithm and Faster Region-based Convolutional Neural Network (Faster R-CNN) are used to detect the presence of the target pet in images and accurately label its location. Classification networks (such as Residual Neural Networks (ResNet)) are used to identify the pet's species or specific identity. Recurrent Neural Network (RNN) or Transformer models are employed, combining time-series and spatial location data, to predict the target pet's activity trends and regional distribution. Regression models (such as UNet) are used to generate heatmaps of activity probability distributions, labeling the probability of pet activity in different areas of the home map.

[0195] The deep learning model continuously accumulates activity data of the target pet and generates a basic heatmap based on historical behavior. During cleaning or search tasks, the cleaning robot collects new data in real time and dynamically adjusts the heatmap. The model generates personalized heatmaps based on the specific pet's behavioral habits (such as preferred hiding places), improving search efficiency.

[0196] Figure 3 shows a structural schematic diagram of another possible robotic vacuum cleaner according to some embodiments of the present disclosure. The AI ​​processor includes a machine learning model and a communication module. An external AI processor can provide more powerful computing capabilities to handle complex deep learning tasks. The machine learning model can perform large-scale image processing, spatiotemporal behavior prediction, and dynamic path optimization, etc. The communication module is used to communicate with the robotic vacuum cleaner's host in real time, receive data from the host, and return analysis results. For example, the host sends image data to the AI ​​processor for deep analysis, and the AI ​​processor returns the detection results.

[0197] This disclosure utilizes deep learning algorithms to transform image and location information collected by cleaning robots during cleaning or pet search tasks into thermal imagery of the target pet's activity. This method combines historical data and real-time information to generate efficient and dynamic heatmaps, significantly improving the accuracy of pet activity area prediction and search efficiency, providing users with an intelligent and efficient pet search solution.

[0198] In embodiments of this disclosure, the aforementioned target pet activity thermal information also includes time information.

[0199] The above-mentioned response to the pet search command issued by the target user, obtaining the target pet's activity heatmap information, includes:

[0200] In response to a pet search command issued by the target user, determine the current time; and

[0201] Based on the current time, obtain the heat map information of the target pet's activity that matches the current time.

[0202] In some embodiments, the target pet activity heatmap includes a time dimension, reflecting the pet's activity habits and frequency at different times. For example, the time period can be divided into morning, daytime, and evening. The heatmap information for each area changes dynamically over time. For example, the kitchen might be a high-probability area in the morning but a low-probability area at night. The time dimension information can be derived from long-term accumulation of cleaning tasks and pet activity data through analysis using a deep learning model.

[0203] The target user issues a search command via a smart terminal (such as a mobile app or voice assistant). Upon receiving the command, the system automatically obtains the current time and uses it as a parameter to match with time-related information. The system then filters the target pet's heatmap information based on the current time, extracting heatmap data relevant to the current time period from a heatmap database. Based on the time-filtered heatmap information, an optimized path is generated to control the cleaning robot to locate the target pet.

[0204] This disclosure combines time information with heatmap data of the target pet's activity to dynamically match the pet's activity area within the current time period based on the user's search instructions, generating more accurate heatmap data and optimizing the cleaning robot's search path. This method significantly improves search efficiency and accuracy by utilizing the time dimension, providing users with an efficient and intelligent pet search solution.

[0205] In the embodiments of this disclosure, Figure 4 is a schematic diagram of a cleaning robot working scenario according to some embodiments of this disclosure. The client's interactive interface provides intuitive visualization functions, allowing the target user to view information such as a home map, the cleaning robot's location, and a pet heatmap. The target user issues commands via a mobile app, such as "find the pet" or "clean the living room." The system provides real-time feedback on the task status, notifying the target user of the search progress or completion status.

[0206] When performing a cleaning task, the target user selects a cleaning mode (such as global cleaning or targeted area cleaning) in the mobile app. The app uploads the instructions to the cloud, which generates the cleaning task and sends it to the cleaning robot. The cleaning robot completes the cleaning according to the planned path and uploads the cleaning log to the cloud. The target user can view the cleaning progress in real time through the app and receive a notification upon completion of the task.

[0207] When performing a pet search task, the target user selects the "Find Pet" function in the app and specifies the target pet (selected by pet name or thumbnail). The system generates a task and sends the search instruction to the cleaning robot. The cleaning robot uses a vision module or sensors to capture the pet's location in real time and prioritizes high-probability areas based on an activity heatmap. After confirming the pet's location, the cleaning robot sends the result back to the cloud, and the target user receives a notification and location information through the app.

[0208] The intelligent cleaning robot system disclosed herein encompasses a cleaning robot main unit, pet location functionality, target user terminal, and cloud services. Through multi-module collaboration, target users can control the cleaning robot in real-time via a mobile app to complete cleaning or pet location tasks and view progress and results. This system design achieves an integrated solution for home cleaning and pet monitoring, characterized by high efficiency, intelligence, and convenience.

[0209] Please refer to Figure 5, a structural schematic diagram of a cleaning robot control device 200 according to some embodiments of the present disclosure, which may include an acquisition module 201 and / or a search module 202.

[0210] The acquisition module 201 is configured to acquire target pet activity heatmap information in response to a pet search command issued by the target user; the target pet activity heatmap information is the distribution information of activity intensity in different areas based on the target pet's historical activity information.

[0211] In some embodiments, the target user can issue a "find pet" command to the cleaning robot via a smart terminal, such as a mobile app or voice assistant. The cleaning robot's acquisition module 201 receives the command and activates the pet search mode. The cleaning robot analyzes the target pet's historical activity data locally or in its corresponding cloud. The target pet's historical activity data can be collected through smart cameras installed in the home, pet wearable devices (such as positioning collars), or sensors built into the cleaning robot. Heatmap information is a distribution map obtained by statistically analyzing the frequency of pet activity in different areas. For example: high-frequency pet activity areas (such as pet beds, kitchens). Low-frequency activity areas (such as corners, hard-to-reach areas). This distribution information is usually presented in the form of a heatmap, indicating the probability of pet activity in different areas.

[0212] The search module 202 is configured to control the movement of the cleaning robot based on the target pet's activity thermal information in order to search for the target pet.

[0213] In some embodiments, the cleaning robot's search module 202 generates a path plan based on a pet activity heatmap, prioritizing navigation to areas where the pet is frequently active. During path planning, the cleaning robot combines built-in maps (such as SLAM technology) with heatmap information to avoid furniture obstacles and ensure efficient movement. If the cleaning robot fails to find the target pet in a high-probability area, the system can switch to a medium- or low-probability area to continue searching, based on a set threshold. During movement, real-time sensor data, such as camera recognition and microphone sound detection, can be used to further narrow down the target area. The cleaning robot's built-in visual module (such as a camera or thermal infrared sensor) or sound recognition module (such as detecting barks or collar signals) confirms the target pet's location. Once the target pet is found, the cleaning robot notifies the user, such as through an app push notification or voice broadcast.

[0214] In embodiments of this disclosure, the search module includes: a first path generation unit and a first control unit.

[0215] The aforementioned first path generation unit is configured to generate a first planned path based on the aforementioned target pet activity heat map information and path planning algorithm;

[0216] The first control unit is configured to control the movement of the cleaning robot based on the first planned path in order to find the target pet.

[0217] In some embodiments, the target pet activity thermal information is activity data of the target pet collected by the cleaning robot through multiple data sources. This data can be pet dynamic monitoring data provided by a home smart camera, location information transmitted by a pet wearable device (such as a positioning collar), or pet activity information recorded by the cleaning robot's own sensors (such as cameras and infrared sensors).

[0218] The first path generation unit generates a preliminary planned path based on the target pet's activity heatmap information and a path planning algorithm. The path planning algorithm can be, for example, A* algorithm or Dijkstra's algorithm. The preliminary planned path prioritizes high-probability areas and gradually covers medium- and low-probability areas. Furthermore, it dynamically optimizes the path considering factors such as the cleaning robot's battery level and obstacle avoidance requirements.

[0219] If the thermal information is updated in real time (e.g., sensors detect a new pet location), the first path generation unit can dynamically adjust the first planned path based on a path planning algorithm. The first planned path can be adjusted based on sensor information to avoid dynamic obstacles such as furniture. When a pet is detected moving in a specific area, the path is directly replanned to reach that area.

[0220] The first control unit directs the cleaning robot to move sequentially according to the generated first planned path, and the navigation module ensures the accuracy of path execution. In high-probability areas, the cleaning robot adopts a small-range search mode (such as a spiral path) to ensure that no key points are missed.

[0221] During movement, the cleaning robot can perceive the pet's location in real time through, but not limited to, the following methods: recognizing the pet's shape, color, and other characteristics through a visual module; capturing the pet's barks or collar signals through an audio module; and detecting the pet's body temperature through a thermal sensor module.

[0222] If the target pet is not found in a high-probability area, the cleaning robot will automatically adjust its path, reorder the remaining low-to-medium probability areas, and continue the search. If the real-time sensors detect the target pet, the system will directly generate a new path and head towards the target location.

[0223] After confirming the target using vision, sound, or infrared modules, the cleaning robot sends a notification to the user's device. The pet's location is marked in the app, and real-time video or voice announcements are provided. The user is given feedback on the covered area and suggestions on whether to expand the search area or regenerate the route.

[0224] In the embodiments of this disclosure, the search module includes: an information display unit, an area determination unit, and a second control unit.

[0225] The aforementioned information display unit is configured to display the heat map information of the target pet's activity on the target user's control terminal;

[0226] The aforementioned area determination unit is configured to acquire the search area determined by the target user based on the aforementioned target pet activity heat map information;

[0227] The second control unit is configured to control the cleaning robot to move within the search area in order to find the target pet.

[0228] In some embodiments, after the cleaning robot acquires the heat map information of the target pet's activity, the information display unit can display it as a visual heat map on the user control terminal (such as a mobile app or tablet). The heat map information of the target pet's activity can be displayed in the form of a heat map, using color gradients to represent the probability of pet activity in different areas. For example, red represents areas with high activity frequency, and yellow or green represents areas with medium or low activity frequency.

[0229] On the control terminal, the target user can independently select or adjust the search area based on the target pet's activity heatmap information and user needs. The area selection unit can acquire the area selected by the target user. The area selection method can be determined manually or through other quick options. Manual selection allows the target user to directly circle a specific area on the heatmap. Quick options allow the user to quickly confirm or adjust the system-recommended high-probability areas. In addition, the target user can select multiple areas to search, which is especially suitable for multiple locations where the pet may frequently be active, such as the pet's bed and balcony.

[0230] The second control unit generates a precise movement path based on the user-specified search area and a path planning algorithm. Within the designated search area, the cleaning robot navigates to high-activity areas according to a probability distribution. Utilizing SLAM technology and sensors, it dynamically avoids obstacles such as furniture and walls, ensuring efficient movement. If the user selects multiple search areas, the cleaning robot automatically switches to the next target area after completing the current one. During the search, the cleaning robot uses multimodal perception modules (such as cameras, microphones, and infrared sensors) to detect the presence of the target pet in real time. Visual recognition involves capturing the pet's morphological features in real time using a camera. Sound localization uses a microphone to detect the pet's barks or signals emitted from its collar, further narrowing the search area. Thermal sensing detects the pet's body temperature to help identify where the pet is hiding (such as under a sofa or bed). If the cleaning robot does not find the target pet in the designated area, the system dynamically adjusts its path based on activity thermal information, prioritizing the inspection of other potential areas.

[0231] When the cleaning robot successfully locates the target pet, it notifies the user of the target location via the user terminal (such as an app push or voice assistant), or marks the pet's location on the terminal map and prompts "Pet found".

[0232] If the target pet is not found, the cleaning robot will provide feedback to the user on the search progress (such as covered areas and remaining uncovered areas). It will suggest whether to expand the search area or enable the whole-house search mode.

[0233] In embodiments of this disclosure, the cleaning robot control device further includes: a marker acquisition module and a manual addition control module.

[0234] The aforementioned marker acquisition module is configured to acquire manually added markers from the target user.

[0235] The aforementioned manually added control module is configured to control the cleaning robot based on the manually added markers to locate the target pet.

[0236] In some embodiments, the target user can manually add target markers by observing pet activity heatmaps using a control terminal (such as a mobile app, tablet, or voice assistant). The marker acquisition module can acquire this input information. Markers are used to specify areas where the pet may be active or hiding, such as a room or near furniture. Markers supplement the heatmaps generated by the system, covering areas the user deems likely to be target areas.

[0237] Marker points can be set using one or more combinations of methods, including direct selection, voice input, and text description. Direct selection involves the user clicking on the target location on the home map interface to generate a marker. Voice input allows the user to specify the location via voice commands, such as "Looking for my pet, it might be on the balcony." Text description allows the user to input a specific location description, and the system automatically locates the corresponding map position.

[0238] Users can set the search priority of markers and allow users to add multiple markers at the same time. Users can add, delete, or adjust the position and priority of markers at any time.

[0239] The manually added control module combines user-marked points with pet activity heatmap information, taking into account the priority of both to generate a path. If a marker has a higher priority than a heatmap area, the cleaning robot will prioritize visiting the marker. Path planning dynamically balances the coverage order between markers and heatmap areas to ensure maximum efficiency.

[0240] In the embodiments of this disclosure, the acquisition module includes: a first search unit and a first matching unit.

[0241] The first search unit mentioned above is configured to determine the target pet based on the pet search command issued by the target user.

[0242] The first matching unit is configured to match the target pet activity heat information corresponding to the target pet in the heat information database.

[0243] In some embodiments, the target user issues a "Find Pet" command via a smart terminal (such as a mobile app or voice assistant). The first search unit can determine the target pet based on the pet search command. For example, there may be multiple target pets in the target user's home, and the heat map database stores heat map information on the activity of multiple target pets. The target user can select one or more target pets to search for. The target user can select a target pet by choosing from a list, by using a voice command, or by inputting an image or feature information on the smart terminal. By choosing from a list, the target user can select the target pet to search for on the interface of the smart terminal. By using a voice command, the target user can directly say the name of the target pet, such as "Find Xiaobai". By inputting an image or feature, the target user can upload a photo of the target pet or select a bound pet feature (such as color, size, etc.).

[0244] The system determines the target pet based on the user's input. If the user owns multiple pets, the system selects and locks the target pet. The system also supports searching for multiple pets at once. The system verifies whether the target pet is already linked to the system (e.g., identified by a collar, tag, etc.). If not linked, the system prompts the user to register or select another pet.

[0245] The heatmap database stores activity heatmap information for multiple target pets. Each pet's activity heatmap is generated based on its historical activity data, and includes, but is not limited to, the following: spatial distribution information, temporal dimension information, and specific behavioral habits. Spatial distribution information includes the frequency of the pet's activity in different areas of the home. Temporal dimension information includes the pet's activity patterns at different times of day. Specific behavioral habits include preferred hiding places, etc.

[0246] Pet activity heatmaps are collected via devices worn by pets, home cameras, or sensors from cleaning robots. The database is categorized by pet identity for easy and quick searching. The first matching unit matches the target pet's activity heatmap against the corresponding information in the heatmap database. The first matching unit searches the database based on the pet identifier entered by the target user. It returns an activity heatmap of the target pet, containing spatial distribution information of activity probability. If multiple target pets are being searched, the system will match multiple heatmaps simultaneously and display the activity distribution for each pet separately. The system also supports dynamic updates to the heatmaps, continuously supplementing them with the latest pet activity data to ensure the heatmaps reflect current activity trends.

[0247] In embodiments of this disclosure, the pet search instruction includes an image matching instruction.

[0248] The first search unit is configured to obtain pet image information corresponding to the image matching instruction.

[0249] The first search unit is also configured to determine the target pet based on the pet image information and the image recognition algorithm.

[0250] In some embodiments, the target user can upload pet photos of the target pet via a smart terminal (such as a mobile app or tablet). The first search unit can retrieve this pet photo information. The pet photo information can be a close-up of the pet, including facial features, body shape, and other information. The system supports multiple image upload formats, such as real-time shooting, album selection, or importing from social media.

[0251] The first search unit uses image recognition algorithms (such as deep learning models: ResNet, YOLO, etc.) to analyze the pet's appearance features. Extracted features may include, but are not limited to, color features, morphological features, facial features, and feature markers. Color features include the distribution of the pet's coat color. Morphological features include the shape of the ears and tail, and the overall body shape. Facial features include the shape of the eyes and the position of the nose. Feature markers include unique markings, scars, or collars.

[0252] The image recognition algorithm compares the extracted features with pet feature templates in the heat map database. It then matches the extracted pet features against the pet information already linked in the heat map database. If a unique match is found, the first search unit confirms the target pet's identity. If multiple pets are matched, the first search unit lists possible options for the target user to confirm. If no target pet is matched, the target user is prompted to link a new pet or re-upload the image. The target user can flexibly operate via image upload commands to adapt to diverse needs.

[0253] In embodiments of this disclosure, the pet search instructions include voice search instructions.

[0254] The first search unit is configured to perform speech recognition on the voice search command to obtain semantic information;

[0255] The first search unit is also configured to determine the target pet based on the semantic information and semantic recognition algorithm.

[0256] In some embodiments, the target user issues voice commands to the system via a smart terminal. The content of the voice commands may include, but is not limited to: the pet's name, the pet's characteristics, location prompts, and action requests.

[0257] The first search unit uses a speech recognition algorithm to convert speech signals into text. During the speech recognition process, an adaptive model can be used to improve recognition accuracy for different language environments or target user accents.

[0258] The first search unit analyzes the semantic information of voice commands using a semantic recognition algorithm, extracting key content. The system then combines this extracted semantic information with the pet's already bound information, searching the heatmap database for records matching the pet's name or characteristics. If a unique pet is matched, the target pet is confirmed. If multiple possible matches exist, the system requests more information from the target user to further confirm the target. If the target user is searching for multiple pets simultaneously, the system processes the semantic information separately, generating corresponding heatmaps and search plans. The system also supports combined commands, such as "Looking for the white cat and the orange cat; the orange cat might be in the living room."

[0259] By combining speech recognition and semantic analysis with thermal information and path planning technology, the cleaning robot can quickly respond to voice search commands. Its efficient semantic extraction, dynamic path adjustment, and real-time notification functions not only improve the accuracy and efficiency of pet retrieval but also significantly enhance the convenience and flexibility of the user experience.

[0260] In embodiments of this disclosure, the pet search instruction includes a target pet selection instruction.

[0261] The first search unit is configured to generate a pet search instruction based on the pet display information selected by the target user on the target user's control terminal; the pet display information includes at least one of a pet icon, a pet name, and a pet thumbnail.

[0262] The first search unit is also configured to determine the target pet based on the pet search instruction.

[0263] In some embodiments, the target user views pet information bound to the system via a control terminal. This information includes one or a combination of pet icons, pet names, and pet thumbnails. For example, the user can view a pet icon and pet name, a pet name and pet thumbnail, a pet icon and pet thumbnail, or a pet icon, pet name, and pet thumbnail. A pet icon is an icon that visually identifies the pet, such as a graphic of a specific color or shape. The pet name is the name set by the target user for the pet. A pet thumbnail is a photo or characteristic image of the pet, such as a realistic photograph or an abstract cartoon representation.

[0264] The target user selects a target pet through interactive operations on the terminal interface. Clicking on a pet in the pet list generates a search command for that target pet. The target user can also select multiple pets, such as clicking on both "Xiaobai" and "Wangcai" simultaneously to generate a command to search for multiple target pets.

[0265] In the embodiments of this disclosure, the above-mentioned search module includes: a first thermal point determination unit and a third control unit.

[0266] The aforementioned first thermal point determination unit is configured to determine thermal points based on the aforementioned target pet activity thermal information and a preset threshold.

[0267] When there is only one heat source location, the third control unit is configured to control the cleaning robot to move to the heat source location in order to find the target pet.

[0268] In some embodiments, heatmap locations are regions or locations extracted by a first heatmap location determination unit based on the target pet's activity heatmap information and a preset threshold, representing areas where the target pet has a high probability of activity. The preset threshold is used to filter high-probability activity areas. For example, when the heatmap information shows that the activity probability of a certain area is greater than a certain threshold (e.g., 80%), that area is marked as a heatmap location. The activity probability distribution of different areas is extracted from the target pet's activity heatmap. The activity probability value of each area is compared with the preset threshold. If the probability value is greater than the threshold, that area is marked as a heatmap location.

[0269] When there is only one heat map location (e.g., the target pet's main activity area is clearly concentrated in one place), the third control unit is configured to control the cleaning robot to move to the heat map location to find the target pet. The cleaning robot is directly controlled to move to the heat map location. Within the heat map location area, the cleaning robot uses a local search mode (such as spiral search or area coverage). Multimodal perception modules, including vision, sound, and thermal sensing, are used to further locate the pet's position. This method is simple and efficient, avoiding unnecessary path planning and improving search efficiency.

[0270] In the embodiments of this disclosure, the search module includes: a second thermal point determination unit, a second path generation unit, and a fourth control unit.

[0271] The aforementioned second thermal location determination unit is configured to determine thermal locations based on the aforementioned target pet activity thermal information and a preset threshold.

[0272] When there are at least two of the aforementioned heat points, the second generation unit is configured to generate a second planned path based on all of the aforementioned heat points.

[0273] The aforementioned fourth control unit is configured to control the movement of the cleaning robot based on the aforementioned second planned path in order to find the aforementioned target pet.

[0274] In some embodiments, when at least two heatmap locations exist (the target pet may be active in multiple high-probability areas), a second generation unit generates a second planned path based on all heatmap locations. Algorithms such as A*, Dijkstra's algorithm, and the Traveling Salesman Problem (TSP) are used to determine the optimal path connecting all heatmap locations. In this approach, several high-probability locations are selected based on the target pet's activity heatmap information, and a search path is generated. The search path can be optimized according to the proximity of the locations to save time. Heatmap information is updated in real time during movement; if the probability of a location changes or the target pet is detected, the path is dynamically adjusted.

[0275] The fourth control unit uses the A* algorithm, Dijkstra's algorithm, or the TSP (Traveling Salesman Problem) algorithm to plan the optimal travel path among all heatmap locations, ensuring that the cleaning robot visits high-probability areas sequentially via the shortest path, avoiding redundant movements or inefficient searches. Through a second planned path generated from multiple heatmap locations, the cleaning robot can intelligently adapt to different pet activity patterns, optimize path selection, and dynamically adjust search strategies, ultimately improving the efficiency and accuracy of finding the target pet and providing users with a more intelligent and convenient pet-finding solution.

[0276] In the embodiments of this disclosure, the pet search instruction includes the pet's location information guessed by the target user.

[0277] The aforementioned device also includes: a third path generation module and a third path control module.

[0278] The aforementioned third path generation module is configured to generate a third planned path based on all the aforementioned heatmap locations and the pet location information guessed by the target user.

[0279] The aforementioned third path control module is configured to control the movement of the cleaning robot based on the aforementioned third planned path in order to find the aforementioned target pet.

[0280] In some embodiments, the pet location information guessed by the target user is the pet's possible location that the target user infers based on experience or observation, and this pet location information guessed by the target user can be transmitted to the system through an interactive terminal.

[0281] Heatmaps are extracted from the target pet's activity heatmap data to represent areas where the pet is likely to be active. These heatmaps are determined based on the activity probability distribution, with higher-probability areas receiving higher priority. The system integrates the target user's guessed location information with the heatmaps to generate a more accurate search strategy. For example, the target user's guessed location information can be given higher weight, prioritizing areas entered by the target user. The heatmaps serve as a reference, covering other potential areas.

[0282] The third path generation module generates a third planned path based on the target user's guessed location information and heatmap points. This module can control the cleaning robot to prioritize covering the area guessed by the target user, and then sequentially cover the heatmap points after completing the guessed area. The third path generation module can also comprehensively consider the user's guessed location and heatmap points, generating an efficient path based on distance or using path optimization algorithms to minimize travel distance and time. If the target pet is detected at a certain stage, the third path control module immediately terminates the subsequent path and notifies the target user. If the target pet is not found in the user's guessed area, the fifth control unit continues to complete the path according to the heatmap points.

[0283] In embodiments of this disclosure, the cleaning robot control device further includes:

[0284] The information acquisition module is configured to acquire image information and location information of the target pet based on cleaning tasks and / or pet search tasks; and

[0285] The thermal information determination module is configured to determine the thermal information of the target pet's activity based on the image information and location information of the target pet using a deep learning algorithm.

[0286] In some embodiments, the generation of target pet activity thermal information relies on image information and pet location information collected by the cleaning robot during the task. The information acquisition module can collect data during the cleaning task; the cleaning robot acquires images of the home environment through its built-in camera during daily cleaning. If the camera detects the target pet's shape, color, or characteristics (such as movement or stillness), its location is recorded. The data source for target pet activity thermal information can also be data collected during pet search tasks. After the target user issues a pet search command, the cleaning robot activates search mode, collecting data on the target pet through multimodal sensors. The location of the target pet detected each time is recorded and uploaded for subsequent analysis of activity thermal information.

[0287] Deep learning models use collected data as input to analyze the activity areas of target pets. The heatmap detection module uses algorithms such as YOLO and Faster R-CNN to detect the presence of the target pet in images and accurately label its location. Classification networks (such as ResNet) identify the pet's species or specific identity. Using RNN (Recurrent Neural Network) or Transformer models, combined with time-series and spatial location data, the activity trends and regional distribution of the target pet are predicted. Regression models (such as UNet) generate heatmaps of activity probability distributions, labeling the probability of pet activity in different areas of the home map.

[0288] The deep learning model continuously accumulates activity data of the target pet and generates a basic heatmap based on historical behavior. During cleaning or search tasks, the cleaning robot collects new data in real time and dynamically adjusts the heatmap. The model generates personalized heatmaps based on the specific pet's behavioral habits (such as preferred hiding places), improving search efficiency.

[0289] In embodiments of this disclosure, the aforementioned target pet activity thermal information further includes time information.

[0290] The aforementioned acquisition module includes: a time determination unit and a second matching unit.

[0291] The aforementioned time determination unit is configured to determine the current time in response to a pet search command issued by the target user;

[0292] The second matching unit is configured to acquire heatmap information of target pet activity that matches the current time based on the current time.

[0293] In some embodiments, the target pet activity heatmap includes a time dimension, reflecting the pet's activity habits and frequency at different times. For example, the time period can be divided into morning, daytime, and evening. The heatmap information for each area changes dynamically over time. For example, the kitchen might be a high-probability area in the morning but a low-probability area at night. The time dimension information can be derived from long-term accumulation of cleaning tasks and pet activity data through analysis using a deep learning model.

[0294] The target user issues a search command via a smart terminal (such as a mobile app or voice assistant). Upon receiving the command, the system's time determination unit obtains the current time, and the second matching unit uses the current time as a parameter to match it with time-dimensional information. The second matching unit then filters the target pet's heatmap information based on the current time, extracting heatmap data related to the current time period from the heatmap information database. Based on the time-filtered heatmap information, an optimized path is generated to control the cleaning robot to locate the target pet.

[0295] As shown in Figure 6, this disclosure also provides an electronic device 300, including a memory 310, a processor 320, and a computer program 311 stored in the memory 310 and executable on the processor. When the processor 320 executes the computer program 311, it implements the steps of any of the above-described methods for controlling the cleaning robot.

[0296] Since the electronic device described in this embodiment is a device used to implement a cleaning device in this disclosure, those skilled in the art can understand the specific implementation method and various variations of the electronic device in this embodiment based on the method described in this disclosure. Therefore, how the electronic device implements the method in this disclosure will not be described in detail here. Any device used by those skilled in the art to implement the method in this disclosure is within the scope of protection of this disclosure.

[0297] In some implementations, when the computer program 311 is executed by a processor, it can implement any of the implementations in the embodiments corresponding to the first aspect.

[0298] This disclosure also proposes a computer-readable storage medium including a computer program stored thereon, which, when executed by a processor, implements the cleaning robot control method of any of the first aspects. Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk drive (HDD), or solid-state drive (SSD), etc., and the storage medium may also include combinations of the above types of memory.

[0299] As shown in Figure 7, this disclosure also provides a cleaning robot 400, including an electronic device 300 as shown in Figure 4.

[0300] It should be noted that the descriptions of each embodiment in the above embodiments have different focuses. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0301] Those skilled in the art will understand that embodiments of this disclosure can be provided as methods, systems, or computer program products. Therefore, this disclosure can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this disclosure can take the form of a computer program product embodied on one or more computer-readable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-readable program code.

[0302] This disclosure is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to this disclosure. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in one or more blocks of the flowchart illustrations and / or one or more blocks of the block diagrams.

[0303] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means that implement the functions specified in one or more flowcharts and / or one or more block diagrams.

[0304] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, such that the instructions, which execute on the computer or other programmable apparatus, provide steps for implementing the functions specified in one or more flowcharts and / or one or more block diagrams.

[0305] This disclosure also provides a computer program product including computer software instructions that, when executed on a processing device, cause the processing device to perform a process of a cleaning robot control method.

[0306] A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, they produce, in whole or in part, the processes or functions according to this disclosure. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transferred from one computer-readable storage medium to another. For example, computer instructions can be transferred from one website, computer, server, or data center to another via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can store or a data storage device such as a server or data center that integrates one or more available media. Available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media (e.g., solid-state disks (SSDs)).

[0307] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0308] In the several embodiments provided in this disclosure, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0309] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0310] Furthermore, the functional units in the various embodiments of this disclosure can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0311] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this disclosure, in essence, or the part that contributes to related technologies, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this disclosure. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0312] In summary, this disclosure proposes a cleaning robot control method that utilizes multiple data sources to generate activity heatmap information of the target pet, comprehensively considering historical behavior data and real-time environmental information, significantly improving the accuracy of pet identification and localization. By generating a heatmap through statistical analysis of the target pet's historical activity data, high-frequency activity areas are prioritized, improving the target-specificity of identification. Furthermore, the heatmap information is presented in the form of a probability distribution, effectively avoiding the problem of missing pets active in low-frequency areas. This disclosure combines activity heatmap information with path planning algorithms to prioritize navigation to high-frequency pet activity areas, avoiding the inefficiency of traditional full-coverage search methods. The cleaning robot control method of this disclosure effectively overcomes the problems of insufficient pet identification accuracy and low efficiency in related technologies through heatmap analysis and path planning optimization, significantly improving the practicality and intelligence level of pet-finding functions.

[0313] The cleaning robot control method disclosed herein, along with other advantages, objectives, and features of this disclosure, will be partly apparent from the following description and partly understood by those skilled in the art through study and practice of this disclosure.

Claims

1. A method for controlling a cleaning robot, comprising: In response to a pet search command issued by a target user, the system acquires heatmap information on the target pet's activity, wherein the heatmap information on the target pet's activity is distribution information on the frequency of its activity in different areas, obtained based on the target pet's historical activity information; and The cleaning robot is controlled to move based on the target pet's activity thermal information in order to locate the target pet.

2. The cleaning robot control method according to claim 1, wherein, The step of controlling the cleaning robot to move based on the target pet's activity thermal information to locate the target pet includes: A first planned path is generated based on the target pet's activity heatmap information and a path planning algorithm; and The cleaning robot is controlled to move based on the first planned path in order to find the target pet.

3. The cleaning robot control method according to claim 1 or 2, wherein, The step of controlling the cleaning robot to move based on the target pet's activity thermal information to locate the target pet includes: The thermal information of the target pet's activity is displayed on the target user's control terminal; Obtain the search area determined by the target user based on the target pet's activity heatmap; and The cleaning robot is controlled to move within the search area to locate the target pet.

4. The cleaning robot control method according to any one of claims 1-3, further comprising: Obtain manually added markers from the target user; as well as The cleaning robot is controlled by manually adding markers to locate the target pet.

5. The cleaning robot control method according to any one of claims 1-4, wherein, The process of responding to a pet search command issued by the target user and obtaining heatmap information on the target pet's activity includes: The target pet is determined based on the pet search command issued by the target user; and Match the target pet's activity heat information to the target pet in the heat information database.

6. The cleaning robot control method according to any one of claims 1-5, wherein, The heatmap database stores heat map information on the activities of multiple target pets.

7. The cleaning robot control method according to any one of claims 1-6, wherein, The pet search command includes an image matching command. The step of determining the target pet based on the pet search command issued by the target user includes: Obtain the pet image information corresponding to the image matching command; and The target pet is determined based on the pet image information and the image recognition algorithm.

8. The cleaning robot control method according to any one of claims 1-1, wherein, The pet search instructions include voice search instructions. The step of determining the target pet based on the pet search command issued by the target user includes: The voice search command is subjected to speech recognition to obtain semantic information; and The target pet is determined based on the semantic information and the semantic recognition algorithm.

9. The cleaning robot control method according to any one of claims 1-8, wherein, The pet search command includes a target pet selection command. The step of determining the target pet based on the pet search command issued by the target user includes: Based on the pet display information selected by the target user on the target user's control terminal, a pet search instruction is generated, wherein the pet display information includes at least one of a pet icon, a pet name, and a pet thumbnail; and The target pet is determined based on the pet search instructions.

10. The cleaning robot control method according to any one of claims 1-9, wherein, The step of controlling the cleaning robot to move based on the target pet's activity thermal information to locate the target pet includes: Based on the target pet's activity thermal information and preset thresholds, determine the thermal points; and If there is only one heat source, the cleaning robot is controlled to move to the heat source in order to find the target pet.

11. The cleaning robot control method according to any one of claims 1-10, wherein, The step of controlling the cleaning robot to move based on the target pet's activity thermal information to locate the target pet includes: Based on the target pet's activity thermal information and preset thresholds, determine the thermal points; When there are at least two thermal points, a second planned path is generated based on all of the thermal points; and The cleaning robot is controlled to move based on the second planned path in order to find the target pet.

12. The cleaning robot control method according to claim 10 or 11, wherein, The pet search command includes the target user's guess of the pet's location. The method further includes: A third planned path is generated based on all the heatmap locations and the target user's guessed pet location information; and The cleaning robot moves according to the third planned path in order to find the target pet.

13. The cleaning robot control method according to any one of claims 1-12, further comprising: Based on the cleaning task and / or pet search task, obtain the image information and pet location information of the target pet; as well as Based on the image information and location information of the target pet, a deep learning algorithm is used to determine the thermal information of the target pet's activity.

14. The cleaning robot control method according to any one of claims 1-13, wherein, The target pet activity heat map information also includes time information. The process of responding to a pet search command issued by the target user and obtaining heatmap information on the target pet's activity includes: In response to a pet search command issued by the target user, determine the current time; and Based on the current time, obtain the heat map information of the target pet activity that matches the current time.

15. A control device for a cleaning robot, comprising: The acquisition module is configured to acquire target pet activity heatmap information in response to a pet search command issued by a target user. This heatmap information is based on the distribution of the target pet's activity intensity in different areas, obtained from its historical activity data. The search module is configured to control the movement of the cleaning robot based on the target pet's activity thermal information in order to locate the target pet.

16. The cleaning robot control device according to claim 15, wherein, The search module includes: a first path generation unit and a first control unit. The first path generation unit is configured to generate a first planned path based on the target pet activity heat map information and the path planning algorithm; The first control unit is configured to control the movement of the cleaning robot based on the first planned path in order to find the target pet.

17. The cleaning robot control device according to claim 15 or 16, wherein, The search module includes: an information display unit, an area determination unit, and a second control unit. The information display unit is configured to display the target pet's activity heat map information on the target user's control terminal; The area determination unit is configured to acquire the search area determined by the target user based on the target pet's activity heat map information; The second control unit is configured to control the cleaning robot to move within the search area to find the target pet.

18. The cleaning robot control device according to any one of claims 15-17, further comprising: The marker acquisition module and the manual addition control module, The marker acquisition module is configured to acquire manually added markers by the target user; The manual addition control module is configured to control the cleaning robot to find the target pet based on the manually added marker points.

19. The cleaning robot control device according to any one of claims 15-18, wherein, The acquisition module includes: a first search unit and a first matching unit. The first search unit is configured to determine the target pet based on the pet search command issued by the target user; The first matching unit is configured to match the target pet's activity thermal information in the thermal information database.

20. The cleaning robot control device according to any one of claims 15-19, wherein, The heatmap database stores heat map information on the activities of multiple target pets.

21. The cleaning robot control device according to claim 19, wherein, The pet search command includes an image matching command. The first search unit is configured to obtain pet image information corresponding to the image matching instruction; The first search unit is also configured to determine the target pet based on the pet image information and the image recognition algorithm.

22. The cleaning robot control device according to claim 19, wherein, The pet search instructions include voice search instructions. The first search unit is configured to perform speech recognition on the voice search command to obtain semantic information; The first search unit is also configured to determine the target pet based on the semantic information and the semantic recognition algorithm.

23. The cleaning robot control device according to claim 19, wherein, The pet search command includes a target pet selection command. The first search unit is configured to generate a pet search instruction based on the pet display information selected by the target user on the target user's control terminal, wherein the pet display information includes at least one of a pet icon, a pet name, and a pet thumbnail; The first search unit is also configured to determine the target pet according to the pet search instruction.

24. The cleaning robot control device according to any one of claims 15-23, wherein, The locating module includes: a first thermal point determination unit and a third control unit. The first thermal point determination unit is configured to determine thermal points based on the thermal information of the target pet's activity and a preset threshold. When there is only one heat source, the third control unit is configured to control the cleaning robot to move to the heat source in order to find the target pet.

25. The cleaning robot control device according to any one of claims 15-24, wherein, The search module includes: a second thermal point determination unit, a second path generation unit, and a fourth control unit. The second thermal point determination unit is configured to determine thermal points based on the thermal information of the target pet's activity and a preset threshold. When there are at least two thermal points, the second generation unit is configured to generate a second planned path based on all of the thermal points; The fourth control unit is configured to control the movement of the cleaning robot based on the second planned path in order to find the target pet.

26. The cleaning robot control device according to claim 24 or 25, wherein, The pet search command includes the target user's guess of the pet's location. The device further includes: a third path generation module and a third path control module. The third path generation module is configured to generate a third planned path based on all the heat map locations and the pet location information guessed by the target user. The third path control module is configured to control the movement of the cleaning robot based on the third planned path in order to find the target pet.

27. The cleaning robot control device according to any one of claims 15-26, further comprising: The information acquisition module is configured to acquire image information and pet location information of the target pet based on cleaning tasks and / or pet search tasks; as well as The thermal information determination module is configured to determine the thermal information of the target pet's activity based on the target pet's image information and the pet's location information using a deep learning algorithm.

28. The cleaning robot control device according to any one of claims 15-27, wherein, The target pet activity heat map information also includes time information. The acquisition module includes: a time determination unit and a second matching unit. The time determination unit is configured to determine the current time in response to a pet search command issued by the target user; The second matching unit is configured to acquire heat information of target pet activity that matches the current time based on the current time.

29. An electronic device comprising: A memory and a processor, the processor being configured to implement the steps of the cleaning robot control method as described in any one of claims 1-14 when executing a computer program stored in the memory.

30. A computer-readable storage medium comprising a computer program stored thereon, which, when executed by a processor, implements the steps of the cleaning robot control method as claimed in any one of claims 1-14.

31. A cleaning robot, including the electronic device as claimed in claim 29.

Images