Stair climbing frame and sweeper cooperative docking method and system

Abstract

The embodiment of the invention provides a cooperative docking method for a stair climbing frame and a sweeper, and belongs to the technical field of intelligent service robots. The method is executed by a stair climbing frame, and comprises the following steps: acquiring sensing data used for describing the spatial position and attitude angle of the sweeper; determining the relative distance and the relative azimuth information with the sweeper, determining a docking motion mode according to the relative distance and the relative azimuth information, and generating a corresponding driving control instruction; docking motion is executed according to the driving control instruction, and the advancing track of the stair climbing frame is corrected in real time in the docking motion process so that the stair climbing frame can be aligned with the sweeper; after docking with the sweeper is completed, locking operation with the sweeper is executed, a docking completion instruction is sent to the sweeper, and unlocking is executed until a preset target position is reached. According to the scheme, autonomous recognition, accurate alignment and stable locking of the stair climbing frame and the sweeper in a complex environment are achieved, and high-reliability automatic cooperative butt joint and carrying connection can be completed among multiple devices.

Description

Technical Field

[0001] This invention relates to the field of intelligent service robot technology, specifically to a method for collaborative docking of a stair climber and a sweeper, a stair climber, a sweeper, a collaborative docking system of a stair climber and a sweeper, a computer-readable storage medium, an electronic device, and a computer program product. Background Technology

[0002] With the development of smart home and service robot technologies, robotic vacuum cleaners have become common household cleaning devices. However, existing robotic vacuum cleaners are generally limited to flat environments and can only complete cleaning tasks in a single-level space. When encountering structures with height differences such as stairs, thresholds, and steps, they usually cannot autonomously overcome obstacles or switch to different floors to work, which has become a key bottleneck restricting their whole-house cleaning capabilities. To address this, the industry has proposed various improvement solutions, such as increasing wheel diameter, adding climbing mechanisms, or using tracked chassis, to improve the obstacle-crossing ability of robotic vacuum cleaners. However, these solutions often lead to complex structures, increased size, and higher costs, and their mobility and energy consumption performance deteriorate significantly when operating on flat ground.

[0003] Another approach involves using a dedicated stair-climbing vehicle (also known as a stair-climbing frame) to assist the robotic vacuum cleaner in traversing floors, allowing the robot to be transported to the target floor and continue cleaning. However, this type of split design still presents significant technical challenges in practical use. On one hand, the stair-climbing vehicle and the robotic vacuum cleaner are typically two independent systems, lacking an effective sensing and communication coordination mechanism. They cannot achieve precise matching in spatial position, attitude angle, and travel path, resulting in a low docking success rate. On the other hand, existing solutions often rely on manual placement or simple infrared sensor guidance, exhibiting poor fault tolerance and insufficient environmental adaptability. Especially in home settings, factors such as ground reflection, changes in illumination, or confined spaces can easily lead to misidentification and deviation, making stable and reliable automatic docking impossible.

[0004] Therefore, how to achieve high-precision recognition, reliable alignment, and automatic connection between the stair-climbing frame and the sweeping robot without changing the existing structure of the sweeping robot has become a key technical problem that urgently needs to be solved in the field of intelligent service robots. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for collaborative docking between a stair climber and a robotic vacuum cleaner, at least to address the problems of low docking success rate and poor environmental adaptability caused by the lack of precise positional perception and communication coordination between the stair climber and the robotic vacuum cleaner in existing solutions. This invention significantly improves docking accuracy and environmental robustness by introducing bidirectional perception data fusion, dynamic motion mode switching based on relative distance and orientation (opposing motion mode and same-direction chasing mode), and real-time trajectory correction closed-loop control. Specifically, the mode switching mechanism allows the device to adaptively select the optimal path in open spaces and narrow passages, avoiding docking failures caused by traditional single motion methods.

[0006] To achieve the above objectives, the first aspect of the present invention provides a method for coordinated docking between a stair climber and a robot vacuum cleaner. The method is executed by the stair climber and includes: receiving a docking request signal from the robot vacuum cleaner and acquiring sensing data describing the spatial position and attitude angle of the robot vacuum cleaner; determining the relative distance and relative orientation information between the stair climber and the robot vacuum cleaner based on the sensing data, and determining a docking motion mode based on the relative distance and relative orientation information, and generating corresponding drive control commands; executing the docking motion according to the drive control commands, and correcting the trajectory of the stair climber in real time during the docking motion to maintain alignment with the robot vacuum cleaner; after docking with the robot vacuum cleaner, performing a locking operation between the stair climber and the robot vacuum cleaner, and sending a docking completion command to the robot vacuum cleaner until a predetermined target position is reached, and then releasing the lock.

[0007] Optionally, receiving a docking request signal from the sweeping machine and acquiring perception data to describe the spatial position and attitude angle of the sweeping machine includes: acquiring image frame data of the sweeping machine and extracting pixel features from the image frame data to describe the boundary shape of the sweeping machine; calculating the displacement vector and attitude angle change of the sweeping machine based on the feature point matching results between adjacent image frames; and generating perception data to describe the spatial position and attitude angle of the sweeping machine based on the displacement vector and attitude angle change.

[0008] Optionally, image frame data of the sweeping machine is acquired, and pixel features describing the boundary morphology of the sweeping machine are extracted from the image frame data. This includes: performing specific visual marker recognition in consecutive frames, extracting pixel positioning points based on preset QR code markers or ArUco markers on the sweeping machine; calculating the pose calculation matrix of the markers according to the distribution relationship of the pixel positioning points in the image frame coordinate system, which is used to determine the azimuth angle and center position of the sweeping machine in the environmental coordinate system; when the recognition result does not meet the feature extraction accuracy requirements, calling the depth gradient data in the image frame to perform edge feature matching to correct the boundary contour position of the sweeping machine.

[0009] Optionally, determining the relative distance and relative orientation information between the robot vacuum and the robot based on the perceived data includes: parsing the coordinate values ​​of the robot vacuum's center position in the perceived data and establishing a unified coordinate reference system in conjunction with the real-time pose parameters of the stair-climbing frame; calculating the planar distance between the robot vacuum and the robot in the horizontal direction and the height difference in the vertical direction based on the unified coordinate reference system to generate the relative distance between the robot vacuum and the robot; calculating the relative orientation angle based on the difference between the robot vacuum's attitude angle and the stair-climbing frame's orientation angle to describe the angle between the forward direction of the stair-climbing frame and the centerline of the robot vacuum; combining the relative distance and the relative orientation angle to form a relative position information data packet between the robot vacuum and the robot, and writing it into the docking motion calculation buffer.

[0010] Optionally, the docking motion mode is determined based on the relative distance and relative orientation information, and a corresponding drive control command is generated, including: reading the distance difference and angle deviation value in the relative distance and relative orientation information, and comparing them with preset thresholds respectively; wherein, the distance difference is obtained based on distance data in the relative distance used to characterize the spatial distance between the stair climber and the sweeper in a unified reference frame, and the angle deviation value is obtained based on azimuth angle data in the relative orientation information used to characterize the angle between the forward direction of the stair climber and the centerline direction of the sweeper; when the angle deviation value is less than the preset angle threshold and the distance difference is greater than the preset distance threshold, it is determined to be a reciprocating motion mode, and a first drive control command including forward direction, travel speed and correction frequency is generated; when the angle deviation value is not less than the preset angle threshold or the environmental constraint information indicates that the channel is limited (such as a narrow channel), it is determined to be a same-direction chasing mode, and a second drive control command including path following parameters, relative speed difference and lateral correction amount is generated; the first drive control command or the second drive control command is written into the control buffer area for executing the corresponding docking motion.

[0011] Optionally, the docking motion is executed according to the drive control command, and the trajectory of the stair-climbing frame is corrected in real time during the docking motion to maintain alignment with the sweeper. This includes: controlling the stair-climbing frame to move in a straight line along the docking direction based on the drive control command; monitoring the position offset and azimuth deviation of the sweeper in real time, and calculating the lateral correction and steering angle correction of the stair-climbing frame in real time based on the offset and azimuth deviation; generating a trajectory adjustment command based on the lateral correction and steering angle correction to dynamically correct the trajectory of the stair-climbing frame; and maintaining the current direction of travel and the alignment between the docking platform and the sweeper when the offset and azimuth deviation are both less than the corresponding preset thresholds in several consecutive frames of perception output.

[0012] Optionally, after docking with the robot vacuum, a locking operation is performed between the robot vacuum and the robot vacuum, and a docking completion command is sent to the robot vacuum until the predetermined target position is reached, at which point the locking is released. This includes: monitoring the level signal status output by the optocoupler sensor in the stair climber, and determining whether the level signal remains stable for a preset duration within a continuous detection cycle; when the level signal remains valid, generating a locking execution command and sending it to the locking drive component of the docking platform of the stair climber to control the docking platform to perform a mechanical locking action; after the locking action is completed, reading the feedback signal of the optocoupler sensor to confirm the locking state, and combining the locking confirmation result with the docking state identifier to generate a docking completion data packet; sending the docking completion data packet to the robot vacuum; and after the stair climber carries the robot vacuum to the predetermined target position, generating a locking release command based on the arrival determination result, and controlling the docking platform to perform the corresponding locking release action. The second aspect of this invention provides a method for the coordinated docking of a stair climber and a sweeper. The method is executed by the sweeper and includes: sending a docking request signal to the stair climber and acquiring identification data describing the spatial position and attitude angle of the stair climber; calculating the relative distance and relative orientation information between the sweeper and the stair climber based on the identification data, determining the cooperative motion mode according to the calculation results, and generating corresponding travel control commands; executing the cooperative motion according to the travel control commands, and correcting the travel trajectory in real time during the motion to keep the central axis of the sweeper aligned with the docking positioning area of ​​the stair climber; and stopping travel and retracting the cleaning components when a docking completion signal is received from the stair climber, entering the transport state.

[0013] Optionally, a docking request signal is sent to the stair climbing frame, and identification data describing the spatial position and attitude angle of the stair climbing frame is obtained, including: acquiring image frame data containing the outline of the stair climbing frame; identifying visual markers set on the outer surface of the stair climbing frame in the image frame data; calculating the attitude angle parameters of the stair climbing frame according to the distribution of the positioning points of the visual markers in the image frame coordinate system, and inferring the spatial position of the stair climbing frame in the environmental coordinate system by combining the image depth information; and combining the calculated spatial position and attitude angle parameters of the stair climbing frame to generate identification data describing the spatial position and attitude angle of the stair climbing frame.

[0014] Optionally, based on the identification data, the relative distance and relative orientation information between the sweeper and the stair climber are calculated, and the coordination movement mode is determined according to the calculation results, generating corresponding travel control commands. This includes: parsing the coordinate parameters and attitude angle parameters of the stair climber's spatial position in the identification data, and establishing a unified coordinate reference system in combination with the sweeper's own position parameters; calculating the planar distance, vertical height difference, and centerline angle between the sweeper and the stair climber under the unified coordinate reference system to form the relative distance and relative orientation information between the sweeper and the stair climber; comparing the relative distance and relative orientation information with a preset movement threshold, determining a straight-line docking coordination mode when the angle deviation value is less than the preset angle threshold, and determining a following docking coordination mode when the angle deviation value is not less than the preset angle threshold or the spatial channel is limited; and generating travel control commands containing travel direction, speed correction value, and attitude correction amount based on the determined coordination mode.

[0015] Optionally, the sweeper performs coordinated movement according to the travel control command, and corrects the travel trajectory in real time during the movement to keep the sweeper's central axis aligned with the stair climber's docking positioning area. This includes: receiving the stair climber's position offset and azimuth deviation value in real time; calculating the sweeper's correction distance in the lateral direction based on the position offset, and calculating the attitude adjustment angle based on the azimuth deviation value; generating a trajectory correction command based on the sweeper's correction distance in the lateral direction and attitude adjustment angle, and synchronously adjusting the travel direction and wheel speed difference in the next movement cycle to correct the sweeper's travel trajectory; when the position offset and azimuth deviation value in multiple consecutive frames of recognition output are both less than the corresponding preset threshold, maintaining the current travel parameters and keeping the sweeper's central axis aligned with the stair climber's docking positioning area.

[0016] Optionally, when a docking completion signal is received from the stair climber, the movement stops and the cleaning components are retracted, entering the transport state. This includes: parsing the received docking completion signal and generating a state switching command after the signal verification is successful; executing the state switching command, stopping the movement and maintaining the current position; driving the cleaning components to perform a retraction action, and turning off the power to the cleaning motor and vacuum fan; after the cleaning components are retracted, updating the current operating mode identifier to the transport state. A third aspect of the present invention provides a stair-climbing frame for performing the aforementioned stair-climbing frame and robot vacuum cleaner collaborative docking method. The stair-climbing frame includes: a main controller for controlling the overall operation of the stair-climbing frame and coordinating the signal interaction of other functional modules within the stair-climbing frame; a drive unit electrically connected to the main controller for executing traveling and turning movements according to drive control commands; a communication module signal-connected to the main controller for receiving docking requests, exchanging status information, and sending signals with the robot vacuum cleaner; a sensing module signal-connected to the main controller for collecting the spatial position and attitude angle information of the robot vacuum cleaner and outputting it to the main controller; a docking platform installed at the front end of the stair-climbing frame and signal-connected to the main controller for supporting the robot vacuum cleaner and performing locking actions during docking; and an optocoupler sensor disposed on the docking platform and signal-connected to the main controller for outputting a trigger signal for the main controller to determine docking when the robot vacuum cleaner enters the docking position.

[0017] A fourth aspect of the present invention provides a sweeping machine, which is used to perform the above-described stair climber and sweeping machine collaborative docking method. The sweeping machine includes: a sweeping machine control unit, used to control the overall operation of the sweeping machine and coordinate the signal interaction of other functional units in the sweeping machine; a travel drive component, electrically connected to the sweeping machine control unit, used to perform forward movement, turning, and speed adjustment according to travel control commands; a communication interaction unit, signal-connected to the sweeping machine control unit, used to send docking requests to the stair climber, receive docking completion signals, and exchange status information; an environmental recognition unit, signal-connected to the sweeping machine control unit, used to collect the spatial position and attitude angle information of the stair climber and output the recognition results to the sweeping machine control unit; and a state switching mechanism, signal-connected to the sweeping machine control unit, used to stop the output of the travel drive component and control the cleaning components to retract and enter the transport state after receiving the docking completion signal.

[0018] The fifth aspect of the present invention provides a stair-climbing frame and a sweeping machine collaborative docking system, the system comprising the aforementioned stair-climbing frame and the aforementioned sweeping machine. The sixth aspect of the present invention provides a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the above-described method for coordinated docking of the stair climber and the sweeping machine.

[0019] A seventh aspect of the present invention provides an electronic device, the electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described method for coordinated docking of the stair climber and the sweeping machine.

[0020] The eighth aspect of the present invention provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method for coordinated docking of a stair climber and a sweeping machine.

[0021] Through the above technical solution, this invention achieves fully closed-loop collaborative control between the stair climber and the robot vacuum during the docking process. Upon receiving a docking request from the robot vacuum, the stair climber automatically activates its sensing module to acquire pose information and calculates relative distance and orientation based on real-time sensing data, thereby dynamically determining the docking motion mode and drive commands. During the motion execution phase, the stair climber continuously corrects its trajectory based on the real-time output of the sensing module, ensuring the docking platform remains precisely aligned with the robot vacuum. When the optocoupler sensor detects a stable trigger signal, the system automatically locks and sends back a docking completion command, achieving fully automatic identification, positioning, and confirmation of the docking process. This process significantly improves the docking accuracy and reliability between multiple robots, avoids errors caused by human intervention, and ensures stable and smooth transfer of the robot vacuum between different floors. In particular, through dynamic motion mode switching (such as opposing motion for open spaces and same-direction chasing for narrow passages), the system can adapt to complex home environments, overcoming the limitations of traditional fixed-path docking.

[0022] Other features and advantages of the embodiments of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0023] The accompanying drawings are provided to further illustrate embodiments of the present invention and form part of the specification. They are used together with the following detailed description to explain the embodiments of the present invention, but do not constitute a limitation thereof. In the drawings: Figure 1 This is a flowchart illustrating the steps of a method for the coordinated docking of a stair-climbing frame and a sweeping machine according to one embodiment of the present invention. Figure 2 This is a schematic diagram of two docking motion modes provided in one embodiment of the present invention; Figure 3 This is a flowchart illustrating the overall process of a method for coordinated docking of a stair-climbing frame and a sweeping machine according to one embodiment of the present invention. Figure 4 This is a schematic diagram of the structure of a stair-climbing frame provided in one embodiment of the present invention; Figure 5 This is a schematic diagram of the structure of a sweeper provided in one embodiment of the present invention; Figure 6 This is a block diagram of the overall structure of the stair-climbing frame and sweeping machine collaborative docking system provided in one embodiment of the present invention. Detailed Implementation

[0024] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0025] Figure 1 This is a flowchart illustrating the steps of a method for the coordinated docking of a stair-climbing frame and a sweeping machine according to one embodiment of the present invention. Figure 1 As shown, an embodiment of the present invention provides a method for the coordinated docking of a stair-climbing frame and a sweeping robot, the method comprising: Step S10: The sweeping robot sends a corresponding docking request signal to the stair climber. The sweeping robot and the stair climber respectively acquire perception data to describe the spatial position and attitude angle of the other.

[0026] Specifically, the stair-climbing frame acquires image frame data of the sweeper and extracts pixel features from the image frame data to describe the boundary morphology of the sweeper; it calculates the displacement vector and attitude angle change of the sweeper based on the feature point matching results between adjacent image frames; and it generates perception data to describe the spatial position and attitude angle of the sweeper based on the displacement vector and attitude angle change.

[0027] Furthermore, image frame data of the sweeping machine is acquired, and pixel features describing the boundary morphology of the sweeping machine are extracted from the image frame data. This includes: performing specific visual marker recognition in consecutive frames, extracting pixel positioning points based on preset QR code markers or ArUco markers on the sweeping machine; calculating the pose solution matrix of the markers according to the distribution relationship of the pixel positioning points in the image frame coordinate system, which is used to determine the azimuth angle and center position of the sweeping machine in the environmental coordinate system; when the recognition result does not meet the feature extraction accuracy requirements, calling the depth gradient data in the image frame to perform edge feature matching to correct the boundary contour position of the sweeping machine.

[0028] Synchronously, the sweeping robot collects image frame data containing the outline of the stair climber; identifies visual markers set on the outer surface of the stair climber in the image frame data; calculates the attitude angle parameters of the stair climber based on the distribution of the positioning points of the visual markers in the image frame coordinate system, and infers the spatial position of the stair climber in the environmental coordinate system by combining the image depth information; and combines the calculated spatial position of the stair climber with the attitude angle parameters to generate recognition data describing the spatial position and attitude angle of the stair climber.

[0029] In this embodiment of the invention, during the docking phase, the robot vacuum cleaner first initiates a collaborative command, that is, sends a docking request signal to the stair climber. The purpose of this request is to establish a clear interaction trigger point, allowing both parties to enter the docking recognition state.

[0030] Upon receiving the request, the stair climber immediately activates its own sensing link, beginning to capture the data stream used to identify the robot vacuum's pose. Simultaneously, the robot vacuum also initiates its own recognition program, acquiring the stair climber's position information from its own perspective. The recognition actions at both ends serve as mutual references, forming a dynamic, two-way spatial perception process. In this way, regardless of changes in lighting or spatial obstructions in the environment, both can achieve spatial information redundancy through independent sensing data, thereby improving the stability and fault tolerance of the recognition process.

[0031] Once activated by the main control unit, the sensing unit on the stair-climbing frame collects image frame data of the robot vacuum cleaner in real time. The image sampling frequency is typically above tens of hertz to ensure a sufficiently continuous frame sequence. In each image frame, boundary feature extraction is performed, primarily by identifying the contour lines and color distribution of the robot vacuum cleaner's casing to generate a preliminary set of pixel features. Next, between adjacent image frames, the sensing unit calculates the motion trajectory of the pixels using a feature point matching algorithm. By comparing the positional differences of the same feature point in different frames, the displacement vector of the robot vacuum cleaner can be obtained, and the change in attitude angle can be calculated by combining the rotation relationship of the matching points. After feature point aggregation and anomaly removal, this displacement and angle data are integrated into a set of sensing data describing the spatial position and attitude angle of the robot vacuum cleaner. The entire process requires no external calibration or additional positioning equipment; dynamic tracking of the target can be achieved solely through visual sequences.

[0032] In some scenarios, relying solely on contour or edge information may not be sufficient to support high-precision recognition. For example, in areas with ground reflection or low texture, the number of feature points is limited, making matching more difficult. Therefore, the perception unit also performs specific visual marker recognition tasks. The stair climber detects QR codes or ArUco (ArUcomarker, a two-dimensional square visual marker used for visual recognition and attitude calculation) on the robot vacuum cleaner within a predetermined field of view. By analyzing the distribution of these markers' positioning points in the image frame coordinate system, it calculates the attitude calculation matrix. The attitude calculation matrix reflects the rotation and translation relationship of the target relative to the camera coordinate system. After matrix decomposition, the azimuth and center position of the robot vacuum cleaner in the environmental coordinate system can be obtained. When the marker recognition confidence level is detected to be below a threshold or there is partial occlusion, the algorithm automatically switches to edge gradient mode, refitting the outer contour by analyzing the depth gradient changes in the image. This dual-channel recognition method effectively improves the continuity and robustness of the data, enabling the stair climber to maintain stable recognition performance under different lighting and angles.

[0033] While the stair climber performs the aforementioned recognition, the robot vacuum cleaner also performs corresponding operations. The robot vacuum cleaner's control unit drives its environmental recognition unit to collect image frame data containing the outline of the stair climber. Similar to the stair climber recognition logic, the robot vacuum cleaner also identifies preset visual markers on the outer surface of the stair climber in the image, such as QR codes or matrix markers. By calculating the relative distribution of these markers in the image frame coordinate system, the robot vacuum cleaner can deduce the attitude angle parameters of the stair climber and combine them with the depth information in the camera model to solve for the spatial position of the stair climber in the environmental coordinate system. Then, the robot vacuum cleaner combines the calculated spatial position and attitude angle parameters to form a complete recognition dataset, which is used for subsequent trajectory planning and path correction in conjunction with motion.

[0034] This bidirectional recognition mechanism ensures that the recognition error at either end is balanced by the measurement result at the other end, effectively introducing a cooperative constraint in spatial positioning. Even if the recognition end of the stair-climbing frame experiences slight vibration or obstruction of its field of vision, the robot vacuum cleaner can still provide a correction reference through its own sensing results, thereby avoiding deviations caused by accumulated errors. Furthermore, after the recognition data from both sides is exchanged in real time through the communication module, joint calculation of relative pose can be achieved, ensuring that the coordinate references of both sides remain consistent, and subsequent docking movements can converge along the optimal path.

[0035] In another possible implementation, to further improve the docking stability of the stair climber and the robot vacuum in complex environments, a dynamic illumination adaptive recognition mechanism can be introduced on the basis of bidirectional recognition. When there are highly reflective floor tiles, glass doors, or changes in natural lighting in the home environment, conventional image features are prone to overexposure or shadow occlusion, leading to unstable feature extraction. Therefore, in this implementation, the sensing units of both the stair climber and the robot vacuum have a dynamic brightness adjustment algorithm, which can automatically adjust the exposure parameters and contrast threshold according to the brightness histogram information of the current frame, so that the recognized image maintains a stable feature distribution in both high-brightness and low-brightness areas.

[0036] Specifically, after acquiring image frame data, the stair climber first performs illumination equalization preprocessing, calculates the global brightness distribution, and applies local gain to dark areas to eliminate edge breaks caused by shadows. Simultaneously, it applies brightness suppression to overexposed areas to preserve the boundary clarity of the QR code or ArUco marker. The robot vacuum cleaner employs the same strategy when recognizing the stair climber's shape, using ambient brightness estimation and dynamic grayscale compensation to ensure the marker detection algorithm maintains consistent response values ​​under different illumination levels.

[0037] Through this light-adaptive recognition mechanism, the stair climber and the sweeper can stably identify each other's marked positions and boundary contours in direct sunlight in the morning, low illumination at night, or mixed indoor lighting environments, greatly reducing attitude calculation errors caused by feature drift, thereby ensuring the continuity of the docking path and locking accuracy.

[0038] Step S20: The stair climber and the sweeper determine the relative distance and relative orientation information between themselves and each other based on the sensing data, and determine the docking motion mode based on the relative distance and relative orientation information, and generate the corresponding drive control command.

[0039] Specifically, the stair-climbing frame determines the relative distance and relative orientation information between itself and the sweeping machine based on the perceived data, including: parsing the coordinate values ​​of the sweeping machine's center position in the perceived data and establishing a unified coordinate reference system in combination with the real-time pose parameters of the stair-climbing frame itself; calculating the planar distance between itself and the sweeping machine in the horizontal direction and the height difference in the vertical direction based on the unified coordinate reference system to generate the relative distance between itself and the sweeping machine; calculating the relative orientation angle based on the difference between the sweeping machine's attitude angle and the stair-climbing frame's orientation angle, which is used to describe the angle relationship between the forward direction of the stair-climbing frame and the center line of the sweeping machine; combining the relative distance and the relative orientation angle to form a relative position information data packet between itself and the sweeping machine, and writing it into the docking motion calculation buffer.

[0040] Furthermore, the stair-climbing frame determines the docking motion mode based on relative distance and relative orientation information, and generates corresponding drive control commands, including: reading the distance difference and angle deviation values ​​in the relative distance and relative orientation information, and comparing them with preset thresholds respectively; when the angle deviation value is less than the preset angle threshold and the distance difference value is greater than the preset distance threshold, it is determined to be a reciprocating motion mode, and a first drive control command including forward direction, travel speed and correction frequency is generated; when the angle deviation value is not less than the preset angle threshold or the environmental constraint information indicates that the passage (such as a narrow passage) is restricted, it is determined to be a same-direction chasing mode, and a second drive control command including path following parameters, relative speed difference and lateral correction amount is generated; the first drive control command or the second drive control command is written into the control buffer area for executing the corresponding docking motion.

[0041] Simultaneously, the sweeper analyzes and identifies the coordinate parameters and attitude angle parameters of the stair climber's spatial position in the data, and establishes a unified coordinate reference system by combining the sweeper's own position parameters. Under the unified coordinate reference system, it calculates the planar distance, vertical height difference, and centerline angle between the sweeper and the stair climber, forming relative distance and relative orientation information. Based on the relative distance and relative orientation information, it compares them with preset motion thresholds. When the angle deviation value is less than the preset angle threshold, it is determined to be a straight-line docking mode. When the angle deviation value is not less than the preset angle threshold or the spatial passage is limited, it is determined to be a following docking mode. Based on the determined docking mode, it generates a travel control command that includes the travel direction, speed correction value, and attitude correction amount.

[0042] In this embodiment of the invention, during the specific implementation process, after the stair-climbing frame and the sweeping robot enter the docking phase, they each independently calculate the spatial relationship parameters between themselves and the other based on the sensing data collected by their own devices. This phase determines whether the two devices can accurately align and smoothly approach each other in three-dimensional space.

[0043] The stair-climbing frame first analyzes the coordinate information of the robot vacuum cleaner from the perceived data, which is usually obtained from the previous stage's visual recognition unit or depth camera. To ensure consistency in calculations, the stair-climbing frame integrates its real-time pose parameters with the recognized center position coordinates of the robot vacuum cleaner into the same coordinate reference system, thus avoiding errors caused by changes in the device's own posture. This unified coordinate reference system typically uses the center of the stair-climbing frame chassis as the origin, the forward direction as the X-axis, the left and right directions as the Y-axis, and the vertical direction as the Z-axis. This spatial definition provides a fixed reference frame for subsequent distance and angle calculations.

[0044] Based on this, the stair climber calculates the horizontal distance between the two components using coordinate differences, and simultaneously calculates the vertical height difference based on height sensors or visual depth information. The former reflects the relative displacement of the two components on the ground, while the latter represents the floor difference in height. For example, if the robot vacuum is at the bottom of a staircase and the stair climber is on the first step, this difference can indicate whether the stair climber needs to adjust its angle or tilt to achieve accurate docking. Next, the stair climber calculates the relative azimuth angle based on the difference between the robot vacuum's attitude angle change and its own orientation angle. This parameter can be understood as the angle between the stair climber's forward direction and the robot vacuum's centerline, playing a role similar to rudder angle correction in control. If the relative azimuth angle deviation is large, it means that the two are not face-to-face but have a certain degree of deflection, and subsequent movement patterns must be adjusted; otherwise, docking misalignment may occur.

[0045] After completing the above calculations, the stair-climbing frame combines the relative distances and relative azimuths to form a complete set of spatial description data. For easy and rapid retrieval, this data is typically packaged into relative position information data packets and written to the motion control buffer. This buffer provides a real-time updated reference data pool, enabling the motion control module to obtain the latest pose parameters within milliseconds. Whenever the sensing module receives new data input, the main control unit immediately updates the contents of this buffer, ensuring that the motion trajectory calculation is based on the latest environmental information without delay or drift.

[0046] After obtaining relative position information, the climbing frame needs to determine its own movement mode and generate corresponding drive control commands. The first step of the control logic is to read the relative distance and relative orientation information stored in the buffer, and then extract the distance difference and angle deviation values ​​respectively. Next, these two key parameters are compared with preset thresholds to determine the most suitable docking method in the current scenario. If the angle deviation value is small, it means that the two are basically face-to-face and the space is open, in which case the opposing movement mode is suitable; conversely, if the angle deviation is large or the environmental sensor information shows that the space is narrow or there are obstacles, then the same-direction chasing mode is safer.

[0047] In the opposing motion mode, the stair climber generates a control command that includes the direction of travel, speed, and correction frequency. The core of this command is to bring the two devices closer together along the same axis, performing minor corrections in each sensing refresh cycle to maintain alignment of the central axis. The correction frequency setting balances response speed and stability, typically adjusted dynamically based on the relative distance. When the distance is greater, the correction frequency is lower to avoid excessive swaying; as the two devices approach, the frequency gradually increases for more precise position convergence. In the same-direction chasing mode, the control logic differs. The stair climber then generates a second control command that includes path following parameters, relative speed difference, and lateral correction. The "path following parameters" define a dynamic target trajectory, and the stair climber gradually approaches along the same path at a speed slightly higher than the robot vacuum. The lateral correction corrects lateral deviations between the two devices, ensuring that the docking platform and the robot vacuum's trajectories overlap during the chasing process.

[0048] After the control command is generated, the main control unit writes it into the motion control buffer to drive the chassis actuators. The chassis drive adjusts the rotational speed of each wheel set or track in real time according to the direction vector and speed parameters defined in the command, thereby achieving a smooth approach. During the docking process, the sensing module remains active, continuously outputting updated pose data. The motion control module performs closed-loop adjustments based on this real-time data, ensuring the entire approach process remains continuous and stable.

[0049] Meanwhile, the robot vacuum is also executing corresponding coordination logic. It analyzes and identifies the spatial coordinates and attitude angle parameters of the stair-climbing frame in the data, establishing a unified coordinate reference system with its own chassis center as the origin. Within this unified coordinate system, it calculates the planar distance, vertical height difference, and centerline angle between the robot vacuum and the stair-climbing frame, thus forming relative distance and orientation information. By comparing this data with preset motion thresholds, the robot vacuum can determine which coordination mode to adopt. If the angle deviation is small, it executes a straight-line docking coordination mode; if the angle deviation is large or the environmental detection indicates limited space, it switches to a following docking coordination mode.

[0050] In the straight-line docking mode, the robot vacuum primarily approaches the stair climber along a straight line, making subtle posture corrections based on the output of the environmental recognition unit to reduce accumulated deviations. In the following mode, the robot vacuum travels at a slightly slower speed than the stair climber, constantly adjusting its direction to keep its central axis aligned with the climber's direction of travel. The two devices continuously exchange pose data through their respective communication modules, ensuring synchronized docking paths. The travel control commands generated by the robot vacuum typically include a travel direction vector, a speed correction value, and a posture correction amount. The speed correction value can be dynamically adjusted in real-time communication feedback to avoid overshoot or delay caused by excessive speed differences.

[0051] Throughout the process, although the stair climber and the sweeper control their movements independently, they share each other's key spatial parameters and execute actions using the same reference coordinate logic. The stair climber's opposing or chasing mode and the sweeper's straight-line or following mode are actually complementary. Any deviation adjustment at one end is instantly transmitted to the other end via the communication link, thus forming a flexible and collaborative motion control loop. This structure allows the docking path to achieve self-organizing coordination without relying on an absolute positioning system, through relative pose calculation.

[0052] This dual-mode motion determination based on relative distance and relative orientation information enables the stair climber and the sweeping robot to adaptively dock under different environmental constraints. The opposing motion mode is suitable for open environments and can quickly converge to the target position; while the chasing mode is particularly suitable for narrow areas such as corridors and doorways, avoiding frequent turning around. Through dual judgment of angle and distance thresholds, the two devices can automatically select the optimal path under different spatial structures, significantly improving the success rate and efficiency of docking.

[0053] Step S30: Execute the docking movement of the stair climber and the sweeper according to the drive control command, and synchronously correct the travel trajectory in real time during the docking movement to maintain the alignment between the two.

[0054] Specifically, based on the drive control command, the stair-climbing frame is controlled to move in a straight line along the docking direction; the position offset and azimuth deviation of the sweeper are monitored in real time, and the lateral correction and steering angle correction of the stair-climbing frame are calculated in real time based on the offset and azimuth deviation; a trajectory adjustment command is generated according to the lateral correction and steering angle correction to dynamically correct the trajectory of the stair-climbing frame; in several consecutive frames of perception output, when the offset and azimuth deviation are both less than the corresponding preset threshold, the current direction of travel is maintained and the alignment between the docking platform and the sweeper is maintained.

[0055] Synchronously, the sweeper receives the position offset and azimuth deviation of the stair climber in real time; it calculates the correction distance of the sweeper in the lateral direction based on the position offset and the attitude adjustment angle based on the azimuth deviation; it generates a trajectory correction command based on the correction distance and attitude adjustment angle of the sweeper in the lateral direction, and synchronously adjusts the travel direction and wheel speed difference in the next motion cycle to correct the sweeper's travel trajectory; when the position offset and azimuth deviation are both less than the corresponding preset threshold in multiple consecutive frames of recognition output, the current travel parameters are maintained and the sweeper's central axis is kept aligned with the stair climber's docking positioning area.

[0056] In this embodiment of the invention, during the docking phase, the stair-climbing frame and the sweeping robot have calculated each other's spatial pose information and determined their movement modes. The next step is for both to synchronize and continuously align their trajectories in a dynamic environment. This phase is a continuous and delicate coordination process: every instantaneous offset and angle change needs to be captured, calculated, and converted into corresponding trajectory correction amounts, so that the two are pulled together stably by an invisible thread, converging towards the same target point.

[0057] In practice, the stair-climbing frame moves in a straight line along the predetermined docking direction according to the previously generated drive control commands. The control logic reads the real-time output from the sensing module in each update cycle. This data includes the lateral offset of the sweeper relative to the central axis of the stair-climbing frame and the azimuth deviation between the two. The offset can be understood as the horizontal misalignment distance between the two devices; the azimuth deviation reflects the angle between their central axes. The stair-climbing frame needs to dynamically input these two parameters into the motion control model and calculate the lateral correction and steering angle correction through mathematical calculations. Generally, this calculation uses a proportional control method, applying a higher correction amplitude when the offset is large, and gradually converging as the offset decreases to prevent oscillations.

[0058] After calculating the correction parameters, the stair-climbing frame generates a trajectory adjustment command. This command includes real-time direction vectors and velocity correction coefficients to control the differential output of the wheelset or tracks. Unlike traditional path following, this adjustment does not follow a predetermined trajectory but rather corrects for deviations in real time, forming a self-closing dynamic trajectory. The stair-climbing frame updates its direction of travel and angular velocity every tens of milliseconds, making the movement smoother and avoiding abrupt direction changes. The advantage of this is that even in a home environment with uneven ground friction, local tilt, or minor obstacles, the accuracy of the final docking posture will not be affected.

[0059] To avoid unnecessary frequent corrections, the algorithm also sets a convergence condition. When the position offset and azimuth deviation of the robot vacuum cleaner are simultaneously less than the corresponding preset thresholds in several consecutive frames of sensing data, the control logic considers that a stable alignment state has been reached. At this point, it stops outputting new correction commands and only maintains the current direction of travel. Through this "threshold-based convergence" design, the stair-climbing frame can reduce unnecessary attitude adjustments while ensuring accuracy, thus making the docking process smoother and reducing drive energy consumption.

[0060] Meanwhile, the robot vacuum's actions are also performed in real time. During its movement, the robot vacuum continuously receives posture deviation information transmitted by the stair climber, including the stair climber's position offset and azimuth deviation. Unlike the stair climber, the robot vacuum emphasizes "attitude response." It calculates its own correction distance in the lateral direction based on the received offset and calculates the attitude adjustment angle based on the azimuth deviation. Combining these two parameters generates a trajectory correction command for the next cycle.

[0061] During the correction process, the sweeper adjusts its direction by controlling the speed difference between the left and right drive wheels, while simultaneously making subtle adjustments to its speed, gradually aligning its central axis with the docking positioning area of ​​the stair-climbing frame. This process is similar to an automated guided vehicle aligning its trajectory on a dynamic path, except that the reference point here is another moving object. To ensure the smoothness of the correction, the algorithm also incorporates dynamic gain adjustment. When a rapid rate of change in deviation is detected, the correction amplitude is reduced to prevent oscillations; while when the deviation is stable but not fully aligned, the response speed is appropriately increased to accelerate convergence.

[0062] In consecutive recognition frames, if the robot vacuum detects that both the position offset and azimuth deviation are less than the corresponding thresholds, it considers the alignment complete and then maintains its current travel parameters, entering a locked following state. At this time, the robot vacuum no longer actively adjusts its posture, but maintains its alignment with the stair climber through inertia until the stair climber sends a final docking confirmation signal.

[0063] The entire docking process is essentially a two-way closed-loop control. The stair climber is responsible for guiding and converging the macroscopic trajectory, while the robot vacuum is responsible for adjusting and coordinating the microscopic posture. Motion data from both sides is exchanged in real time via a communication module, forming a stable dynamic feedback loop: changes in the stair climber's movement are immediately reflected in the robot vacuum's recognition results, and the robot vacuum's posture adjustments, in turn, influence the stair climber's next correction. This complementary design allows for centimeter-level spatial alignment without the need for external navigation base stations or fixed markers.

[0064] The trajectory correction mechanism in this stage not only improves docking accuracy but, more importantly, enhances environmental adaptability. In traditional solutions, when faced with smooth ground, changes in slope, or confined spaces, unilateral control often struggles to maintain stable alignment, easily leading to accumulated deviations. However, in this solution, because both the stair-climbing frame and the sweeping robot possess autonomous correction capabilities, even if recognition on one side is temporarily limited, the deviation adjustment on the other side can still compensate, ensuring that the overall trajectory does not deviate.

[0065] Furthermore, this bidirectional real-time correction method significantly optimizes the smoothness of movement. Because each adjustment is completed through continuous changes in relative pose, there is no abrupt "pause, correction, and then forward" process. For the user, the entire docking process appears more natural and seamless, as if two devices are slowly approaching each other while "searching for each other" until they are firmly connected. More importantly, after multiple dockings, the stair climber and the robot vacuum cleaner can automatically fine-tune thresholds through algorithmic learning, making subsequent actions more efficient. This adaptive closed-loop control approach also provides greater potential for future multi-robot collaboration.

[0066] Overall, through dual-end execution of drive control commands and real-time trajectory correction, the stair climber and the sweeper achieve true dynamic coordinated movement. Whether on open ground or at complex stairwell entrances, the two devices can achieve alignment and contact with minimal deviation, laying a precise foundation for subsequent locking and handling.

[0067] Step S40: After the stair climber docks with the sweeper, it performs a locking operation with the sweeper and sends a docking completion command to the sweeper until the predetermined target position is reached. Then, it releases the lock. When the sweeper receives the docking completion signal from the stair climber, it stops moving and retracts the cleaning components, entering the transport state.

[0068] Specifically, the stair climber monitors the level signal status output by the optocoupler sensor in the stair climber and determines whether the level signal remains stable for a preset duration within a continuous detection cycle. When the level signal remains valid, a locking execution command is generated and sent to the locking drive component of the docking platform of the stair climber, controlling the docking platform to perform a mechanical locking action. After the locking action is completed, the feedback signal of the optocoupler sensor is read to confirm the locking state, and the locking confirmation result is combined with the docking state identifier to generate a docking completion data packet. The docking completion data packet is sent to the sweeper. After the stair climber carries the sweeper to the predetermined target position, an unlocking command is generated based on the arrival determination result, and the docking platform is controlled to perform the corresponding unlocking action.

[0069] Synchronously, the robot vacuum cleaner parses the received docking completion signal and generates a state switching command after the signal verification is successful; it executes the state switching command, stops moving and maintains the current position; drives the cleaning components to perform a retraction action, and turns off the power of the cleaning motor and the vacuum fan; after the cleaning components have retracted, it updates the current operating mode identifier to the transport state.

[0070] In this embodiment of the invention, after achieving precise spatial alignment, the docking process enters its final stage: locking and state switching. The main task of this stage is to transition the stair-climbing frame and the sweeper from a relatively independent state of motion to a stable physically connected state, preparing for subsequent handling actions. Although it may seem like just a locking and command-sending action, this stage actually involves multiple steps, including signal determination, mechanical action confirmation, feedback verification, and dual-end synchronous switching. Every detail determines the reliability of the entire docking process.

[0071] Once physical contact is established, the optocoupler sensor inside the stair climber begins its crucial detection task. This sensor, consisting of a transmitter and receiver located on either side of the docking platform, works by detecting changes in light blocking or reflection to determine if the robot vacuum has accurately entered the designated position. To avoid false alarms caused by brief contact or slight vibrations, the stair climber is designed with a continuous detection cycle. Only when the sensor's output signal remains stable and valid for a preset duration, such as several hundred milliseconds to one second, will the main control logic classify this state as "valid contact." This filters out transient vibrations and prevents false triggering caused by light interference.

[0072] Once the voltage level signal is confirmed to be stable, the control logic immediately generates a locking execution command and sends it to the locking drive component of the docking platform. The locking mechanism typically consists of an electromagnetic latch or a mechanical gear lock claw, which completes the locking action through short-pulse current drive. During execution, the platform's micro-displacement is monitored in real time to ensure that the robot vacuum chassis is completely seated in the support groove. To prevent damage to the outer casing due to over-clamping, a mechanical limit is provided for the locking stroke, and the current detection module also monitors the current change at the moment of locking. When the damping rises to a set threshold, the drive automatically stops.

[0073] After the locking action is completed, the control logic does not immediately end the process. Instead, it reads the feedback signal from the optocoupler sensor again and compares it with the locking status signal. If they match, it means the robot vacuum has been firmly fixed to the docking platform. At this point, the control logic packages the locking confirmation result and the docking status identifier together to generate a "docking completion data packet." This data packet contains not only the docking completion signal but also a timestamp and locking status code for verification by the robot vacuum. Subsequently, the data packet is sent to the robot vacuum via the communication module, announcing that the physical docking and signal alignment are complete.

[0074] Meanwhile, after receiving the docking completion signal from the stair climber, the robot vacuum initiates its own state switching process. First, it parses and verifies the signal, comparing the checksum in the data packet to ensure uninterrupted or unlost signal transmission. Once confirmed, it generates a state switching command and sends it to the execution module. Based on this command, the robot vacuum immediately stops moving, and the drive unit output resets to zero, ensuring stable stillness on the docking platform. Next, the state switching module begins driving the retraction mechanism of the cleaning components, typically including side brush folding, roller brush lifting, and shutting off the vacuum fan. To prevent mechanical interference during movement, the retraction sequence follows a "front, middle, rear" logic, ensuring each component gradually detaches from the ground during retraction.

[0075] Once all cleaning components have retracted, the robot vacuum will update its internal operating mode, switching the current status indicator from "cleaning operation" to "transportation mode." This switch ensures that all subsequent control logic operates as if the device is being transported, including pausing autonomous navigation, disabling obstacle avoidance algorithms, and locking wheel brakes. This prevents unnecessary malfunctions or interference during the climbing and movement of the stair climber.

[0076] The stair-climbing robot needs to confirm reliable mechanical locking and consistent signal feedback, while the robot vacuum cleaner must also verify its own status is synchronized with the stair-climbing robot through data verification. The verification actions of both parties almost overlap in time, ensuring that any delay on either side will not lead to malfunctions. For example, if communication is momentarily interrupted, the robot vacuum cleaner will not prematurely enter the carrying state; it will only switch after receiving a complete and verified signal. This dual verification mechanism provides the entire process with extremely high safety and fault tolerance.

[0077] Specifically, the unlocking process follows a joint judgment rule based on task progress and position status. While carrying the sweeper, the stair-climbing frame continuously collects position information, posture data, and motion status data, comparing this data with a preset target position to determine if the arrival conditions are met. Once the arrival conditions are met, the control component enters the unlocking judgment phase, monitoring whether the movement speed is below a preset threshold, whether the platform posture is within a stable range, and whether external interference is within an acceptable range to confirm the safety of the unlocking environment. Subsequently, the control component generates an unlocking command and sends the corresponding unlocking action control quantity to the locking drive component of the docking platform, causing the mechanical locking structure to complete the unlocking process in a predetermined sequence. The status signal of the docking area is read again by an optocoupler sensor to verify that the sweeper has completely broken free of the locking constraint, thus completing the entire unlocking process.

[0078] The design combining optocoupler detection and mechanical locking achieves both high-precision position determination and reliable mechanical fixation. Optical detection provides sub-centimeter-level position resolution, while mechanical locking ensures structural stability under complex vibration environments. Furthermore, subsequent data feedback and communication confirmation create a complete closed loop for the collaborative process between the stair climber and the sweeper, encompassing perception, judgment, execution, and feedback.

[0079] In another possible implementation, to further improve the reliability of the stair climber and the sweeper during the docking phase, a multi-signal fusion judgment mechanism can be introduced into the locking operation. While traditional single photoelectric detection can identify whether docking is successful, false triggering may still occur under specific lighting conditions, such as strong light reflection, dust obstruction, or signal jitter caused by slight angular deviations. To address this, this implementation adds angle attitude detection and micro-displacement feedback in addition to the photoelectric coupling sensor to jointly determine whether the sweeper has reached a stable docking position.

[0080] Specifically, two sets of displacement sensing units are arranged on the docking platform of the stair climber to detect minute deformations of the contact surface between the platform and the bottom of the sweeper. When the photoelectric sensor outputs a stable signal, if the displacement sensing unit detects that the rate of change of deformation is less than a set threshold, it indicates that the sweeper has landed smoothly on the platform and is stationary. Conversely, if continuous shaking or deformation fluctuations are detected, the main control logic will delay the locking action and re-verify the docking position. At the same time, the stair climber will also call the internal attitude detection module to determine the attitude stability of the platform before and after locking by measuring the instantaneous changes in its pitch and roll angles. If the attitude change of the stair climber is less than a preset threshold after the locking action begins, it is determined that the locking execution environment is stable and the next action can proceed.

[0081] In the locking confirmation process, this implementation uses dual-channel feedback, meaning that both photoelectric and displacement signals are simultaneously returned to the main control logic. A locking confirmation flag is only generated when both signals remain stable and continuous. At this point, the main control logic generates a docking completion data packet and sends it to the robot vacuum. This data packet additionally includes a locking redundancy status code and a posture stability flag to improve the confirmation accuracy on the robot vacuum's end. Upon receiving this signal, the robot vacuum not only verifies signal integrity but also reads the posture stability flag to determine if there is any slight tilt. If the stair climber's posture is detected to be unstable, the robot vacuum will delay the timing of retracting the cleaning components, waiting for the locking to be fully stable before switching to transport mode. Through these improvements, even on stair edges, on sloping surfaces, or in environments with strong vibrations, the stair climber and robot vacuum can still ensure the stability and safety of the docking operation through multi-signal redundancy judgment.

[0082] like Figure 2In one specific implementation, the stair-climbing frame and the sweeping robot perform corresponding collaborative docking processes under different relative movement scenarios. Figure 2 In the example on the left, the sweeper moves along a straight line towards the stair-climbing frame. Upon receiving the docking request, the stair-climbing frame initiates image acquisition, identifying the sweeper's outline or markers, and calculating the sweeper's displacement direction and attitude angle based on consecutive image frames. The stair-climbing frame writes the identification results and its own pose data into a unified reference frame to obtain the relative distance and relative orientation between the two. With this relative position stable and the centerlines basically aligned, the stair-climbing frame moves forward and performs minor lateral corrections, gradually aligning its front end with the docking area on the bottom of the sweeper, thus completing the forward approach. Figure 2 In the example on the right, the sweeper and the stair climber move in the same direction, and the sweeper is initially not at the center line of the stair climber. After recognizing the sweeper's movement trend, the stair climber calculates the relative positional offset between the two, and while maintaining its own direction of travel, it moves closer to the sweeper through continuous small lateral movements and attitude fine-tuning. As the offset gradually decreases, the stair climber eventually converges with the center line of the sweeper during its movement in the same direction, maintaining a dockable relative position and providing conditions for subsequent physical locking and collaborative handling.

[0083] During the docking process, the stair-climbing frame and the sweeper followed... Figure 3 The docking process is as follows: When the robotic vacuum cleaner navigates to target point A, it sends navigation arrival information to the stair climber. The stair climber then activates its navigation module and moves to target point B. As the stair climber approaches target point B, it compares its current position with the preset docking position. Once the proximity condition is met, it actively notifies the robotic vacuum cleaner to enter the docking preparation stage. Upon receiving the stair climber's position update, the robotic vacuum cleaner begins searching for the stair climber's signal. It uses its visual or communication recognition capabilities to acquire the docking signal from the stair climber and feeds the recognition result back to it. Based on the received signal, the stair climber determines that the robotic vacuum cleaner has entered the docking range and initiates docking actions based on the signal characteristics. These actions include adjusting the alignment angle, correcting lateral offset, and making slight forward movements to ensure that its docking position coincides with the robotic vacuum cleaner's docking area.

[0084] Once the docking signal between the robot vacuum and the stair climber is stable, both will perform the final docking action: the robot vacuum will enter a stationary state after receiving a signal confirmation from the stair climber, while the stair climber will continue to perform subsequent carrying or cross-floor tasks after confirming successful docking, thus completing the entire docking process.

[0085] like Figure 4The present invention also provides a stair-climbing frame for performing the above-described stair-climbing frame and sweeping robot collaborative docking method. The stair-climbing frame includes: a main controller for controlling the overall operation of the stair-climbing frame and coordinating the signal interaction of other functional modules in the stair-climbing frame; a drive unit electrically connected to the main controller for executing traveling and turning movements according to drive control commands; a communication module signal-connected to the main controller for receiving docking requests, exchanging status information, and sending signals with the sweeping robot; a sensing module signal-connected to the main controller for collecting the spatial position and attitude angle information of the sweeping robot and outputting it to the main controller; a docking platform installed at the front end of the stair-climbing frame and signal-connected to the main controller for supporting the sweeping robot and performing locking actions during docking; and an optocoupler sensor disposed on the docking platform and signal-connected to the main controller for outputting a trigger signal for the main controller to determine docking when the sweeping robot enters the docking position.

[0086] In this embodiment of the invention, the main controller, as the core control component of the stair-climbing frame, establishes data connections with various functional modules through an internal bus, undertaking core tasks such as motion coordination, signal distribution, and logical judgment. The main controller can receive spatial recognition data from the sensing module in real time and calculate motion command parameters by combining internal attitude sensing information. In terms of control logic, the main controller outputs graded command signals to the drive unit, corresponding to three motion states: forward, turning, and deceleration, enabling the stair-climbing frame to achieve refined motion control during the docking phase.

[0087] The drive unit employs a differential drive or tracked structure, achieving minute steering adjustments through dual-path independent control, enabling stable movement in confined spaces. The communication module serves as a data exchange channel, handling docking request reception, status synchronization, and the transmission of docking completion signals, ensuring information consistency between the stair climber and the sweeper throughout the docking process. The sensing module typically consists of a camera unit and attitude measurement components, used to collect the sweeper's spatial position and attitude angle information in real time, and send the processed sensing data to the main controller for calculations.

[0088] The stair-climbing frame has a docking platform at its front end, which supports the sweeper and performs the locking operation after docking. The platform structure adopts a composite design of rigid support and buffer layer, which can absorb some mechanical impact at the moment of locking. Optocoupler sensors are installed on both sides of the docking platform to detect whether the sweeper has accurately entered the designated position. When the sweeper blocks the sensor's light path, the main controller can determine that the contact has been established and trigger the locking operation.

[0089] With the above-mentioned structural cooperation, the stair climbing frame can achieve integrated operation of spatial recognition, motion control and status feedback when performing collaborative docking. It not only has high docking accuracy, but also can maintain a stable operating posture in various ground environments, providing reliable support for the sweeper to clean across floors.

[0090] like Figure 5 The present invention also provides a sweeping machine, which is used to perform the above-mentioned stair climber and sweeping machine collaborative docking method. The sweeping machine includes: a sweeping machine control unit, used to control the overall operation of the sweeping machine and coordinate the signal interaction of other functional units in the sweeping machine; a travel drive component, electrically connected to the sweeping machine control unit, used to perform forward movement, turning and speed adjustment according to travel control commands; a communication interaction unit, signal-connected to the sweeping machine control unit, used to send docking requests to the stair climber, receive docking completion signals and exchange status information; an environmental recognition unit, signal-connected to the sweeping machine control unit, used to collect the spatial position and attitude angle information of the stair climber and output the recognition results to the sweeping machine control unit; and a state switching mechanism, signal-connected to the sweeping machine control unit, used to stop the output of the travel drive component and control the cleaning components to retract and enter the transport state after receiving the docking completion signal.

[0091] In this embodiment of the invention, the sweeping robot control unit, as the core of the entire machine, is responsible for operational status management and command allocation. It integrates multi-threaded control logic, capable of simultaneously processing motion control, environmental recognition, and communication interaction signals, ensuring coordination and consistency among functional units. The driving component is electrically connected to the control unit, achieving precise path control through dual-motor differential speed or wheel drive. This driving component supports independent adjustment of linear and angular velocity, enabling it to perform minute steering and deceleration operations in narrow passages or at the edges of stairs, providing a hardware foundation for maintaining central axis alignment during docking.

[0092] The communication interaction unit is responsible for sending docking requests, synchronizing status, and confirming signals. It can operate in local wireless or near-field communication environments, ensuring low-latency data interaction between two devices when they are close to each other. After the stair climber locks in place, the unit receives and verifies the docking completion signal from the stair climber and feeds the result back to the control unit for status switching determination.

[0093] The environmental recognition unit is designed with a dual recognition path. First, it acquires the spatial contour and attitude angle information of the stair climber through a forward-looking camera or laser rangefinder. Second, when the recognition signal is unstable, it uses a specific image feature matching algorithm to identify marker points on the stair climber surface, thus maintaining reliable recognition even in complex environments. The recognition results are then fused internally and directly transmitted to the control unit to participate in subsequent trajectory correction and motion calculations.

[0094] After docking is complete, the state switching mechanism enters the working phase. Connected to the control unit, it automatically executes the retraction logic of the cleaning components based on the docking completion signal. Specifically, it first controls the travel drive component to stop outputting and maintain a stable position, then drives the side brush to fold, the roller brush to rise, and the vacuum fan to shut down, ensuring that the cleaning components do not malfunction due to vibration during transport. Finally, the state switching mechanism updates the robot vacuum's operating mode, switching it from cleaning mode to transport mode, providing conditions for the subsequent climbing and cross-floor transport by the stair-climbing frame.

[0095] Through the above structural design, the sweeping robot can not only complete cleaning tasks independently, but also has the ability to coordinate with the stair climber with high precision. The addition of a two-way communication and environmental recognition module enables it to respond to the movement adjustments of the stair climber in real time during docking, thereby achieving automatic and stable spatial coordination and significantly improving the intelligence level and execution reliability of multi-robot collaboration.

[0096] The present invention also provides a stair climbing frame and a sweeping machine collaborative docking system, the system including the stair climbing frame and the sweeping machine described above.

[0097] In one specific implementation, such as Figure 6 The robotic vacuum cleaner and the stair-climbing frame each have multiple functional modules that share task commands and relative position information via a two-way communication link. The robotic vacuum cleaner includes a motion control module, an infrared signal receiving module, a WiFi communication module, a positioning module, and a sensing module. The motion control module executes movement and adjustment actions, the infrared signal receiving module receives guidance signals from the stair-climbing frame, and the positioning and sensing modules acquire information about the robotic vacuum cleaner's own position and surrounding environment. The stair-climbing frame is equipped with corresponding motion control, infrared signal transmitting, WiFi communication, positioning, and sensing modules, enabling it to actively send task commands and pose prompts to the robotic vacuum cleaner and adjust its own position based on the output of the sensing module. Both maintain a stable data connection via the WiFi communication module, initiating and confirming the docking process through task commands, and using infrared signals to complete relative position calibration at close range, thus ensuring smooth execution of the docking action.

[0098] In this embodiment of the invention, the stair-climbing frame and sweeping robot collaborative docking system adopts a dual-body collaborative structure, consisting of the aforementioned stair-climbing frame and sweeping robot, which establish a real-time interactive connection through a wireless communication link. During collaborative operation, the stair-climbing frame undertakes the functions of identification, support, and locking, while the sweeping robot is responsible for positioning, response, and coordinated movement. Both parties periodically exchange pose data, motion status, and perception results through their respective configured communication modules, forming a closed-loop data channel. The perception data is parsed and updated in real time at both ends to determine the current docking stage and adjust the relative trajectory. When the sensing module of the stair-climbing frame detects that the sweeping robot has arrived and outputs a stable signal, it drives its docking platform to perform a locking action; simultaneously, the sweeping robot enters the transport state after receiving a confirmation signal.

[0099] The entire process is fully automated, requiring no human intervention and encompassing everything from docking request and motion calibration to lock confirmation. This system enables two independent mobile devices to achieve high-precision docking and status synchronization in different environments, providing a complete pathway for the coordinated control of multi-layered home cleaning equipment.

[0100] The present invention also provides a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the above-described method for coordinated docking of the stair climber and the sweeping robot.

[0101] The present invention also provides an electronic device, the electronic device including a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described method for coordinated docking of the stair climber and the sweeping machine.

[0102] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method for coordinated docking of the stair climber and the sweeping machine.

[0103] A fourth aspect of the present invention provides an electronic device, the electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described method for evaluating the propagation mode of hydraulic fractures.

[0104] The fifth aspect of the present invention provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method for evaluating the propagation mode of hydraulic cracks.

[0105] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. This program is stored in a storage medium and includes several instructions to cause a microcontroller, chip, or processor to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0106] The optional embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the embodiments of the present invention are not limited to the specific details described above. Within the scope of the technical concept of the embodiments of the present invention, various simple modifications can be made to the technical solutions of the embodiments of the present invention, and these simple modifications all fall within the protection scope of the embodiments of the present invention. It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, the embodiments of the present invention will not further describe the various possible combinations.

[0107] Furthermore, various different embodiments of the present invention can be combined in any way, as long as they do not violate the spirit of the embodiments of the present invention, they should also be regarded as the content disclosed by the embodiments of the present invention.

Claims

1. A method for coordinated docking of a stair-climbing frame and a sweeping robot, characterized in that, The method is performed by a stair-climbing frame, and the method includes: Receive docking request signals from the sweeping robot and acquire sensing data to describe the spatial position and attitude angle of the sweeping robot; Based on the sensed data, the relative distance and relative orientation information between the robot vacuum and the robot vacuum are determined, and the docking motion mode is determined according to the relative distance and relative orientation information, and corresponding drive control commands are generated. The docking motion is executed according to the drive control command, and the trajectory of the stair climbing frame is corrected in real time during the docking motion to keep it aligned with the sweeping machine. After docking with the sweeper, a locking operation is performed between the sweeper and the sweeper, and a docking completion command is sent to the sweeper until the predetermined target position is reached, at which point the lock is released.

2. The method for coordinated docking of the stair-climbing frame and the sweeping machine according to claim 1, characterized in that, Receive docking request signals from the robot vacuum cleaner and acquire sensing data describing the robot vacuum cleaner's spatial position and attitude angles, including: Acquire image frame data of the sweeping robot and extract pixel features from the image frame data to describe the boundary shape of the sweeping robot; The displacement vector and attitude angle change of the sweeping robot are calculated based on the feature point matching results between adjacent image frames. Based on the displacement vector and the change in attitude angle, sensing data is generated to describe the spatial position and attitude angle of the sweeping robot.

3. The method for coordinated docking of the stair-climbing frame and the sweeping machine according to claim 2, characterized in that, Acquire image frame data of the sweeping robot and extract pixel features from the image frame data to describe the boundary morphology of the sweeping robot, including: Perform specific visual marker recognition in consecutive frames, and extract pixel positioning points based on the QR code markers or ArUco markers preset on the robot vacuum body; The pose calculation matrix of the marker is calculated based on the distribution relationship of the pixel positioning points in the image frame coordinate system, which is used to determine the azimuth and center position of the sweeping robot in the environmental coordinate system. When the recognition result does not meet the feature extraction accuracy requirements, the depth gradient data in the image frame is called to perform edge feature matching in order to correct the boundary contour position of the sweeping robot.

4. The method for coordinated docking of the stair-climbing frame and the sweeping machine according to claim 1, characterized in that, Based on the sensed data, the relative distance and relative orientation information between the robot vacuum and the robot vacuum are determined, including: The coordinates of the center position of the sweeping robot in the sensor data are analyzed, and a unified coordinate reference system is established by combining the real-time pose parameters of the stair-climbing frame. The relative distance between the sweeper and the sweeper is generated by calculating the horizontal distance and the vertical height difference between the sweeper and the sweeper based on a unified coordinate reference system. The relative azimuth angle is calculated based on the difference between the change in the sweeper's attitude angle and the orientation angle of the stair climber, and is used to describe the angular relationship between the front direction of the stair climber and the center line of the sweeper's body. The relative distance and the relative azimuth angle are combined to form a relative position information data packet between the robot vacuum and the robot vacuum, and then written into the docking motion calculation cache.

5. The method for coordinated docking of the stair-climbing frame and the sweeping machine according to claim 4, characterized in that, Based on the relative distance and relative orientation information, the docking motion mode is determined, and corresponding drive control commands are generated, including: Read the distance difference and angle deviation values ​​from the relative distance and relative orientation information, and compare them with preset thresholds respectively; wherein, The distance difference is obtained based on the distance data used to characterize the spatial distance between the stair climber and the sweeper in a unified reference system in the relative distance, and the angle deviation is obtained based on the azimuth data used to characterize the angle between the forward direction of the stair climber and the centerline direction of the sweeper in the relative orientation information. When the angle deviation value is less than a preset angle threshold and the distance difference value is greater than a preset distance threshold, it is determined to be a reciprocating motion mode, and a first drive control command containing the forward direction, travel speed and correction frequency is generated. When the angle deviation value is not less than the preset angle threshold or the environmental constraint information indicates that the channel is restricted, it is determined to be in the same direction chasing mode, and a second drive control command containing path following parameters, relative speed difference and lateral correction amount is generated. The first drive control instruction or the second drive control instruction is written into the control buffer area to execute the corresponding docking motion.

6. The method for coordinated docking of the stair-climbing frame and the sweeping machine according to claim 1, characterized in that, The docking motion is executed according to the drive control command, and the trajectory of the stair-climbing frame is corrected in real time during the docking motion to maintain alignment with the sweeper, including: Based on the drive control command, the stair climbing frame is controlled to move in a straight line along the docking direction; The sweeper's position offset and azimuth deviation are monitored in real time, and the lateral correction and steering angle correction of the stair climbing frame are calculated in real time based on the offset and azimuth deviation. Based on the lateral correction amount and the steering angle correction amount, a trajectory adjustment command is generated to dynamically correct the travel trajectory of the stair climbing frame; In several consecutive frames of perception output, when the offset and azimuth deviation values ​​are both less than the corresponding preset thresholds, the current direction of travel is maintained and the alignment between the docking platform and the sweeping machine is maintained.

7. The method for coordinated docking of the stair-climbing frame and the sweeping machine according to claim 1, characterized in that, After docking with the robot vacuum, a locking operation is performed between the robot vacuum and the docking completion command is sent to the robot vacuum. The process continues until the predetermined target position is reached, at which point the lock is released, including: Monitor the level signal status output by the optocoupler sensor in the stair climbing frame, and determine the duration for which the level signal remains stable within a preset time during a continuous detection cycle; When the level signal remains valid, a locking execution command is generated and sent to the locking drive component of the docking platform of the climbing frame to control the docking platform to perform a mechanical locking action; After the locking action is completed, the feedback signal of the optocoupler sensor is read to confirm the locking state, and the locking confirmation result is combined with the docking status identifier to generate a docking completion data packet; Send the docking completion data packet to the sweeping machine; After the stair-climbing frame carries the sweeper to the predetermined target location, it generates an unlocking command based on the arrival determination result and controls the docking platform to perform the corresponding unlocking action.

8. A method for coordinated docking of a stair-climbing frame and a sweeping machine, characterized in that, The method is performed by a sweeping machine, and the method includes: Send a docking request signal to the stair climbing frame and obtain identification data to describe the spatial position and attitude angle of the stair climbing frame; Based on the identification data, the relative distance and relative orientation information between the climber and the stair climber are calculated, and the coordination movement mode is determined according to the calculation results to generate corresponding movement control commands. The robot vacuum cleaner performs coordinated movements according to the aforementioned movement control commands, and corrects its movement trajectory in real time during the movement to keep the central axis of the robot vacuum cleaner aligned with the docking and positioning area of ​​the stair climbing frame. When a docking completion signal is received from the stair-climbing frame, it stops moving, retracts the cleaning components, and enters the transport state.

9. The method for coordinated docking of the stair-climbing frame and the sweeping machine according to claim 8, characterized in that, Send a docking request signal to the stair-climbing frame and acquire identification data describing the spatial position and attitude angles of the stair-climbing frame, including: Collect image frame data containing the outline of the stair climbing frame; Visual markings set on the outer surface of the stair climbing frame are identified in the image frame data; The attitude angle parameters of the climbing frame are calculated based on the distribution of the positioning points of the visual markers in the image frame coordinate system, and the spatial position of the climbing frame in the environmental coordinate system is inferred by combining the image depth information. The calculated spatial position and attitude angle parameters of the stair climber are combined to generate identification data describing the spatial position and attitude angle of the stair climber.

10. The method for coordinated docking of the stair-climbing frame and the sweeping machine according to claim 9, characterized in that, Based on the identified data, the relative distance and relative orientation information between the climber and the stair climber are calculated, and the coordinated movement mode is determined according to the calculation results, generating corresponding movement control commands, including: The coordinate parameters and attitude angle parameters of the stair climbing frame in the data are analyzed and identified, and a unified coordinate reference system is established by combining the position parameters of the sweeping machine itself. Calculate the planar distance, vertical height difference, and centerline angle between the sweeping machine and the stair climbing frame under a unified coordinate reference system to form the relative distance and relative orientation information between them; Based on the comparison of relative distance and relative orientation information with preset motion thresholds, when the angle deviation value is less than the preset angle threshold, it is determined to be a straight-line docking mode; when the angle deviation value is not less than the preset angle threshold or the space channel is limited, it is determined to be a following docking mode. Based on the determined coordination pattern, a travel control command is generated that includes the travel direction, speed correction value, and attitude correction amount.

11. The method for coordinated docking of the stair-climbing frame and the sweeping machine according to claim 8, characterized in that, The sweeper executes coordinated movements according to the aforementioned travel control commands, and corrects its travel trajectory in real time during the movement to ensure that the central axis of the sweeper remains aligned with the docking and positioning area of ​​the stair-climbing frame, including: Real-time reception of the stair climbing frame's position offset and azimuth deviation values; The correction distance of the sweeper in the lateral direction is calculated based on the position offset, and the attitude adjustment angle is calculated based on the azimuth deviation value. Based on the correction distance and attitude adjustment angle of the sweeper in the lateral direction, a trajectory correction command is generated. In the next motion cycle, the travel direction and wheel speed difference are adjusted synchronously to correct the sweeper's travel trajectory. When the position offset and azimuth deviation values ​​in the recognition output of multiple consecutive frames are both less than the corresponding preset threshold, the current travel parameters are maintained and the central axis of the sweeper is kept aligned with the docking and positioning area of ​​the stair climber.

12. The method for coordinated docking of the stair-climbing frame and the sweeping machine according to claim 8, characterized in that, Upon receiving a docking completion signal from the stair-climbing frame, the system stops moving, retracts the cleaning components, and enters the transport state, including: The received docking completion signal is parsed, and a state switching command is generated after the signal verification is passed; Execute the state switching command to stop moving and maintain the current position; Drive the sweeping components to perform a retraction action, and turn off the power to the sweeping motor and vacuum fan; After the cleaning components are retracted, update the current operating mode identifier to transport status.

13. A stair-climbing frame, characterized in that, The stair-climbing frame is used to perform the collaborative docking method between the stair-climbing frame and the sweeping machine as described in any one of claims 1-7, and the stair-climbing frame comprises: The main controller is used to control the overall operation of the stair climbing frame and coordinate the signal interaction of other functional modules in the stair climbing frame. The drive unit, electrically connected to the main controller, is used to execute driving and steering movements according to drive control commands; The communication module is connected to the main controller and is used to receive docking requests, exchange status information, and send signals with the sweeping machine. The sensing module is connected to the main controller and is used to collect the spatial position and attitude angle information of the sweeper and output it to the main controller. The docking platform, installed at the front end of the stair-climbing frame and connected to the main controller via signal, is used to support the sweeper and perform locking actions during the docking process; An optocoupler sensor is mounted on the docking platform and connected to the main controller for signal transmission. It is used to output a trigger signal when the sweeper enters the docking position for the main controller to determine the docking.

14. A sweeping machine, characterized in that, The sweeper is used to perform the collaborative docking method between the stair climber and the sweeper as described in any one of claims 8-12, wherein the sweeper comprises: The sweeper control unit is used to control the overall operation of the sweeper and coordinate the signal interaction of other functional units in the sweeper; The travel drive component is electrically connected to the sweeper control unit and is used to perform forward movement, steering, and speed adjustment according to travel control commands. The communication interaction unit is connected to the sweeper control unit and is used to send docking requests to the stair climbing frame, receive docking completion signals, and exchange status information. An environmental recognition unit is signal-connected to the sweeper control unit and is used to collect the spatial position and attitude angle information of the stair climber and output the recognition result to the sweeper control unit. The state switching mechanism is connected to the sweeper control unit and is used to stop the output of the driving component and control the sweeping parts to retract and enter the transport state after receiving the docking completion signal.

15. A stair-climbing frame and sweeping machine collaborative docking system, characterized in that, The system includes the stair-climbing frame as described in claim 13 and the sweeper as described in claim 14.

16. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores instructions that, when executed on a computer, cause the computer to perform the collaborative docking method of the stair climber and sweeper as described in any one of claims 1-12.

17. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the collaborative docking method between the stair climbing frame and the sweeping machine as described in any one of claims 1-12.

18. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the collaborative docking method between the stair climbing frame and the sweeping machine according to any one of claims 1-12.

Images