Dynamic obstacle avoidance method, device and system for roadway cleaning robot

Abstract

The invention relates to a dynamic obstacle avoidance method, device and system for a roadway cleaning robot. The method comprises the steps that environment data of the environment where the cleaning robot is located in a roadway is obtained; performing fusion processing on the environment data to obtain target features, identifying a dynamic obstacle by using the target features, and predicting a motion track of the dynamic obstacle; determining attribute information of the dynamic obstacle, and performing risk assessment on the dynamic obstacle based on the motion trail and the attribute information to obtain a risk level; according to the risk level, an improved A * algorithm is combined with a Bezier curve to carry out dynamic path re-planning, and an obstacle avoidance path is obtained; and controlling the cleaning robot to advance according to the obstacle avoidance path, and controlling the cleaning robot to execute an obstacle avoidance task matched with the risk level. According to the scheme, continuous and automatic roadway cleaning is achieved.

Description

Technical Field

[0001] This disclosure relates to the field of tunnel cleaning technology, and in particular to a dynamic obstacle avoidance method, device and system for a tunnel cleaning robot. Background Technology

[0002] In related technologies, when roadway cleaning robots operate in large deformation roadways in coal mines, they face complex and ever-changing dynamic environments, and existing technologies have significant limitations. Traditional obstacle avoidance solutions mostly rely on single LiDAR data, which can only identify static obstacles. They are slow to identify sudden dynamic targets underground (such as falling coal blocks, temporary work tools, personnel movement, etc.), easily leading to collision risks. At the same time, existing obstacle avoidance logic is disconnected from the robot's cleaning and support operation process. After obstacle avoidance, the work area needs to be repositioned, resulting in long operation interruptions and low efficiency. In addition, traditional methods lack the ability to predict the movement trend of obstacles, and the obstacle avoidance path planning is rigid, easily deviating from the work trajectory or interfering with the operation modules such as the robotic arm and drilling rig, seriously affecting the continuity and safety of the entire automated operation. Summary of the Invention

[0003] To overcome the problems existing in related technologies, this disclosure provides a dynamic obstacle avoidance method, device and system for a roadway cleaning robot.

[0004] According to a first aspect of the present disclosure, a dynamic obstacle avoidance method for a roadway cleaning robot is provided, comprising: Acquire environmental data of the environment in which the cleaning robot is located within the alleyway; the environmental data includes lidar point clouds, camera images, and ultrasonic blind spot data. The environmental data is fused to obtain target features, and the target features are used to identify dynamic obstacles and predict the movement trajectory of the dynamic obstacles. The attribute information of the dynamic obstacle is determined, and a risk assessment is performed on the dynamic obstacle based on the motion trajectory and the attribute information to obtain a risk level; the attribute information includes the category, the distance and relative speed between the dynamic obstacle and the cleaning robot; Based on the risk level, an improved A* algorithm combined with Bézier curves is used for dynamic path replanning to obtain an obstacle avoidance path. Control the cleaning robot to travel along the obstacle avoidance path, and control the cleaning robot to perform obstacle avoidance tasks that match the risk level.

[0005] According to a second aspect of the present disclosure, a dynamic obstacle avoidance device for a roadway cleaning robot is provided, comprising: The acquisition unit is used to acquire environmental data of the environment in which the cleaning robot is located in the alleyway; the environmental data includes lidar point clouds, camera images, and ultrasonic blind spot data. The prediction unit is used to fuse the environmental data to obtain target features, use the target features to identify dynamic obstacles, and predict the movement trajectory of the dynamic obstacles. An assessment unit is used to determine the attribute information of the dynamic obstacle, and to perform a risk assessment on the dynamic obstacle based on the motion trajectory and the attribute information to obtain a risk level; the attribute information includes the category, the distance and relative speed between the dynamic obstacle and the cleaning robot; The planning unit is used to perform dynamic path replanning based on the risk level, using an improved A* algorithm combined with Bézier curves, to obtain an obstacle avoidance path. The control unit is used to control the cleaning robot to travel along the obstacle avoidance path and to control the cleaning robot to perform obstacle avoidance tasks that match the risk level.

[0006] According to a third aspect of the present disclosure, an electronic device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the method as described in any one of the first aspects.

[0007] According to a fourth aspect of the present disclosure, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the method as described in any one of the first aspects.

[0008] According to a fifth aspect of the present disclosure, a computer program product is provided, including a computer program that, when executed by a processor, implements the method as described in any one of the first aspects.

[0009] The technical solution provided by the embodiments of this disclosure can include the following beneficial effects: acquiring environmental data of the environment in which the cleaning robot is located in the alleyway; fusing the environmental data to obtain target features, using the target features to identify dynamic obstacles, and predicting the movement trajectory of the dynamic obstacles; determining the attribute information of the dynamic obstacles, and based on the movement trajectory and attribute information, performing a risk assessment on the dynamic obstacles to obtain a risk level; according to the risk level, using an improved A* algorithm combined with Bézier curves for dynamic path replanning to obtain an obstacle avoidance path; controlling the cleaning robot to move according to the obstacle avoidance path, and controlling the cleaning robot to perform obstacle avoidance tasks matching the risk level. By fusing multi-source perception data from lidar, vision, and ultrasound, and performing real-time risk assessment based on the type and movement trajectory of dynamic obstacles, using improved A* and Bézier curves to plan a smooth obstacle avoidance path, and finally coordinating the robot's movement and robotic arm operation according to the risk level, continuous automated cleaning is achieved, effectively solving the problems of delayed recognition, rigid paths, and operation interruptions in traditional solutions.

[0010] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0011] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0012] Figure 1 This is a flowchart illustrating a dynamic obstacle avoidance method for a roadway cleaning robot according to an exemplary embodiment.

[0013] Figure 2 This is a block diagram illustrating a dynamic obstacle avoidance device for a roadway cleaning robot according to an exemplary embodiment.

[0014] Figure 3 This is a block diagram illustrating an apparatus for a dynamic obstacle avoidance method for a roadway cleaning robot according to an exemplary embodiment. Detailed Implementation

[0015] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Rather, they are merely examples of apparatuses and methods consistent with some aspects of the invention as detailed in the appended claims.

[0016] The terminology used in this disclosure is for the purpose of describing particular embodiments only and is not intended to be limiting of the present disclosure. The singular forms “a” and “the” as used in this disclosure and the appended claims are also intended to include the plural forms, unless the context clearly indicates otherwise.

[0017] It should be understood that although the terms first, second, third, etc., may be used to describe various information in embodiments of this disclosure, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, first information may also be referred to as second information without departing from the scope of embodiments of this disclosure, and similarly, second information may also be referred to as first information. Depending on the context, the words “if” and “suppose” as used herein may be interpreted as “when”, “when”, or “in response to a determination”.

[0018] Furthermore, various forms of processes shown in the embodiments of this disclosure can be used to reorder, add, or delete steps. For example, the steps described in this application can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and no limitation is imposed herein.

[0019] It should be noted that the collection, storage, use, processing, transmission, provision, and disclosure of user personal information involved in the technical solution disclosed herein all comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0020] Figure 1 This is a flowchart illustrating a dynamic obstacle avoidance method for a roadway cleaning robot according to an exemplary embodiment, such as... Figure 1 As shown, it should be noted that the dynamic obstacle avoidance method for the roadway cleaning robot in this embodiment is applied to the dynamic obstacle avoidance device of the roadway cleaning robot. For example... Figure 1 As shown, the method may include the following steps: Step 101: Obtain environmental data of the environment in which the cleaning robot is located in the alley.

[0021] The environmental data includes lidar point clouds, camera images, and ultrasonic blind spot data.

[0022] In one embodiment, a front-mounted LiDAR scanner installed on the cleaning robot can collect real-time 3D point cloud data of the tunnel environment, while a camera simultaneously acquires RGB image data to capture visual features of obstacles. An ultrasonic sensor array (installed around the robot's body and at the end of the robotic arm) enables near-range obstacle detection, compensating for blind spots in LiDAR detection in obstructed areas. All sensing devices can communicate with the central control PLC system via a TP-Link industrial-grade switch, using the NTP protocol to align multi-source data timestamps and ensure data synchronization.

[0023] Step 102: Perform fusion processing on the environmental data to obtain target features, use the target features to identify dynamic obstacles, and predict the movement trajectory of the dynamic obstacles.

[0024] In some embodiments of this application, step 102, which involves fusing environmental data to obtain target features, may specifically include the following steps: Preprocessing is performed on lidar point cloud data, camera image data, and ultrasonic blind spot data respectively; The preprocessed multi-source data is time-aligned based on a unified timestamp. Multi-source data that has been aligned in time are fused to form target features for identification and tracking.

[0025] In one embodiment, LiDAR point cloud data can be Gaussian filtered to remove dust and noise interference, retaining valid obstacle point clouds; point cloud data within the robot's operating height range can be extracted using pass-through filtering, reducing unnecessary calculations. Camera image data undergoes histogram equalization to enhance contrast, and edge detection algorithms can be used to extract obstacle contour features; ultrasonic data can be mean filtered to remove outliers, outputting stable obstacle distance information.

[0026] As an example of a possible implementation, obstacle location data from LiDAR and ultrasound can be dynamically tracked based on the Kalman filter algorithm to predict obstacle trajectories (such as the falling path of a coal block or the direction of personnel movement). A lightweight YOLOv8 deep learning model is introduced, fusing point cloud geometric features with image visual features to achieve dynamic obstacle classification and recognition (distinguishing between personnel, coal blocks, tools, etc.).

[0027] In some embodiments of this application, step 102, which involves identifying dynamic obstacles using target features and predicting the trajectory of the dynamic obstacles, may specifically include the following steps: When the dynamic obstacle is a coal block or a tool-like inanimate obstacle, the Kalman filter algorithm is used to fuse its position and velocity information to predict its future motion trajectory in the time series. When dynamic obstacles are the target of people, the behavioral posture features of the dynamic obstacles are extracted, and a temporal deep learning model is used to identify their behavioral intentions based on the behavioral posture features, and the movement trajectory of the dynamic obstacles is predicted based on the behavioral intentions.

[0028] In one embodiment, when the dynamic obstacle is a coal block or a tool-like inanimate obstacle, a Kalman filter algorithm is used to accurately predict its motion trajectory. Specifically, a motion model with the obstacle's position and velocity as state vectors is constructed. Multi-source data fusion is performed based on continuous frame point cloud data provided by lidar and real-time distance information collected by ultrasonic sensors. The Kalman filter is recursively calculated in a "prediction-update" cycle. In the prediction phase, the state of the obstacle at the next moment is deduced based on Newtonian mechanics models. In the update phase, the prediction results are optimally corrected using the latest observation data, thereby effectively suppressing sensor noise interference and significantly improving the stability and reliability of trajectory prediction. This process can predict the motion trajectory of inanimate obstacles such as coal blocks within a 1-3 second time series, including continuous changes in position, velocity, and acceleration, providing high-precision forward-looking input for subsequent risk assessment and path planning.

[0029] In another embodiment, when the dynamic obstacle is a person, proactive obstacle avoidance is achieved through behavior recognition and intent prediction. Specifically, the coordinates of the main joints of the human body can be extracted in real time from a continuous image sequence captured by a camera using a deep learning algorithm to construct a dynamically changing posture feature vector. These temporal posture features are then input into a lightweight long short-term memory (LSTM) network. This network analyzes the motion patterns and trajectory features of the joints to identify typical behavioral intentions such as "normal walking," "stopping to observe," "bending over to work," and "emergency avoidance." Based on the identified behavioral intentions, different motion prediction models are established. For example, when "emergency avoidance" is identified, the prediction model increases the radius of the possible movement range of the person and prioritizes the tangential direction, thereby generating the movement trajectory with the highest probability in the next 3 seconds. This improves the accuracy of the person's trajectory prediction, provides the robot with a crucial decision-making buffer time, and achieves a safety upgrade from passive obstacle avoidance to proactive prediction.

[0030] Step 103: Determine the attribute information of the dynamic obstacle. Based on the motion trajectory and attribute information, conduct a risk assessment of the dynamic obstacle to obtain the risk level.

[0031] The attribute information includes the category, the distance and relative speed between the dynamic obstacle and the cleaning robot.

[0032] In one embodiment, a risk assessment model can be constructed, using the real-time distance, relative speed, motion trend, and category of obstacles and the robot as assessment indicators, and employing the analytic hierarchy process (AHP) to calculate the risk level (high, medium, and low). For example, risk thresholds can be set as follows: obstacles involving people with a distance ≤ 5m are considered high risk; obstacles involving coal blocks with a distance ≤ 2m and a speed ≥ 0.5m / s are considered medium risk; and obstacles involving tools with a distance ≤ 1m are considered low risk.

[0033] In this embodiment, a multi-dimensional risk assessment model is constructed to comprehensively evaluate the safety of dynamic obstacles. Using obstacle category as the core distinguishing element, and combining its real-time distance, relative speed, and predicted motion trajectory characteristics, a hierarchical analysis method is employed for quantitative assessment: For personnel, a high-risk trigger threshold of 5 meters is set, focusing on analyzing the probability of intersection between their motion trajectory and the robot's operating area; for loose objects such as coal, the distance (e.g., within 2 meters), falling speed (e.g., ≥0.5 m / s), and acceleration characteristics (e.g., automatically upgrading the risk level when >0.2 m / s²) are comprehensively considered; for static obstacles such as tools, the judgment is mainly based on the safe distance boundary (e.g., 1 meter). Simultaneously, an environmental adaptive mechanism is introduced: when the dust concentration in the tunnel exceeds 20 mg / m³ or the light intensity is below 50 lux, the judgment thresholds for each risk level are automatically tightened. Finally, three levels of risk labels (high, medium, and low) are output. Obstacles identified as high-risk trigger the system's priority response mechanism, achieving a multi-dimensional risk assessment that integrates motion trends and environmental factors, moving beyond simple distance judgment, thus reducing the false alarm rate.

[0034] In some embodiments of this application, step 103 may specifically include the following steps: Based on pre-set risk thresholds for different categories of obstacles, the risk thresholds corresponding to the categories of dynamic obstacles are determined; the risk thresholds include distance thresholds and speed thresholds. The distance is compared with a distance threshold to obtain a first comparison result, and the relative speed is compared with a speed threshold to obtain a second comparison result. The risk level is determined based on the first comparison result and the second comparison result.

[0035] In one embodiment, based on the identified dynamic obstacle category, the corresponding risk threshold parameter set is retrieved from a pre-built knowledge base. For example, personnel are associated with a 5-meter distance threshold, coal blocks with a 2-meter distance threshold and a 0.5 m / s speed threshold, and tools with a 1-meter distance threshold. Then, the actual distance and relative speed of the obstacle calculated in real-time are compared in parallel with the corresponding thresholds: distance comparison generates a first judgment result indicating whether the obstacle has entered a warning zone, and speed comparison generates a second judgment result indicating whether the obstacle exhibits a dangerous movement trend. Finally, the two comparison results are combined using logic and / or rules (e.g., for coal blocks, both distance ≤ 2 meters and speed ≥ 0.5 m / s must be met simultaneously to trigger a medium risk level), generating high, medium, and low risk levels. This multi-parameter threshold joint judgment method retains the interpretability advantages of a rule-based system while controlling the risk assessment delay to within 10 milliseconds through a structured judgment process, providing crucial decision-making basis for subsequent real-time path planning.

[0036] Step 104: Based on the risk level, use the improved A* algorithm combined with Bézier curves to perform dynamic path replanning to obtain the obstacle avoidance path.

[0037] In one embodiment, differentiated safety boundaries are configured according to risk levels (high risk ≥ 1.5 meters, medium-low risk ≥ 0.8 meters). Then, an improved A* algorithm is used to quickly search for key points of feasible paths under the constraints of the work area. Subsequently, the key point sequence is smoothed and optimized using a third-order Bézier curve. The number of control points can be dynamically adjusted according to the obstacle distribution density (e.g., 2 more control points are added when the density is > 5 / 10㎡). At the same time, a multi-objective optimization function is introduced to dynamically balance the three indicators of path length, smoothness, and work continuity. When encountering high-risk obstacles, the obstacle avoidance time is shortened first, while the work continuity is emphasized when facing medium-low risks.

[0038] In some embodiments of this application, step 104 may specifically include the following steps: Using the current position of the cleaning robot and the target work area as constraints, and based on the safe distance corresponding to the risk level, the key point sequence of the obstacle avoidance path is searched and generated by the improved A* algorithm; A third-order Bézier curve is used to smoothly fit the key point sequence to generate a final obstacle avoidance path with continuous curvature; the number of control points of the Bézier curve is dynamically adjusted according to the obstacle density of the area traversed by the key point sequence.

[0039] In one embodiment, the robot's real-time positioning coordinates and the diagonal points (P1, P2) of the preset work area are used as path start and end constraints. Based on the real-time risk level assessment, the corresponding safety distance parameters are called. A buffer space of ≥1.5 meters is reserved for high-risk obstacles, and an operational margin of ≥0.8 meters is reserved for medium- and low-risk obstacles. Subsequently, the improved A* algorithm is used to search in the gridded environment model. Its heuristic function introduces an additional safety distance weight factor, so that the generated path key point sequence not only satisfies the shortest path principle, but also ensures that each path point maintains a risk-appropriate safety distance from the obstacle. The final output key point sequence forms a basic framework for obstacle avoidance path that takes into account both safety and efficiency, providing an optimization basis for subsequent Bézier curve smoothing processing.

[0040] In one embodiment, two adjustable control points can be introduced to construct a continuous curvature path. A control point number adaptive mechanism is specially designed: when the obstacle density in the area traversed by the key point sequence is less than 3 per 10 meters, a standard 4-control-point configuration is adopted; when the obstacle density reaches 5 per 10 meters, the number of control points is automatically increased to 6, thereby improving the degree of freedom of the curve and enabling fine-grained navigation in complex obstacle environments.

[0041] Step 105: Control the cleaning robot to move along the obstacle avoidance path and control the cleaning robot to perform obstacle avoidance tasks that match the risk level.

[0042] In this embodiment, the collaborative control of the walking system and the working system can be achieved through a central control PLC. Upon receiving the planned path, the navigation walking module tracks the Bézier curve path and simultaneously activates differentiated obstacle avoidance strategies based on the risk level: when encountering high-risk obstacles, the walking system immediately brakes, the robotic arm performs an emergency reset (retracting the attachments to a vertical state), and the drill rig stops drilling; when facing medium-risk obstacles, the walking system reduces its speed to 50% of its original speed and detours along a smooth path, while the robotic arm synchronously adjusts its joint angles to avoid motion interference with the obstacle; when handling low-risk obstacles, the walking system maintains its normal operating speed and performs a slight path deviation, while the robotic arm continues the cleaning operation. This hierarchical collaborative control can shorten the operation interruption time during obstacle avoidance, realizing an upgrade from simple obstacle avoidance to integrated intelligent control of "walking-operation," effectively ensuring the continuity and safety of roadway cleaning operations.

[0043] In some embodiments of this application, step 105 may specifically include the following steps: Based on the risk level, control the robotic arm of the cleaning robot to perform the corresponding operation. Specifically, when the risk level is high, the robotic arm is controlled to perform an emergency reset action; when the risk level is medium, the robotic arm is controlled to adjust its working posture to avoid interference with dynamic obstacles.

[0044] As an example, for high-risk obstacles: an emergency stop command is immediately sent to the navigation module, the robotic arm module performs an emergency reset (retracting the working attachments and maintaining a vertical position), and the drilling rig module stops drilling operations. For medium-risk obstacles: a deceleration (speed reduced to 50% of the original speed) and path adjustment command is sent to the navigation module, and the robotic arm module fine-tunes its working posture to avoid interference with the obstacle. For low-risk obstacles: only a path deviation command is sent to the navigation module, while other operating modules continue to operate normally.

[0045] In one embodiment, after obstacle avoidance is completed, the navigation and walking module of the cleaning robot can be controlled to reposition itself using a rear-mounted LiDAR and quickly return to the original working path; the central control PLC system simultaneously notifies the robotic arm and drilling rig module to resume operation, seamlessly connecting the interrupted process based on historical operation data.

[0046] According to the dynamic obstacle avoidance method for a roadway cleaning robot proposed in this disclosure, environmental data of the environment in which the cleaning robot is located in the roadway is acquired; the environmental data is fused to obtain target features; dynamic obstacles are identified using the target features, and the movement trajectory of the dynamic obstacles is predicted; the attribute information of the dynamic obstacles is determined; based on the movement trajectory and attribute information, a risk assessment is performed on the dynamic obstacles to obtain a risk level; according to the risk level, a modified A* algorithm combined with Bézier curves is used for dynamic path replanning to obtain an obstacle avoidance path; the cleaning robot is controlled to move according to the obstacle avoidance path, and the cleaning robot is controlled to perform obstacle avoidance tasks matching the risk level. By fusing multi-source perception data from lidar, vision, and ultrasound, and performing real-time risk assessment based on the type and movement trajectory of dynamic obstacles, a smooth obstacle avoidance path is planned using modified A* and Bézier curves, and finally, the robot's movement and robotic arm operation are coordinated according to the risk level, achieving continuous automated cleaning and effectively solving the problems of delayed recognition, rigid paths, and operation interruptions in traditional solutions.

[0047] Figure 2 This is a block diagram illustrating a dynamic obstacle avoidance device for a roadway cleaning robot according to an exemplary embodiment. (Refer to...) Figure 2 The device includes an acquisition unit 201, a prediction unit 202, an evaluation unit 203, a planning unit 204, and a control unit 205.

[0048] The acquisition unit 201 is used to acquire environmental data of the environment in which the cleaning robot is located in the alleyway; the environmental data includes lidar point cloud, camera images and ultrasonic blind spot data. The prediction unit 202 is used to fuse environmental data to obtain target features, use the target features to identify dynamic obstacles, and predict the movement trajectory of the dynamic obstacles. The evaluation unit 203 is used to determine the attribute information of dynamic obstacles, and to perform risk assessment on dynamic obstacles based on motion trajectory and attribute information to obtain risk level; attribute information includes category, distance and relative speed between dynamic obstacle and cleaning robot; Planning unit 204 is used to perform dynamic path replanning based on the risk level, using an improved A* algorithm combined with Bézier curves, to obtain an obstacle avoidance path; The control unit 205 is used to control the cleaning robot to travel along the obstacle avoidance path and to control the cleaning robot to perform obstacle avoidance tasks that match the risk level.

[0049] In this embodiment of the application, the evaluation unit 203 can be specifically used for: Based on pre-set risk thresholds for different categories of obstacles, the risk thresholds corresponding to the categories of dynamic obstacles are determined; the risk thresholds include distance thresholds and speed thresholds. The distance is compared with a distance threshold to obtain a first comparison result, and the relative speed is compared with a speed threshold to obtain a second comparison result. The risk level is determined based on the first comparison result and the second comparison result.

[0050] In this embodiment of the application, the risk levels include high risk, medium risk, and low risk, and the control unit 205 can be specifically used for: Based on the risk level, control the robotic arm of the cleaning robot to perform the corresponding operation. Specifically, when the risk level is high, the robotic arm is controlled to perform an emergency reset action; when the risk level is medium, the robotic arm is controlled to adjust its working posture to avoid interference with dynamic obstacles.

[0051] In this embodiment of the application, the prediction unit 202 can specifically be used for: When the dynamic obstacle is a coal block or a tool-like inanimate obstacle, the Kalman filter algorithm is used to fuse its position and velocity information to predict its future motion trajectory in the time series. When dynamic obstacles are the target of people, the behavioral posture features of the dynamic obstacles are extracted, and a temporal deep learning model is used to identify their behavioral intentions based on the behavioral posture features, and the movement trajectory of the dynamic obstacles is predicted based on the behavioral intentions.

[0052] In this embodiment of the application, the planning unit 204 can be specifically used for: Using the current position of the cleaning robot and the target work area as constraints, and based on the safe distance corresponding to the risk level, the key point sequence of the obstacle avoidance path is searched and generated by the improved A* algorithm; A third-order Bézier curve is used to smoothly fit the key point sequence to generate a final obstacle avoidance path with continuous curvature; the number of control points of the Bézier curve is dynamically adjusted according to the obstacle density of the area traversed by the key point sequence.

[0053] In this embodiment of the application, the prediction unit 202 can specifically be used for: Preprocessing is performed on lidar point cloud data, camera image data, and ultrasonic blind spot data respectively; The preprocessed multi-source data is time-aligned based on a unified timestamp. Multi-source data that has been aligned in time are fused to form target features for identification and tracking.

[0054] Regarding the apparatus in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0055] According to the dynamic obstacle avoidance device for a roadway cleaning robot proposed in this disclosure, environmental data of the environment in which the cleaning robot is located in the roadway is acquired; the environmental data is fused to obtain target features; dynamic obstacles are identified using the target features, and the movement trajectory of the dynamic obstacles is predicted; the attribute information of the dynamic obstacles is determined; based on the movement trajectory and attribute information, a risk assessment is performed on the dynamic obstacles to obtain a risk level; according to the risk level, a dynamic path replanning is performed using an improved A* algorithm combined with Bézier curves to obtain an obstacle avoidance path; the cleaning robot is controlled to move according to the obstacle avoidance path, and the cleaning robot is controlled to perform obstacle avoidance tasks matching the risk level. By fusing multi-source perception data from lidar, vision, and ultrasound, and performing real-time risk assessment based on the type and movement trajectory of dynamic obstacles, a smooth obstacle avoidance path is planned using improved A* and Bézier curves, and finally, the robot's movement and robotic arm operation are coordinated according to the risk level, achieving continuous automated cleaning and effectively solving the problems of delayed recognition, rigid paths, and operation interruptions in traditional solutions.

[0056] Figure 3 This is a block diagram illustrating an apparatus for a dynamic obstacle avoidance method for a roadway cleaning robot according to an exemplary embodiment. For example, apparatus 300 may be an electronic device, such as a mobile phone, computer, digital broadcasting terminal, messaging device, personal digital assistant, etc.

[0057] Reference Figure 3 The device 300 may include one or more of the following components: a processing component 302, a memory 304, a power component 306, a multimedia component 308, an audio component 310, an input / output (I / O) interface 312, a sensor component 314, and a communication component 316.

[0058] Processing component 302 typically controls the overall operation of device 300, such as operations associated with display, telephone calls, data communication, camera operation, and recording. Processing component 302 may include one or more processors 320 to execute instructions to perform all or part of the steps of the methods described above. Furthermore, processing component 302 may include one or more modules to facilitate interaction between processing component 302 and other components. For example, processing component 302 may include a multimedia module to facilitate interaction between multimedia component 308 and processing component 302.

[0059] Memory 304 is configured to store various types of data to support the operation of device 300. Examples of such data include instructions for any application or method operating on device 300, contact data, phonebook data, messages, pictures, videos, etc. Memory 304 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0060] The power supply component 306 provides power to the various components of the device 300. The power supply component 306 may include a power management system, one or more power sources, and other components associated with generating, managing, and distributing power to the device 300.

[0061] Multimedia component 308 includes a screen that provides an output interface between the device 300 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touchscreen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touches, swipes, and gestures on the touch panel. The touch sensors may sense not only the boundaries of the touch or swipe action but also the duration and pressure associated with the touch or swipe operation. In some embodiments, multimedia component 308 includes a front-facing camera and / or a rear-facing camera. When the device 300 is in an operating mode, such as a shooting mode or a video mode, the front-facing camera and / or the rear-facing camera may receive external multimedia data. Each front-facing camera and rear-facing camera may be a fixed optical lens system or have focal length and optical zoom capabilities.

[0062] Audio component 310 is configured to output and / or input audio signals. For example, audio component 310 includes a microphone (MIC) configured to receive external audio signals when device 300 is in an operating mode, such as call mode, recording mode, and voice recognition mode. The received audio signals may be further stored in memory 304 or transmitted via communication component 316. In some embodiments, audio component 310 also includes a speaker for outputting audio signals.

[0063] I / O interface 312 provides an interface between processing component 302 and peripheral interface modules, such as keyboards, click wheels, buttons, etc. These buttons may include, but are not limited to, home buttons, volume buttons, power buttons, and lock buttons.

[0064] Sensor assembly 314 includes one or more sensors for providing status assessments of various aspects of device 300. For example, sensor assembly 314 may detect the on / off state of device 300, the relative positioning of components such as the display and keypad of device 300, changes in the position of device 300 or a component of device 300, the presence or absence of user contact with device 300, the orientation or acceleration / deceleration of device 300, and temperature changes of device 300. Sensor assembly 314 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact. Sensor assembly 314 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, sensor assembly 314 may also include an accelerometer, a gyroscope, a magnetometer, a pressure sensor, or a temperature sensor.

[0065] Communication component 316 is configured to facilitate wired or wireless communication between device 300 and other devices. Device 300 can access wireless networks based on communication standards, such as WiFi, 2G, or 3G, or combinations thereof. In one exemplary embodiment, communication component 316 receives broadcast signals or broadcast-related information from an external broadcast management system via a broadcast channel. In one exemplary embodiment, communication component 316 also includes a near-field communication (NFC) module to facilitate short-range communication. For example, the NFC module may be implemented based on radio frequency identification (RFID) technology, Infrared Data Association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology, and other technologies.

[0066] In an exemplary embodiment, the apparatus 300 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the methods described above.

[0067] In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions is also provided, such as a memory 304 including instructions, which can be executed by a processor 320 of the device 300 to perform the above-described method. For example, the non-transitory computer-readable storage medium may be a ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage device, etc.

[0068] In an exemplary embodiment, a computer program product is also provided, including a computer program that implements the above-described method when executed by the processor 320 of the device 300.

[0069] Other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This disclosure is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of the invention are indicated by the following claims.

[0070] It should be understood that the present invention is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A dynamic obstacle avoidance method for a roadway cleaning robot, characterized in that, include: Acquire environmental data of the environment in which the cleaning robot is located within the alleyway; the environmental data includes lidar point clouds, camera images, and ultrasonic blind spot data. The environmental data is fused to obtain target features, and the target features are used to identify dynamic obstacles and predict the movement trajectory of the dynamic obstacles. The attribute information of the dynamic obstacle is determined, and a risk assessment is performed on the dynamic obstacle based on the motion trajectory and the attribute information to obtain a risk level; The attribute information includes the category, the distance and relative speed between the dynamic obstacle and the cleaning robot; Based on the risk level, an improved A* algorithm combined with Bézier curves is used for dynamic path replanning to obtain an obstacle avoidance path. Control the cleaning robot to travel along the obstacle avoidance path, and control the cleaning robot to perform obstacle avoidance tasks that match the risk level.

2. The dynamic obstacle avoidance method for a roadway cleaning robot according to claim 1, characterized in that, The risk assessment of the dynamic obstacle based on the motion trajectory and the attribute information to obtain a risk level includes: Based on pre-set risk thresholds for different categories of obstacles, the risk thresholds corresponding to the categories of the dynamic obstacles are determined; the risk thresholds include distance thresholds and speed thresholds. The distance is compared with a distance threshold to obtain a first comparison result, and the relative speed is compared with a speed threshold to obtain a second comparison result; The risk level is determined based on the first comparison result and the second comparison result.

3. The dynamic obstacle avoidance method for the roadway cleaning robot according to claim 1, characterized in that, The risk levels include high risk, medium risk, and low risk; controlling the cleaning robot to perform obstacle avoidance tasks matching the risk levels includes: Based on the risk level, the robotic arm of the cleaning robot is controlled to perform the corresponding operation. Specifically, when the risk level is high, the robotic arm is controlled to perform an emergency reset action; when the risk level is medium, the robotic arm is controlled to adjust its working posture to avoid interference with the dynamic obstacle.

4. The dynamic obstacle avoidance method for the roadway cleaning robot according to claim 1, characterized in that, The step of identifying dynamic obstacles using the target features and predicting the trajectory of the dynamic obstacles includes: When the dynamic obstacle is a coal block or a tool-like inanimate obstacle, the Kalman filter algorithm is used to fuse its position and velocity information to predict its future motion trajectory in the time series. When the dynamic obstacle is a person target, the behavioral posture features of the dynamic obstacle are extracted, and a temporal deep learning model is used to identify its behavioral intention based on the behavioral posture features, and the movement trajectory of the dynamic obstacle is predicted based on the behavioral intention.

5. The dynamic obstacle avoidance method for the roadway cleaning robot according to claim 1, characterized in that, The step of using an improved A* algorithm combined with Bézier curves for dynamic path replanning based on the risk level to obtain an obstacle avoidance path includes: Using the current position of the cleaning robot and the target work area as constraints, and based on the safe distance corresponding to the risk level, the improved A* algorithm is used to search for and generate a sequence of key points for the obstacle avoidance path. The key point sequence is smoothly fitted using a third-order Bézier curve to generate a final obstacle avoidance path with continuous curvature; wherein, the number of control points of the Bézier curve is dynamically adjusted according to the obstacle density of the area traversed by the key point sequence.

6. The dynamic obstacle avoidance method for a roadway cleaning robot according to claim 1, characterized in that, The process of fusing the environmental data to obtain target features includes: The lidar point cloud data, camera image data, and ultrasonic blind spot data are preprocessed respectively. The preprocessed multi-source data is time-aligned based on a unified timestamp. The time-aligned multi-source data is fused to form the target features used for identification and tracking.

7. A dynamic obstacle avoidance device for a roadway cleaning robot, characterized in that, include: The acquisition unit is used to acquire environmental data of the environment in which the cleaning robot is located in the alleyway; the environmental data includes lidar point clouds, camera images, and ultrasonic blind spot data. The prediction unit is used to fuse the environmental data to obtain target features, use the target features to identify dynamic obstacles, and predict the movement trajectory of the dynamic obstacles. An assessment unit is used to determine the attribute information of the dynamic obstacle, and to perform a risk assessment on the dynamic obstacle based on the motion trajectory and the attribute information to obtain a risk level; The attribute information includes the category, the distance and relative speed between the dynamic obstacle and the cleaning robot; The planning unit is used to perform dynamic path replanning based on the risk level, using an improved A* algorithm combined with Bézier curves, to obtain an obstacle avoidance path. The control unit is used to control the cleaning robot to travel along the obstacle avoidance path and to control the cleaning robot to perform obstacle avoidance tasks that match the risk level.

8. An electronic device, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the method as described in any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, The computer program, when executed by a processor, implements the method as described in any one of claims 1 to 6.

Images