Dynamic adjustment traffic management robot displacement method and traffic management robot

Abstract

The application discloses a dynamic adjustment traffic management robot displacement method and a traffic management robot, and belongs to the technical field of robot motion control. The method comprises the following steps: acquiring a target scene image collected by a traffic management robot; iteratively performing discrete displacement adjustment on the traffic management robot according to the positional relationship between a target marker in the target scene image and a preset reference region until the positional relationship between the target marker and the preset reference region meets a specified condition. The above scheme can reduce the influence of single motion error on the final adjustment result, and improve the stability and controllability of the traffic management robot when performing small-range position adjustment in a narrow space.

Description

Technical Field

[0001] This application relates to the field of robot motion control technology, and more specifically, to a method for dynamically adjusting the displacement of a traffic management robot and a traffic management robot. Background Technology

[0002] In traffic management applications, humanoid bipedal robots with humanoid structures are used as traffic management robots, which can perform tasks such as traffic command, information interaction, and on-site auxiliary management in complex environments such as road intersections.

[0003] However, due to the complex mechanical structure of humanoid bipedal robots, their movement is easily affected by factors such as joint clearances, ground elasticity, changes in the robot's center of gravity, and bipedal gait disturbances, resulting in deviations between the robot's actual trajectory and the preset trajectory. Especially in applications such as traffic management where robots need to accurately face target areas, after completing a movement or turning action, the robot usually requires further small-scale position and posture adjustments to eliminate accumulated errors generated during movement. This makes traditional open-loop gait planning or coordinate-based closed-loop control prone to overshoot and repeated oscillations when performing precise small-scale position adjustments in confined spaces (such as charging station docking, control panel alignment, and traffic control position adjustments), failing to achieve rapid and stable single-incremental displacement. Summary of the Invention

[0004] Therefore, it is necessary to provide a method for dynamically adjusting the displacement of traffic management robots and a traffic management robot in response to the above problems.

[0005] This application provides a method for dynamically adjusting the displacement of a traffic management robot, which is applied to a traffic management robot deployed in a target scene. The target scene is provided with a target marker. The method includes: acquiring an image of the target scene collected by the traffic management robot; and iteratively adjusting the discrete displacement of the traffic management robot according to the positional relationship between the target marker and the preset reference area in the target scene image until the positional relationship between the target marker and the preset reference area meets the specified conditions. Specifically, based on the positional relationship between the target marker in the target scene image and the preset reference area, the traffic management robot is iteratively adjusted for discrete displacement. This includes: when the positional relationship does not meet the specified conditions, determining a discrete displacement adjustment command based on the positional relationship; the discrete displacement adjustment command is used to control the traffic management robot to complete a discrete displacement action with a preset displacement amount; and the traffic management robot is adjusted for discrete displacement based on the discrete displacement adjustment command.

[0006] In one optional implementation, before determining the positional relationship between the target marker in the target scene image and the preset reference area, the method further includes: detecting the target marker in the target scene image to determine the image region of the target marker in the target scene image; and determining the positional state of the target marker relative to the preset reference area based on the positional relationship between the image region and the preset reference area.

[0007] In one optional implementation, determining discrete displacement adjustment commands based on positional relationships includes: determining that the positional relationship satisfies specified conditions when the position state is a first position state located within a preset reference area; determining a backward discrete displacement adjustment command when the position state is a second position state located above the preset reference area; determining a forward discrete displacement adjustment command when the position state is a third position state located below the preset reference area; determining a left turn discrete displacement adjustment command when the position state is a fourth position state located to the left of the preset reference area; and determining a right turn discrete displacement adjustment command when the position state is a fifth position state located to the right of the preset reference area. The backward, forward, left, and right discrete displacement adjustment commands are all used to control the traffic management robot to complete a discrete displacement action with a preset displacement amount.

[0008] In one optional implementation, the traffic management robot is subjected to discrete displacement adjustment according to the discrete displacement adjustment command, including: when the discrete displacement adjustment command is a left turn discrete displacement adjustment command or a right turn discrete displacement adjustment command: acquiring the posture detection data corresponding to the traffic management robot; determining the actual rotation angle of the traffic management robot according to the posture detection data; and adjusting the rotation action of the traffic management robot according to the deviation between the actual rotation angle and the target rotation angle.

[0009] In one alternative implementation, the target markers include visual markers set in the target scene; Target marker detection is performed on the target scene image to determine the image region of the target marker in the target scene image, including: converting the target scene image to a preset color space; segmenting the target scene image into regions based on the color threshold corresponding to the visual marker to obtain candidate marker regions; and determining the image region corresponding to the target marker based on the region area, shape features, and position features of the candidate marker regions.

[0010] In one optional implementation, after performing discrete displacement adjustment on the traffic management robot according to the discrete displacement adjustment command, the method further includes: controlling the traffic management robot to stop moving and maintaining the current posture for a preset time to allow the traffic management robot to achieve posture stabilization; after the traffic management robot achieves posture stabilization, re-acquiring the target scene image; redetermining the positional relationship of the target marker relative to the preset reference area based on the re-acquiring target scene image; and determining whether to continue discrete displacement adjustment based on the redetermined positional relationship of the target marker relative to the preset reference area.

[0011] In one optional implementation, determining whether to continue discrete displacement adjustment based on the redefined positional relationship of the target marker relative to the preset reference area includes: stopping the displacement adjustment of the traffic management robot when the redefined positional relationship meets the specified conditions; and determining the next discrete displacement adjustment instruction based on the redefined positional relationship when the redefined positional relationship does not meet the specified conditions.

[0012] In one optional implementation, the preset displacement is determined by the following steps: obtaining the positional deviation between the target marker and the preset reference area; and determining the corresponding preset displacement based on the deviation range to which the positional deviation belongs.

[0013] In another aspect, this application also provides a traffic management robot, comprising: The image acquisition module is used to acquire images of the target scene; The processing module is used to determine the position state of the target marker relative to the preset reference area based on the target scene image, and to determine the corresponding discrete pose adjustment command based on the position state. The discrete pose adjustment command is used to instruct the traffic management robot to perform a pose adjustment action with a preset adjustment step size and / or a preset adjustment angle. The motion control module is used to control the traffic management robot to execute the pose adjustment action corresponding to the discrete pose adjustment command, so as to complete a discrete pose adjustment. The processing module is also used to control the image acquisition module to reacquire the target scene image after completing a discrete pose adjustment, and to redetermine the position of the target marker relative to the preset reference area based on the reacquired target scene image, and to determine whether the redetermined position meets the preset adjustment conditions. If the preset adjustment conditions are not met, the processing module will redetermine the corresponding discrete pose adjustment command and control the motion control module to execute the next discrete pose adjustment until the preset adjustment conditions are met.

[0014] In another aspect, this application also provides a computer-readable storage medium storing computer instructions, which are used to cause a computer to execute the above-described method for dynamically adjusting the displacement of a traffic management robot.

[0015] Compared to existing technologies, the advantages of this application are as follows: This application acquires target scene images collected by a traffic management robot, and determines whether the current target position meets the adjustment requirements based on the positional relationship between the target marker in the target scene image and a preset reference area. When the positional relationship does not meet the specified conditions, a corresponding discrete displacement adjustment command is determined based on the positional relationship, and the traffic management robot is controlled to complete a discrete displacement action with a preset displacement amount using the discrete displacement adjustment command. By converting the positional deviation between the target marker and the preset reference area into a corresponding discrete displacement adjustment command, the robot can perform targeted incremental displacement adjustments based on the actual positional deviation, without relying on a single gait planning to directly reach the target position. Simultaneously, after completing a discrete displacement action, the robot continues to make the next adjustment judgment based on the positional relationship between the target marker and the preset reference area, enabling the robot to gradually correct the positional deviation through multiple discrete displacement adjustments. Therefore, in cases where bipedal robots have joint gaps, ground elasticity, and center of gravity changes leading to single gait displacement errors, the impact of single motion errors on the final adjustment result can be reduced, improving the stability and controllability of the traffic management robot when performing small-range position adjustments in confined spaces. Attached Figure Description

[0016] Figure 1 A flowchart of a method for dynamically adjusting the displacement of a traffic management robot provided in an embodiment of this application; Figure 2 A flowchart of a method for dynamically adjusting the displacement of a traffic management robot, provided in another embodiment of this application. Detailed Implementation

[0017] To make the above-mentioned objectives, features, and advantages of this application more apparent and understandable, the specific embodiments of this application are described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to provide a thorough understanding of this application. However, this application can be implemented in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of this application. Therefore, this application is not limited to the specific embodiments disclosed below.

[0018] In traffic management applications, traffic management robots typically need to move and adjust their posture in different directions based on road conditions and traffic control requirements to accurately face the corresponding road area and perform traffic management tasks. For example, when a traffic management robot needs to switch from its current direction of traffic control to another intersection, it needs to perform corresponding movement, turning, and position calibration operations to ensure that the robot can stably remain in the target working position. However, due to the complex mechanical structure of humanoid bipedal robots, their movement is easily affected by factors such as joint clearances, ground elasticity, changes in the robot's center of gravity, and bipedal gait disturbances, resulting in deviations between the robot's actual movement trajectory and the preset trajectory.

[0019] Therefore, to address the aforementioned issues, this application provides a method for dynamically adjusting the displacement of a traffic management robot, applied to a traffic management robot deployed in a target scene where target markers are present. By acquiring an image of the target scene and adjusting the traffic management robot's displacement iteratively based on the positional relationship of the target markers relative to a preset reference area, the traffic management robot can gradually approach the target position through multiple small-range displacement movements, thereby improving the accuracy of the traffic management robot when making fine adjustments near the target position.

[0020] Figure 1 A flowchart of a method for dynamically adjusting the displacement of a traffic management robot according to an embodiment of this application is shown. (Refer to...) Figure 1 As shown, the method includes the following steps: Step 101: Obtain the target scene image collected by the traffic management robot.

[0021] Specifically, when performing displacement adjustment tasks, traffic management robots acquire images of the target scene using image acquisition devices mounted on them.

[0022] The target scene image includes at least the target marker and image information used to determine the positional relationship of the target marker relative to a preset reference area. The target scene image can be a single frame image acquired by the robot in real time, or any image frame from a continuously acquired video image; this application does not limit this.

[0023] Step 102: Based on the positional relationship between the target marker in the target scene image and the preset reference area, iteratively perform discrete displacement adjustments on the traffic management robot until the positional relationship between the target marker and the preset reference area meets the specified conditions.

[0024] Specifically, after acquiring the target scene image, the positional relationship of the target marker relative to the preset reference area is determined based on the target scene image, and it is determined whether the current traffic management robot needs to make displacement adjustments based on the positional relationship.

[0025] Optionally, the discrete displacement adjustment of the traffic management robot can be achieved in the following way: when the positional relationship does not meet the specified conditions, a discrete displacement adjustment command is determined according to the positional relationship; and the discrete displacement adjustment of the traffic management robot is performed according to the discrete displacement adjustment command.

[0026] When the positional relationship meets the specified conditions, it indicates that the current target marker is already within the position range corresponding to the preset reference area, and the current displacement state of the traffic management robot meets the adjustment requirements. At this time, there is no need to continue to perform displacement adjustment.

[0027] Among them, the discrete displacement adjustment command is used to control the traffic management robot to complete a discrete displacement action with a preset displacement amount.

[0028] Unlike continuous displacement adjustment, in this embodiment, the traffic management robot performs only one corresponding discrete displacement action each time. After completing one adjustment, it determines whether to continue the next adjustment based on the positional relationship between the target marker and the preset reference area.

[0029] Specifically, after the traffic management robot completes a discrete displacement action, it reacquires the target scene image and redetermines the positional relationship of the target marker relative to the preset reference area based on the reacquired target scene image. If the redetermined positional relationship still does not meet the specified conditions, the discrete displacement adjustment command is determined again based on the redetermined positional relationship, and the traffic management robot is controlled to execute the next discrete displacement action. The above process is repeated until the positional relationship between the target marker and the preset reference area meets the specified conditions.

[0030] Specifically, in the scenario of traffic management at intersections, humanoid bipedal robots can be deployed as traffic management robots in road intersection areas to perform corresponding traffic control tasks according to the traffic conditions in different directions.

[0031] When a traffic management robot needs to direct traffic at a certain intersection, it can acquire a target scene image in the corresponding direction through a camera device, and determine whether the robot's current orientation meets the command requirements based on the positional relationship between the target landmark set in the road direction and the preset reference area in the target scene image.

[0032] For example, when a traffic management robot is currently facing south to perform traffic control tasks, but needs to switch to the east direction due to traffic conditions, the robot can determine the adjustment direction based on the positional relationship of the target landmarks corresponding to the target road direction, and perform the corresponding discrete displacement adjustment action.

[0033] After completing a turning adjustment, due to the robot's inertia during movement and the stability of its bipedal structure, the robot's actual orientation may deviate from the target road direction. Therefore, the robot stops moving and re-acquires images of the target scene. Based on the positional relationship between the newly acquired target landmarks and the preset reference area, it determines whether the current orientation meets the specified conditions.

[0034] When the target marker does not enter the preset reference area, the next discrete adjustment command is determined based on the current positional relationship, enabling the robot to make further adjustments; when the target marker enters the preset reference area, the robot is determined to have completed the direction adjustment, thus enabling the robot to accurately face the target road direction to perform traffic control tasks.

[0035] By using the above method, the continuous pose adjustment process of the traffic management robot is transformed into multiple discrete displacement adjustment processes based on visual feedback results. This allows the robot to reacquire the target state after each small-range adjustment and make the next adjustment based on the latest positional relationship, thereby avoiding adjustment deviations caused by continuous movement at once.

[0036] For example, this application also provides another embodiment of a method for dynamically adjusting the displacement of a traffic management robot. Figure 2 This is a flowchart illustrating the method for dynamically adjusting the displacement of traffic management robots. (For example...) Figure 2 As shown: Step 201: Obtain the target scene image collected by the traffic management robot.

[0037] Specifically, when performing displacement adjustment tasks, traffic management robots acquire images of the target scene using image acquisition devices installed on themselves to obtain environmental information about the target scene.

[0038] The target scenarios can be road intersections, road entrances and exits, construction sections, accident handling areas, traffic control areas, or other areas where traffic management robots need to perform traffic control, traffic guidance, traffic prompts, or safety warnings. Before performing the corresponding traffic management task, the traffic management robot can first move to the target duty position corresponding to the traffic management task and maintain a preset orientation at the target duty position so that the traffic lights, displays, speakers, gesture execution mechanisms, or other traffic control devices set up by the traffic management robot are oriented towards the vehicles or pedestrians to be managed.

[0039] In this embodiment, by acquiring images of the target scene, the traffic management robot can determine the position of the target marker in the current scene based on visual information, and further determine whether displacement adjustment is needed based on the positional relationship between the target marker and the preset reference area.

[0040] It should be noted that in road traffic scenarios, the duty position and orientation of traffic management robots typically need to maintain a relatively stable relative position with lanes, stop lines, pedestrian crossings, road edges, or traffic control facilities. If the traffic management robot deviates from its target duty position, it may result in traffic guidance information not being directed towards the target lane, the robot entering the vehicle passage area, or occupying pedestrian passage space. Therefore, this embodiment of the application adjusts the position and orientation of the traffic management robot using a target scene image, enabling the traffic management robot to gradually reach the target duty position corresponding to the current traffic management task without relying on a precise 3D map of the road scene.

[0041] Step 202: Detect target landmarks in the target scene image to determine the image region of the target landmark in the target scene image; determine the position state of the target landmark relative to the preset reference region based on the positional relationship between the image region and the preset reference region.

[0042] Specifically, after acquiring the target scene image, target marker detection is performed on the target scene image to determine the image region corresponding to the target marker. By determining the image region of the target marker in the target scene image, the positional distribution information of the target marker in the image coordinate system can be obtained.

[0043] In one alternative implementation, the target marker includes a visual marker placed in the target scene. The visual marker can be distinguished from the background area in the target scene by a preset color, shape, or other visual features.

[0044] In traffic management scenarios, visual markers can be placed near the target duty positions that traffic management robots need to reach, and are used to represent the target duty position or its reference direction. For example, visual markers can be placed on road edges, roadside facilities, traffic islands, barriers, construction barriers, temporary traffic control facilities, or pre-deployed positioning bases. Visual markers can also be dedicated visual identifiers that are distinct from road traffic facilities to avoid misidentifying traffic lights, vehicle taillights, traffic signs, or construction warning facilities as target markers.

[0045] Optionally, the image region of the target marker in the target scene image can be determined by: converting the target scene image to a preset color space; segmenting the target scene image according to the color threshold corresponding to the visual marker to obtain candidate marker regions; and determining the image region corresponding to the target marker based on the region area, shape features, and position features of the candidate marker regions.

[0046] By using the above methods, the impact of background information on the target scene on the target marker detection process can be reduced, thereby improving the accuracy of target marker detection.

[0047] Furthermore, after determining the image region corresponding to the target marker, the positional state of the target marker relative to the preset reference region is determined based on the positional relationship between the image region and the preset reference region. The positional state characterizes the current position of the target marker relative to the preset reference region, facilitating the determination of corresponding discrete displacement adjustment methods based on different positional states.

[0048] In this embodiment, the preset reference area can be an image region pre-set in the target scene image. This preset reference area characterizes the area in the target scene image where the target marker should be located when the traffic management robot arrives at the target location or forms a preset relative positional relationship with the target marker.

[0049] The target location can be the duty position corresponding to the current traffic management task. The target duty position can be located outside the normal vehicle passage area and maintain a preset safe distance from the target lane, stop line, pedestrian crossing, road edge, or traffic control area. When the traffic management robot maintains a preset target posture at the target duty position, its traffic indication device can be oriented towards vehicles or pedestrians that need to receive traffic instructions.

[0050] In other words, when the image area corresponding to the target marker is within the preset reference area, the current position of the traffic management robot can be considered to meet the preset position adjustment requirements; when the image area corresponding to the target marker is outside the preset reference area, the displacement adjustment direction that the traffic management robot needs to perform can be determined based on the direction in which the target marker deviates from the preset reference area.

[0051] Optionally, the preset reference area can be predetermined based on the installation location, installation angle, field of view of the image acquisition device, and the target relative position between the traffic management robot and the target marker.

[0052] Specifically, the traffic management robot can be pre-controlled to move to the target location and maintain a preset target posture; at the target location and in the preset target posture, a calibration image including the target marker is acquired by an image acquisition device; a preset reference position is determined based on the imaging position of the target marker in the calibration image, and a preset reference area is delineated in the target scene image based on the preset reference position.

[0053] The preset target posture can be determined based on the traffic management object of the traffic management robot. For example, when the traffic management robot is used to provide traffic guidance to vehicles entering an intersection, its traffic indicator can be oriented towards the corresponding approach lane; when the traffic management robot is used to guide pedestrians, its traffic indicator can be oriented towards the crosswalk or pedestrian waiting area; when the traffic management robot is used to guard against accidents, its warning device can be oriented towards the direction of oncoming traffic.

[0054] Step 203: Based on the positional relationship between the target marker in the target scene image and the preset reference area, iteratively perform discrete displacement adjustments on the traffic management robot until the positional relationship between the target marker and the preset reference area meets the specified conditions.

[0055] Specifically, after determining the positional relationship of the target marker relative to the preset reference area, the system judges whether the current target marker meets the specified conditions based on the positional relationship. When the positional relationship does not meet the specified conditions, the system determines the corresponding discrete displacement adjustment command based on the positional relationship, and controls the traffic management robot to complete a discrete displacement action based on the discrete displacement adjustment command.

[0056] In traffic management scenarios, vehicles, non-motorized vehicles, or pedestrians may continuously be present around traffic management robots. Compared to continuously performing long-distance displacements, breaking down the overall position adjustment process into multiple discrete displacement actions allows the robot to reconfirm its position after each small-amplitude movement. This reduces the possibility of the robot entering vehicle traffic areas, crossing target duty areas, or encroaching on pedestrian traffic space due to excessively large single movement distances.

[0057] Meanwhile, discrete displacement adjustment ensures that each movement has a clearly defined start and stop state. During the robot's stop and stabilization period, it provides more predictable motion states for surrounding vehicles and pedestrians, reducing the disturbance caused to traffic participants by continuous robot movement or sudden changes in direction.

[0058] Optionally, the discrete displacement adjustment command can be determined as follows: when the position state is a first position state located within a preset reference area, the positional relationship is determined to meet the specified conditions; when the position state is a second position state located above the preset reference area, a backward discrete displacement adjustment command is determined; when the position state is a third position state located below the preset reference area, a forward discrete displacement adjustment command is determined; when the position state is a fourth position state located to the left of the preset reference area, a left turn discrete displacement adjustment command is determined; and when the position state is a fifth position state located to the right of the preset reference area, a right turn discrete displacement adjustment command is determined.

[0059] It should be noted that the correspondence between the position of the target marker in the image and the direction of movement of the traffic management robot can be predetermined based on the installation direction of the image acquisition device and the calibration relationship between the traffic management robot and the target marker. For example, if the target marker deviates upward from the preset reference area in the image, it indicates that the distance between the traffic management robot and the target marker is less than the target distance, thus requiring a backward adjustment; if the target marker deviates left or right from the preset reference area in the image, it indicates that the orientation of the traffic management robot deviates relative to the target traffic management direction, thus requiring a rotation adjustment in the corresponding direction.

[0060] Among them, the backward discrete displacement adjustment command, the forward discrete displacement adjustment command, the left turn discrete displacement adjustment command, and the right turn discrete displacement adjustment command are all used to control the traffic management robot to complete a discrete displacement action with a preset displacement amount.

[0061] Specifically, after determining the discrete displacement adjustment command, the corresponding preset displacement amount is determined based on the positional deviation between the target marker and the preset reference area. Specifically, the preset displacement amount can be determined as follows: first, the positional deviation between the target marker and the preset reference area is obtained; then, based on the deviation range to which the positional deviation belongs, the preset displacement amount corresponding to the current deviation range is determined.

[0062] By determining a preset displacement based on the positional deviation, the traffic management robot can perform discrete displacement actions of corresponding magnitude according to the current degree of deviation. When the positional deviation between the target marker and the preset reference area is small, a smaller displacement can be used for adjustment; when the positional deviation is large, a larger displacement can be used for adjustment, thereby improving the flexibility of the robot's displacement adjustment process.

[0063] In one optional implementation, when controlling the traffic management robot to perform a discrete displacement action, the discrete displacement action is divided into two consecutive time phases: an acceleration phase and a stop buffer phase. The acceleration phase is used to control the traffic management robot to move in the corresponding direction according to the current discrete displacement adjustment command, and the stop buffer phase is used to gradually reduce the robot's speed before the end of this discrete displacement action, so that the robot can smoothly transition from the moving state to the stationary state.

[0064] Since traffic management robots typically need to move on hard surfaces, road markings, manhole covers, road joints, or roads with a certain slope, if the robot stops immediately after reaching a preset displacement, it may experience significant swaying due to inertia, changes in road surface adhesion, or shifts in the center of gravity of its two feet. This swaying may not only affect the robot's posture stability but may also cause changes in the orientation of traffic signs and reduce the quality of subsequent image acquisition of the target scene.

[0065] Therefore, dividing discrete displacement actions into an active movement phase and a stopping and buffering phase allows the robot to gradually reduce its speed as it approaches the end point of the discrete action. This helps the robot to remain stably stationary at the target duty position and keeps the traffic signage device facing a stable direction.

[0066] In this embodiment, the acceleration phase and the stop buffer phase are executed sequentially in time. That is, after receiving the discrete displacement adjustment command, the traffic management robot first enters the acceleration phase and automatically switches to the stop buffer phase after the acceleration phase ends; after the stop buffer phase ends, the discrete displacement action is completed, and the traffic management robot returns to an idle state.

[0067] Specifically, after determining the forward discrete displacement adjustment command, the backward discrete displacement adjustment command, the left turn discrete displacement adjustment command, or the right turn discrete displacement adjustment command, the corresponding motion direction, preset speed, acceleration phase duration, and stop buffer phase duration can be determined based on the discrete displacement adjustment command. Subsequently, based on the motion direction, preset speed, and duration of each time phase, motion control parameters are generated to control the traffic management robot to perform this discrete displacement action.

[0068] In one alternative implementation, the traffic management robot can be controlled with a fixed deceleration, so that the robot's speed gradually decreases from its current speed when entering the stop buffer zone to zero.

[0069] The fixed deceleration can be preset based on the robot's mass, gait control capability, ground friction conditions, and allowable range of posture changes.

[0070] In another optional implementation, the speed reduction during the stop buffer phase can be dynamically adjusted based on the robot's posture detection data. For example, information such as torso tilt angle, angular velocity, acceleration, or center of gravity change collected by the inertial measurement unit can be acquired. When the posture detection data indicates that the robot has a large tendency to tilt forward, backward, or laterally, the speed reduction amount in the current control cycle can be reduced to prolong the speed reduction process. When the robot's posture remains stable, the movement speed can continue to decrease according to a preset speed reduction relationship.

[0071] Furthermore, during the discrete displacement adjustment process, when the discrete displacement adjustment command is a left-turn discrete displacement adjustment command or a right-turn discrete displacement adjustment command, the rotational motion of the traffic management robot can be adjusted using attitude detection data.

[0072] Specifically, the posture detection data of the traffic management robot is obtained, the actual rotation angle of the traffic management robot is determined based on the posture detection data, and the rotation action of the traffic management robot is adjusted based on the deviation between the actual rotation angle and the target rotation angle.

[0073] In this way, during the rotation adjustment process, the rotation action is adjusted according to the deviation between the robot's actual rotation state and the target rotation state, enabling the traffic management robot to more accurately complete the discrete rotation adjustment in the corresponding direction.

[0074] In traffic management scenarios, the accuracy of a robot's rotational adjustments directly impacts the orientation of traffic signs. For example, when a traffic management robot needs to direct traffic towards a target lane at an intersection, if the actual rotation angle deviates from the target angle, traffic lights, displays, or hand signals may not be directly facing the target vehicle. Therefore, using posture detection data to provide feedback and adjust rotational movements can improve the visibility and directional accuracy of traffic management information.

[0075] The execution logic of discrete rotation adjustment is similar to that of discrete displacement adjustment, and will not be repeated here.

[0076] Step 204: Control the traffic management robot to stop moving and maintain the current posture for a preset time to allow the traffic management robot to achieve posture stabilization; after the traffic management robot achieves posture stabilization, reacquire the target scene image; based on the reacquired target scene image, redetermine the positional relationship of the target marker relative to the preset reference area; based on the redetermined positional relationship of the target marker relative to the preset reference area, determine whether to continue discrete displacement adjustment.

[0077] Specifically, even if the discrete displacement action is divided into two consecutive time phases—an acceleration phase and a stopping buffer phase—to keep the traffic management robot as stationary as possible during the discrete displacement action, the robot's current posture may still change within a short period due to its own motion inertia and the influence of the bipedal gait adjustment process. Therefore, after completing one discrete displacement adjustment, the traffic management robot is controlled to stop moving and maintain its current posture for a preset duration to ensure posture stability.

[0078] After the robot achieves posture stabilization, it reacquires the target scene image and redetermines the positional relationship of the target marker relative to the preset reference area based on the reacquired target scene image to determine whether the current adjustment result meets the preset requirements.

[0079] Optionally, the decision to continue discrete displacement adjustment can be made as follows: when the redefined positional relationship meets the specified conditions, the displacement adjustment of the traffic management robot is stopped; when the redefined positional relationship does not meet the specified conditions, the next discrete displacement adjustment instruction is determined based on the redefined positional relationship, and the traffic management robot is controlled to continue executing discrete displacement adjustment.

[0080] Through the above iterative process, each discrete displacement adjustment is based on feedback adjustment of the robot's latest state information, thereby improving the stability and accuracy of the robot when adjusting the target position in complex environments.

[0081] In summary, the dynamic adjustment method for the displacement of a traffic management robot provided in this embodiment detects target landmarks in the target scene image to determine the position of the target landmarks relative to a preset reference area. Based on different position states, it determines corresponding discrete displacement adjustment commands, controlling the traffic management robot to execute a discrete displacement action with a preset displacement amount. Simultaneously, after completing one adjustment, it re-acquires the target scene image and determines whether to continue adjusting until specified conditions are met. Through this method, the traffic management robot can gradually adjust according to the target position deviation, avoiding the accumulation of position errors caused by continuous displacement control, and improving the accuracy and stability of the traffic management robot's displacement adjustment process.

[0082] In one specific implementation, the above-mentioned method for dynamically adjusting the displacement of a traffic management robot is further explained by taking the task of adjusting the displacement of a target object pair performed by a traffic management robot as an example.

[0083] Specifically, the traffic management robot in this embodiment includes a perception layer, a decision-making layer, and an execution layer. The perception layer collects target scene information and the robot's own posture information; the decision-making layer determines the position of the target marker relative to a preset reference area based on the target scene information and generates corresponding discrete displacement adjustment commands based on the position status; the execution layer controls the robot to complete the corresponding displacement actions according to the discrete displacement adjustment commands.

[0084] In this embodiment, the perception layer includes a depth camera and an inertial measurement unit (IMU). The depth camera is a RealSense D435i camera used to acquire images of the target scene at a frame rate of 30fps and an image resolution of 640×480. The inertial measurement unit is used to acquire the attitude data of the traffic management robot, including Euler angles, angular velocity, and linear acceleration, with an attitude data sampling frequency of no less than 500Hz.

[0085] The decision-making layer includes visual perception nodes and motion control nodes. The visual perception nodes process the acquired target scene images, detect target markers through color space conversion, image segmentation, and morphological processing, and determine the current position state based on the positional relationship of the target markers relative to a preset reference area. The motion control nodes generate corresponding speed control commands based on the current position state and control the traffic management robot to perform corresponding displacement actions.

[0086] The execution layer is used to receive speed control commands output by the motion control nodes and drive the robot joints to perform corresponding displacement actions through the robot's underlying motion controller.

[0087] Furthermore, in this embodiment, the discrete displacement adjustment process includes piecewise time state control, closed-loop rotation control of the inertial measurement unit, orientation self-calibration and shortest path steering, and rotation speed superposition compensation.

[0088] Specifically, when performing a discrete displacement action, the displacement action is divided into two continuous time phases: an acceleration phase and a stop buffer phase.

[0089] During the acceleration phase, the traffic management robot moves according to the preset speed. After the preset duration is reached, it automatically switches to the stop buffer phase, where the speed is gradually reduced to zero to reduce the impact on the robot's posture during the stopping process.

[0090] After the stop buffer ends, the robot returns to an idle state, waiting for the next discrete displacement adjustment command.

[0091] By using the above-mentioned segmented time state control method, each discrete displacement action has an independent start-up and stop process, which improves the repeatability and stability of the robot's small step displacement adjustment process.

[0092] Furthermore, when the traffic management robot performs rotational adjustment actions, closed-loop rotational control is performed through an inertial measurement unit.

[0093] Specifically, at the start of the rotation, the initial angle value output by the inertial measurement unit is recorded. During the rotation, the cumulative rotation angle is calculated based on the angle values ​​collected in real time by the inertial measurement unit. The calculation formula is as follows: ; in, Indicates the first The angle value output by the inertial measurement unit at any given moment; Indicates the first The angle value output by the inertial measurement unit at any given time.

[0094] To avoid angle wrapping errors during angle changes, a ±π jump is performed during each angle increment calculation. The calculation formula is as follows: ; in, Indicates the first The incremental change in robot posture angle during the next sampling process. hour, ;when hour, .

[0095] When the cumulative rotation angle reaches the tolerance range corresponding to the target rotation angle, the control angular velocity is reduced to zero and the system enters a stop buffer state.

[0096] The rotation stopping condition satisfies: ; in, Indicates the target rotation angle. This indicates the allowable error for the angle.

[0097] Through the closed-loop rotation control method of the inertial measurement unit described above, the rotation process can be corrected in real time according to the actual rotation state of the robot, enabling the robot to complete the target direction adjustment more accurately.

[0098] Furthermore, when performing a direction adjustment task, the rotation direction is determined by direction self-calibration and the shortest path steering method.

[0099] Specifically, the reference angle values ​​of the inertial measurement unit are collected by the robot when it faces the four directions of east, south, west, and north through assisted operation, and are recorded as follows: Upon receiving a direction rotation command, the rotation direction is determined based on the difference between the current inertial measurement unit angle value and the corresponding direction reference value.

[0100] Specifically, the left or right turn direction is determined based on the difference between the current angle value and the target direction reference angle value, combined with the π conversion result. The rotation termination condition is met: ; in, This indicates the robot's real-time attitude angle currently output by the inertial measurement unit; The reference angle representing the target direction is selected from a pre-defined set of reference angles based on the target's rotation direction, and its value satisfies the following: ; in, , , , These represent the reference angles of the inertial measurement unit when the robot is facing north, south, east, and west, respectively. This indicates the permissible error range for the direction angle, used to limit the allowable deviation between the robot's current posture angle and the target direction reference angle.

[0101] By using the above methods, robots can perform autonomous orientation calibration based on pre-calibrated orientation references without relying on external positioning equipment, thereby improving the accuracy of orientation adjustment.

[0102] Furthermore, during the visual position state determination process, target landmarks in the target scene image are detected.

[0103] Specifically, the BGR image acquired by the camera is converted to the HSV color space, and a target color region mask is constructed based on a preset color threshold.

[0104] In one specific implementation, taking the yellow visual marker as an example, the HSV threshold range is set as follows: ; ; .in, Represents the hue components in the HSV color space. Represents the saturation component. Indicates the brightness component.

[0105] The target scene image is segmented using the aforementioned color threshold, and morphological opening and closing operations are performed using a 5×5 kernel to remove noise and fill the target region.

[0106] Furthermore, the candidate regions are filtered based on their area range, aspect ratio, contour fill rate, convex hull density, and quadrilateral approximation.

[0107] The candidate region area range is as follows: The aspect ratio range is Contour fill rate Convex hull density .

[0108] From the candidate regions that meet the above conditions, select the largest effective region as the target marker region, and calculate its center coordinates. Relative to the center of the preset reference area and reference area size The positional relationship. Among them: The x-coordinate of the center point of the target marker image region; The ordinate of the center point of the target marker image region; Indicates the x-coordinate of the center point of the preset reference area; Indicates the ordinate of the center point of the preset reference area; Indicates the width of the preset reference area; This indicates the height of the preset reference area.

[0109] The specific judgment method is as follows: when and When the target marker is located within a preset reference area, its position status is INSIDE. When the target marker is located above the preset reference area, its position status is ABOVE. When the target marker is located below a preset reference area, the position state is set to BELOW. When no valid target marker area is detected, the position state is set to NO_DETECTION. Using this method, the robot's current movement direction can be determined based on the visual inspection results.

[0110] Furthermore, during the rotation adjustment process of the robot, in order to reduce the shift of the body's center of gravity caused by the bipedal robot's rotation in place, a small linear velocity compensation is superimposed during the rotation action.

[0111] Specifically, during the rotation process, a linear velocity component matching the rotation direction is simultaneously output; for example, the linear velocity in the x-direction is approximately... The linear velocity in the y-direction is approximately By superimposing compensation when the online velocity direction matches the rotation direction, the robot maintains stability during rotation, reducing the risk of tipping over.

[0112] In a specific application example, the above method is used to perform the target object centering task.

[0113] First, the system is initialized by loading the configuration file, which includes parameters such as camera serial number, image resolution 640×480@30fps, visual marker color threshold, and preset reference area coordinates. Then, the RealSense D435i camera and inertial measurement unit are started to acquire data.

[0114] Subsequently, the visual perception node acquires images of the target scene at a frequency of approximately 33Hz, performs color space conversion, color region segmentation, and target region filtering on the images, and determines the position and status of the target markers.

[0115] When the target marker is located below the preset reference area, it is determined that the robot needs to perform forward adjustment.

[0116] After receiving the forward adjustment command, the motion control node switches the robot's state from idle to forward and sets the linear velocity. .

[0117] The robot is driven to move continuously for 0.4s by a 50Hz control cycle, then switches to a stop buffer state, holds for 0.5s, and then returns to an idle state.

[0118] After completing one displacement adjustment, the visual inspection step is repeated to determine whether the target marker has entered the preset reference area.

[0119] If the requirements are still not met, repeat the above adjustment process.

[0120] When indoor lighting conditions are between 300 lux and 500 lux, the detection accuracy of the color markers is greater than 95% when tested using the above method, the single forward discrete displacement is about 0.32 m, and the final stopping error is less than 0.02 m.

[0121] Furthermore, in the direction rotation task, rotation adjustment is performed through direction self-calibration. The robot is controlled by an auxiliary control device to rotate to the four directions of east, south, west, and north, and the corresponding yaw angle of the inertial measurement unit is recorded as the direction reference value. Example reference values ​​are: ; ; ; .

[0122] Upon receiving a control command to rotate to the target direction, the shortest rotation path is calculated based on the current inertial measurement unit angle value and the corresponding directional reference value, and the left or right turn direction is determined.

[0123] Set the angular velocity during rotation. Simultaneously, linear velocity compensation is added: when When the rotation reaches the target range, the angular velocity is controlled to return to zero, and after a 0.2s buffer, it returns to the idle state.

[0124] Through the above methods, the direction adjustment accuracy reaches... Approximately: Furthermore, the robot is able to maintain its upright stability during rotation.

[0125] This invention also provides a traffic management robot that can be applied to target scenarios that require target position alignment, posture adjustment, or motion positioning. By acquiring images of the target scenario and adjusting the robot sequentially based on the positional relationship between the target marker and the preset reference area in the target scenario image, the robot gradually moves the target marker closer to the preset reference area.

[0126] Specifically, the traffic management robot includes an image acquisition module, a processing module, and a motion control module.

[0127] The image acquisition module is used to acquire images of the target scene.

[0128] Specifically, the image acquisition module can be installed on the traffic management robot to collect image information of the target scene in which the robot is currently located. The target scene image includes image information corresponding to the target landmarks. The processing module can identify the target landmarks based on the target scene image to determine the position and status of the target landmarks in the current field of view.

[0129] The processing module is used to determine the position of the target marker relative to the preset reference area based on the target scene image, and to determine the corresponding discrete pose adjustment command based on the position.

[0130] Specifically, after acquiring the target scene image, the processing module analyzes and processes the target scene image to determine the image region corresponding to the target marker, and determines the position state of the target marker relative to the preset reference region based on the positional relationship between the image region corresponding to the target marker and the preset reference region.

[0131] When the positional relationship between the target marker and the preset reference area does not meet the preset adjustment conditions, the processing module determines the corresponding discrete pose adjustment command based on the current positional state.

[0132] Among them, the discrete pose adjustment command is used to instruct the traffic management robot to perform a pose adjustment action with a preset adjustment step size and / or a preset adjustment angle.

[0133] The motion control module is used to control the traffic management robot to execute the pose adjustment action corresponding to the discrete pose adjustment command, so as to complete a discrete pose adjustment.

[0134] Specifically, after receiving the discrete pose adjustment command sent by the processing module, the motion control module controls the robot to perform the corresponding motion action according to the discrete pose adjustment command, so that the robot completes a discrete pose adjustment with a preset adjustment step size and / or a preset adjustment angle.

[0135] Furthermore, after completing a discrete pose adjustment, the processing module is also used to control the image acquisition module to reacquire the target scene image and redetermine the position of the target marker relative to the preset reference area based on the reacquired target scene image.

[0136] Specifically, since traffic management robots may experience motion errors, posture changes, or deviations between actual and expected motion effects during the execution of pose adjustment actions, after completing a discrete pose adjustment, the processing module determines whether the preset adjustment conditions are met based on the redefined position state.

[0137] When the redefined position meets the preset adjustment conditions, it indicates that the target marker has reached the target range corresponding to the preset reference area, and the processing module controls the traffic management robot to stop adjusting.

[0138] When the redefined position does not meet the preset adjustment conditions, the processing module redefined the corresponding discrete pose adjustment command based on the redefined position and controlled the motion control module to execute the next discrete pose adjustment until the preset adjustment conditions are met.

[0139] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code, which, when accessed and executed by the computer, processor, or hardware, implements the methods shown in the above embodiments.

[0140] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0141] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A method for dynamically adjusting the displacement of a traffic management robot, characterized in that, The method is applied to traffic management robots deployed in the target scenario; The target scene is equipped with target markers, including: Acquire the target scene image collected by the traffic management robot; Based on the positional relationship between the target marker and the preset reference area in the target scene image, the traffic management robot is iteratively adjusted by discrete displacement until the positional relationship between the target marker and the preset reference area meets the specified conditions; Specifically, based on the positional relationship between the target landmark in the target scene image and a preset reference area, the traffic management robot is iteratively subjected to discrete displacement adjustments, including: When the positional relationship does not meet the specified conditions, a discrete displacement adjustment command is determined based on the positional relationship; the discrete displacement adjustment command is used to control the traffic management robot to complete a discrete displacement action with a preset displacement amount; The traffic management robot is subjected to discrete displacement adjustment according to the discrete displacement adjustment command.

2. The method for dynamically adjusting the displacement of a traffic management robot according to claim 1, characterized in that, Before determining the positional relationship between the target landmark in the target scene image and the preset reference area, the method further includes: Target landmark detection is performed on the target scene image to determine the image region of the target landmark in the target scene image; The positional state of the target marker relative to the preset reference area is determined based on the positional relationship between the image area and the preset reference area.

3. The method for dynamically adjusting the displacement of a traffic management robot according to claim 2, characterized in that, The step of determining the discrete displacement adjustment command based on the positional relationship includes: When the position state is a first position state located within the preset reference area, it is determined that the position relationship satisfies the specified condition; When the position state is the second position state located above the preset reference area, a backward discrete displacement adjustment command is determined. When the position state is the third position state located below the preset reference area, a forward discrete displacement adjustment command is determined; When the position state is the fourth position state located to the left of the preset reference area, a left turn discrete displacement adjustment command is determined; When the position state is the fifth position state located to the right of the preset reference area, a right turn discrete displacement adjustment command is determined; The backward discrete displacement adjustment command, the forward discrete displacement adjustment command, the left turn discrete displacement adjustment command, and the right turn discrete displacement adjustment command are all used to control the traffic management robot to complete a discrete displacement action with a preset displacement amount.

4. The method for dynamically adjusting the displacement of a traffic management robot according to claim 3, characterized in that, The step of adjusting the discrete displacement of the traffic management robot according to the discrete displacement adjustment command includes: When the discrete displacement adjustment command is a left turn discrete displacement adjustment command or a right turn discrete displacement adjustment command: Obtain the posture detection data corresponding to the traffic management robot; The actual rotation angle of the traffic management robot is determined based on the posture detection data. The rotation motion of the traffic management robot is adjusted based on the deviation between the actual rotation angle and the target rotation angle.

5. The method for dynamically adjusting the displacement of a traffic management robot according to claim 2, characterized in that, The target markers include visual markers set in the target scene; The step of detecting target landmarks in the target scene image and determining the image region of the target landmark in the target scene image includes: Convert the target scene image to a preset color space; The target scene image is segmented based on the color threshold corresponding to the visual marker to obtain candidate marker regions; The image region corresponding to the target marker is determined based on the area, shape features, and positional features of the candidate marker region.

6. The method for dynamically adjusting the displacement of a traffic management robot according to claim 1, characterized in that, After adjusting the discrete displacement of the traffic management robot according to the discrete displacement adjustment command, the process further includes: The traffic management robot is controlled to stop moving and maintain its current posture for a preset time to allow the traffic management robot to achieve posture stabilization. After the traffic management robot achieves posture stabilization, the target scene image is reacquired. Based on the reacquired target scene image, the positional relationship of the target marker relative to the preset reference area is re-determined; Based on the redefined positional relationship between the target marker and the preset reference area, it is determined whether to continue discrete displacement adjustment.

7. The method for dynamically adjusting the displacement of a traffic management robot according to claim 6, characterized in that, The step of determining whether to continue discrete displacement adjustment based on the redefined positional relationship of the target marker relative to the preset reference area includes: When the redefined positional relationship satisfies the specified conditions, the displacement adjustment of the traffic management robot is stopped; If the redefined positional relationship does not meet the specified conditions, the next discrete displacement adjustment command is determined based on the redefined positional relationship.

8. The method for dynamically adjusting the displacement of a traffic management robot according to claim 1, characterized in that, The preset displacement is determined through the following steps; Obtain the positional deviation between the target marker and the preset reference area; The corresponding preset displacement is determined based on the deviation range to which the position deviation belongs.

9. A traffic management robot, characterized in that, include: The image acquisition module is used to acquire images of the target scene; The processing module is used to determine the position state of the target marker relative to the preset reference area based on the target scene image, and to determine the corresponding discrete pose adjustment instruction based on the position state. The discrete pose adjustment instruction is used to instruct the traffic management robot to perform a pose adjustment action with a preset adjustment step size and / or a preset adjustment angle. The motion control module is used to control the traffic management robot to execute the pose adjustment action corresponding to the discrete pose adjustment command, so as to complete a discrete pose adjustment. The processing module is further configured to, after completing one discrete pose adjustment, control the image acquisition module to reacquire the target scene image, and redetermine the position state of the target marker relative to the preset reference area based on the reacquired target scene image, and determine whether the redetermined position state meets the preset adjustment conditions. If the preset adjustment conditions are not met, the processing module re-determines the corresponding discrete pose adjustment command and controls the motion control module to execute the next discrete pose adjustment until the preset adjustment conditions are met.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to execute the method for dynamically adjusting the displacement of the traffic management robot as described in any one of claims 1 to 8.

Images