A robot posture updating method, device and storage medium

Abstract

The present application provides a method, device and storage medium for updating the posture of a robot, the method comprising: identifying whether the target robot has human position intervention behavior according to at least one of the real-time posture angle, real-time acceleration and real-time motion speed of the target robot; if it is identified that there is human position intervention behavior, determining whether the target robot performs the first positioning and retrieval according to the posture accuracy evaluation result of the target robot and the position evaluation result of the target robot relative to a specific area; if it is determined that the target robot performs the first positioning and retrieval, then updating the current posture of the target robot according to the current recorded posture, sensor data and search strategy of the target robot; wherein the search strategy is provided with an initial search step length and a step length adjustment strategy. In this way, the present scheme can effectively improve the efficiency and success rate by actively monitoring and adopting a targeted method for posture retrieval, thereby ensuring the normal operation of the robot.

Description

Technical Field

[0001] The present application relates to the field of robotics technology, and in particular to a method, device and storage medium for updating the posture of a robot. Background Art

[0002] With the development of automation technology, robots have been applied in many areas of life. For example, cleaning robots are widely used in large commercial scenarios such as shopping malls and office buildings. Cleaning robots perform efficient cleaning tasks with their compact models, automatic navigation, and intelligent systems with real-time feedback. However, in actual use, especially in commercial places such as large shopping malls or supermarkets, cleaning robots may lose their positioning due to various reasons, affecting their normal operation.

[0003] The current method of updating the position of cleaning robots often relies on users to manually push the robot to the charging station for repositioning. This process not only increases the cost of manual operation, but also reduces the robot's work efficiency. If the cleaning robot frequently loses its position or times out, frequent manual intervention is required, resulting in increased usage costs and other problems. In other solutions, when the robot loses its position, it does not distinguish between the cause and position, and uses a unified retrieval strategy for position retrieval and update, with a low success rate. Summary of the invention

[0004] The applicant has found that in commercial environments, cleaning robots are often accidentally touched, pushed or lifted to other places due to the mobility of customers. Especially in areas with high customer traffic, the robot is sometimes accidentally pushed to a place unrelated to the original working path, causing its positioning system to lose its reference point, resulting in positioning loss. Once the positioning is lost, if the positioning cannot be restored in time, navigation timeout or machine crash may occur. In view of this, the purpose of this application is to provide a method, device and storage medium for updating the position of a robot.

[0005] The embodiment of the present application provides a method for updating the posture of a robot, the updating method comprising:

[0006] Identify whether there is human position intervention behavior on the target robot according to at least one of the real-time posture angle, real-time acceleration, and real-time motion speed of the target robot;

[0007] If human position intervention behavior is identified, determining whether the target robot should perform a first positioning retrieval according to a position and posture accuracy evaluation result of the target robot and a position evaluation result of the target robot relative to a specific area;

[0008] If it is determined that the target robot performs the first positioning and retrieval, the current posture of the target robot is updated according to the current recorded posture of the target robot, sensor data and search strategy; wherein the search strategy is provided with an initial search step length and a step length adjustment strategy.

[0009] Optionally, identifying whether the target robot has human position intervention behavior according to at least one of a real-time posture angle, a real-time acceleration, and a real-time motion speed of the target robot includes:

[0010] When the angle difference between the real-time posture angle of the target robot and the corresponding reference angle is higher than the angle threshold for a cumulative number of times and / or a duration that exceeds a first preset persistence condition, it is determined that the target robot has human position intervention behavior;

[0011] And / or, when the acceleration difference between the real-time acceleration of the target robot and the corresponding reference acceleration is higher than the acceleration threshold value for a cumulative number of times and / or the duration exceeds a second preset persistence condition, it is determined that the target robot has human position intervention behavior;

[0012] And / or, when the speed difference between the real-time movement speed of the target robot and the corresponding expected speed is higher than the speed threshold for a cumulative number of times and / or the duration exceeds a third preset continuity condition, it is determined that there is human position intervention behavior in the target robot.

[0013] Optionally, the sensor data includes laser point cloud data, and the pose accuracy evaluation result of the target robot is determined by the following steps:

[0014] Acquire the current recorded position and posture of the target robot and the real-time laser point cloud data at the current recorded position and posture;

[0015] According to the currently recorded posture, matching processing is performed using the real-time laser point cloud data and a known environment map to determine a posture matching score;

[0016] The pose accuracy evaluation result of the target robot is determined according to the pose matching score within a preset time length.

[0017] Optionally, determining whether to perform a first positioning retrieval of the target robot according to a pose accuracy evaluation result of the target robot and a position evaluation result of the target robot relative to a specific area includes:

[0018] When the pose accuracy evaluation result of the target robot indicates that the current pose does not match and the target robot is not located in the specific area, it is determined that the target robot performs a first positioning retrieval; the specific area includes an elevator area and a charging pile area.

[0019] Optionally, updating the current posture of the target robot according to the current recorded posture of the target robot, sensor data, and search strategy includes:

[0020] According to the initial search step length set in the search strategy, the target robot is controlled to perform map search and matching; wherein the initial search step length includes an initial moving distance and / or an initial rotation angle;

[0021] If the match is successful, the current position and posture of the target robot is updated according to the calculation result;

[0022] If the match fails but the update end condition is not met, the initial search step size is adjusted according to the step size adjustment strategy, and the target robot is re-controlled to perform map search and matching according to the adjusted initial search step size until the match is successfully updated to update the current position of the target robot or the update end condition is met;

[0023] If the matching fails and the update end condition is met, an artificial posture recovery operation reminder is executed, and the posture obtained based on the artificial posture recovery operation is determined as the current posture of the target robot.

[0024] Optionally, after determining that the target robot needs to perform a first positioning and retrieval, the updating method further includes:

[0025] Determine whether the current task status of the target robot is task execution;

[0026] If so, suspend the currently executed task;

[0027] After updating the current posture of the target robot, the updating method further includes:

[0028] Determine whether the current task status of the target robot is paused;

[0029] If yes, restart the suspended task.

[0030] Optionally, the sensor data is a frame of laser point cloud data, or a point cloud sub-image composed of multiple frames of laser point cloud data.

[0031] Optionally, identifying whether the target robot has human position intervention behavior according to at least one of a real-time posture angle, a real-time acceleration, and a real-time motion speed of the target robot includes:

[0032] When the current driving area of ​​the target robot meets the flat terrain state condition, and the cumulative number of times and / or the duration of the angle difference between the real-time posture angle of the target robot and the corresponding reference angle is higher than the angle threshold exceeds the first preset persistence condition, it is determined that the target robot has human position intervention behavior;

[0033] When the acceleration difference between the real-time acceleration of the target robot and the corresponding reference acceleration is higher than the acceleration threshold value for a cumulative number of times and / or the duration exceeds a second preset persistence condition, it is determined that the target robot has human position intervention behavior;

[0034] When the target robot is not in an emergency braking state, and the speed difference between the real-time movement speed of the target robot and the corresponding expected speed is higher than the speed threshold for a cumulative number of times and / or a duration that exceeds a third preset continuity condition, it is determined that the target robot has human position intervention behavior.

[0035] The embodiment of the present application also provides a robot posture updating device, the updating device comprising:

[0036] A behavior recognition module, used to identify whether the target robot has human position intervention behavior based on at least one of the real-time posture angle, real-time acceleration and real-time movement speed of the target robot;

[0037] A positioning and retrieval judgment module is used to determine whether the target robot should perform a first positioning and retrieval according to a position and posture accuracy evaluation result of the target robot and a position evaluation result of the target robot relative to a specific area if human position intervention behavior is identified;

[0038] An updating module is used to update the current posture of the target robot according to the current recorded posture, sensor data and search strategy of the target robot if it is determined that the target robot performs the first positioning retrieval; wherein the search strategy is provided with an initial search step length and a step length adjustment strategy.

[0039] An embodiment of the present application also provides an electronic device, comprising: a processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor and the memory communicate via the bus, and when the machine-readable instructions are executed by the processor, the steps of the update method as described above are performed.

[0040] An embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the updating method described above are executed.

[0041] The embodiments of the present application provide a method, device and storage medium for updating the posture of a robot. The updating method comprises: identifying whether there is any human intervention in the position of the target robot, such as being moved, based on at least one of the real-time posture angle, real-time acceleration and real-time movement speed of the target robot; if it is identified that there is any human intervention, determining whether the target robot performs a first positioning and retrieval based on the posture accuracy evaluation result of the target robot and the position evaluation result of the target robot relative to a specific area; if it is determined that the target robot performs a first positioning and retrieval, updating the current posture of the target robot based on the current recorded posture of the target robot, sensor data and search strategy; wherein the search strategy is provided with an initial search step size and a step size adjustment strategy.

[0042] In this way, by actively monitoring the real-time data of the target robot, it is possible to timely detect whether the target robot has any human position intervention behavior. For the positioning loss caused by human position intervention behavior, the robot often does not deviate too much from the original position, and is suitable for automatic positioning and retrieval; and after identifying the presence of human position intervention behavior, by combining the robot's posture accuracy evaluation and position evaluation relative to a specific area, it is determined whether positioning and retrieval are required, which can improve the intelligence of the judgment method, and can also reduce unnecessary positioning attempts, which is helpful to improve the robot's adaptability and task execution accuracy in complex environments; after comprehensive judgment determines that the first positioning and retrieval is required, this scheme updates the robot's positioning and retrieval according to the initial search step size and step size adjustment strategy, which can improve the success rate of retrieval.

[0043] In summary, this solution can monitor the abnormal posture of the robot caused by specific human behaviors more timely by actively monitoring the status of the robot, and adopt targeted methods to recover the posture, which can improve the success rate and efficiency of recovery, and effectively reduce the occurrence of navigation timeouts or crashes, thereby further enhancing the performance of the robot in practical applications.

[0044] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are specifically cited below and described in detail with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0046] Figure 1A flowchart of a method for updating a robot posture provided in an embodiment of the present application;

[0047] Figure 2 A partial process diagram of a posture updating method provided in this application;

[0048] Figure 3 One of the structural schematic diagrams of a robot posture updating device provided in an embodiment of the present application;

[0049] Figure 4 A second structural schematic diagram of a robot posture updating device provided in an embodiment of the present application;

[0050] Figure 5 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0051] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application usually described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, each other embodiment obtained by those skilled in the art without making creative work belongs to the scope of protection of the present application.

[0052] First, the application scenarios to which the present application can be applied are introduced. The present application can be applied to robots, and specifically to robots that need to perform autonomous positioning and navigation in a dynamic environment, such as delivery robots, commercial cleaning robots, and inspection robots.

[0053] Research has found that with the development of automation technology, robots have been used in many areas of life. For example, cleaning robots have begun to be used in large commercial scenarios such as shopping malls and office buildings. Cleaning robots perform efficient cleaning tasks with their compact models, automatic navigation, and intelligent systems with real-time feedback. However, in actual use, especially in commercial places such as large shopping malls or supermarkets, cleaning robots may lose their positioning due to various reasons, affecting their normal work.

[0054] Moreover, in commercial environments, cleaning robots are often accidentally touched, pushed or moved to other places due to the mobility of customers. Especially in areas with high passenger flow, robots are sometimes accidentally pushed to places unrelated to the original working path, causing their positioning system to lose reference points, resulting in positioning loss. In this case, the robot is generally not too far away from the original positioning position. If the robot is not in a specific area, it can try to automatically retrieve the position. Once the positioning is lost, if the positioning cannot be restored in time, navigation timeout or machine crash may occur. The current method for updating the position of cleaning robots often relies on users to manually push the robot to the charging station for repositioning. This process not only increases the cost of manual operation, but also reduces the work efficiency of the robot. If the cleaning robot frequently loses its positioning or times out, frequent manual intervention is required, resulting in problems such as increased usage costs.

[0055] Based on this, the embodiments of the present application provide a method, device and storage medium for updating the robot's posture. By actively monitoring and adopting targeted methods to retrieve the posture, the efficiency and success rate can be effectively improved, thereby ensuring the normal operation of the robot and improving the robot's degree of automation.

[0056] See also Figure 1 , Figure 1 This is a flow chart of a method for updating the robot posture provided by an embodiment of the present application. Figure 1 As shown in , the updating method provided by the embodiment of the present application includes:

[0057] S101. Identify whether there is any human position intervention behavior on the target robot according to at least one of the real-time posture angle, real-time acceleration, and real-time movement speed of the target robot.

[0058] That is, in this solution, the robot actively monitors the motion data in real time during operation. These real-time data can help the robot identify motion data anomalies caused by external intervention, and thus can help the robot actively identify human position intervention behaviors, such as being pushed or moved by people.

[0059] S102: If human intervention is detected, determine whether to perform a first positioning retrieval for the target robot based on a posture accuracy assessment result of the target robot and a position assessment result of the target robot relative to a specific area.

[0060] After human intervention is detected, the results of the posture accuracy evaluation and the position evaluation results relative to a specific area are combined to determine whether to perform the first positioning and retrieval. When there is human intervention, the robot may lose its position. At this time, the posture accuracy evaluation results are abnormal, or it may not have lost its position. At this time, there is no need to perform positioning and retrieval; therefore, after human intervention is detected, the posture accuracy evaluation results are obtained, and it is possible to detect whether the robot has lost its position in time, so as to facilitate timely retrieval; at the same time, it is determined whether the robot is located in a specific area, such as a charging pile area. If the robot is located in the charging pile area, it is more efficient and accurate to use the charging pile to retrieve it, and there is no need to perform the first positioning and retrieval; therefore, according to the judgment results, it is determined whether to perform the first positioning and retrieval, which can detect the robot's positioning in time, reduce invalid retrieval, and improve the robot's intelligence. If the robot loses its position, but no human intervention is actively detected, the position loss is likely to be caused by environmental changes. At this time, even if the robot performs automatic positioning and retrieval, it is difficult to accurately retrieve it. Therefore, in this case, the staff can be notified to manually intervene in the retrieval, which is more efficient.

[0061] S103. If it is determined that the target robot performs the first positioning and retrieval, the current posture of the target robot is updated according to the current recorded posture, sensor data and search strategy of the target robot; wherein the search strategy is provided with an initial search step length and a step length adjustment strategy.

[0062] The search strategy is provided with an initial search step length and a step length adjustment strategy. The retrieved initial position and angle are determined according to the current recorded posture of the robot, and the current recorded posture may be the current posture data of the robot that does not meet the posture accuracy evaluation result.

[0063] The sensor data can be lidar data, image sensor data, etc.

[0064] In this embodiment, the robot may be provided with a device capable of determining angle, acceleration and speed in real time, such as an IMU sensor, an odometer, a wheel speed sensor, etc., so as to determine whether human position intervention occurs based on the real-time determined data. When intervention is identified and the first positioning retrieval condition is met, the position and posture may be updated timely and accurately based on the sensor information collected by the corresponding device on the robot, thereby ensuring the accuracy of the robot's position and posture.

[0065] The present application provides a method for updating the position and posture of a robot. First, by actively monitoring the real-time data of the target robot, it is possible to timely detect whether there is any human position intervention behavior in the target robot, for example: according to at least one of the real-time posture angle, the real-time acceleration and the real-time movement speed, determine whether there is any human moving behavior or human pushing behavior during the operation of the robot; then, after identifying the presence of human position intervention behavior, by combining the robot's position and posture accuracy evaluation and the position evaluation relative to a specific area, it is determined whether positioning and retrieval are required, which can improve the intelligence of the judgment method, and can also avoid unnecessary positioning attempts, which is helpful to improve the robot's adaptive ability and task execution accuracy in complex environments. In this way, when it is determined that the target robot is located in a specific area (on a charging pile or in an elevator), this solution is not used for positioning and retrieval (directly using the position of the charging pile for retrieval) or no retrieval is performed, which can improve the efficiency of position determination and avoid invalid operations; after comprehensive judgment determines that positioning and retrieval are required, this solution performs robot positioning and retrieval update according to the initial search step size and step size adjustment strategy, which can improve the success rate of retrieval.

[0066] In summary, the above embodiments constitute a complete process of self-correction and positioning retrieval, ensuring that the target robot can quickly identify problems and return to the accurate position when facing human intervention or even other external interference, thereby improving the robot's autonomy, stability and reliability.

[0067] The following is a description of the exemplary steps of the present application embodiment:

[0068] With respect to step S101, in this step, when identifying whether the target robot has any human position intervention behavior based on at least one of the real-time posture angle, real-time acceleration and real-time movement speed of the target robot, it may include: identifying whether the target robot has any human position intervention behavior based on the real-time posture angle of the target robot; identifying whether the target robot has any human position intervention behavior based on the real-time acceleration of the target robot; identifying whether the target robot has any human position intervention behavior based on the real-time movement speed of the target robot; identifying whether the target robot has any human position intervention behavior based on the real-time posture angle and real-time acceleration of the target robot; identifying whether the target robot has any human position intervention behavior based on the real-time posture angle and real-time movement speed of the target robot; identifying whether the target robot has any human position intervention behavior based on the real-time acceleration and real-time movement speed of the target robot; identifying whether the target robot has any human position intervention behavior based on the real-time posture angle, real-time acceleration and real-time movement speed of the target robot.

[0069] This step may also include identifying whether there is human position intervention behavior in the target robot based on at least one of at least one real-time posture angle, at least one real-time acceleration, and at least one real-time movement speed of the target robot.

[0070] Here, the human position intervention behavior may refer to external factors, especially the active or passive intervention of personnel on the posture, position or motion state of the robot, causing the robot to deviate from the predetermined trajectory or position. Such intervention behavior may be intentional or unintentional, but it may affect the normal operation and positioning of the robot.

[0071] For example, when step S101 identifies whether the target robot has human position intervention behavior, it can specifically include: identifying whether the target robot has human moving behavior, or identifying whether the target robot has human pushing behavior. The specific type of the human position intervention behavior can be determined according to different real-time data.

[0072] For example, if it is determined based on real-time data that the robot does not produce continuous movement, or the acceleration and speed change suddenly, it is determined that there is human movement; if it is determined based on real-time data that the robot produces continuous movement, and the acceleration and speed change gradually over a period of time, rather than sudden instantaneous changes, it is determined that there is human pushing. For example, the maximum speed of the robot is 0.5m / s, but the current speed is detected to be 1m / s, which may be caused by human pushing. The determination of human moving behavior and human pushing behavior here is only for illustrative purposes, and is not the only way to determine.

[0073] Continuing with step S101 , the real-time attitude angle generally refers to the direction or tilt angle of the robot, reflecting the spatial orientation of the robot. The real-time attitude angle may include a real-time pitch angle, and / or a real-time roll angle, and / or a real-time yaw angle.

[0074] The real-time acceleration refers to the acceleration change rate of the robot in each direction, and the real-time acceleration may include real-time pitch acceleration, and / or real-time roll acceleration, and / or real-time yaw acceleration.

[0075] The real-time movement speed reflects the movement rate of the robot in a certain direction, and the real-time speed may include the real-time wheel speed and / or angular velocity.

[0076] Among them, the acquisition of real-time posture angle, real-time acceleration and real-time motion speed, this application also provides several acquisition methods:

[0077] Devices for acquiring real-time attitude angles include but are not limited to the following devices: inertial measurement unit, gyroscope, etc.

[0078] Devices for acquiring real-time acceleration include but are not limited to the following devices: inertial measurement unit, accelerometer, etc.

[0079] Devices for obtaining real-time motion speed include but are not limited to the following devices: inertial measurement unit, wheel speed meter, etc.

[0080] In one embodiment provided in the present application, the identification of whether the target robot has human position intervention behavior is based on at least one of the real-time posture angle, real-time acceleration and real-time movement speed of the target robot, including: when the angle difference between the real-time posture angle of the target robot and the corresponding reference angle is higher than the angle threshold for a cumulative number of times and / or the duration exceeds a first preset continuity condition, it is determined that the target robot has human position intervention behavior.

[0081] And / or, when the acceleration difference between the real-time acceleration of the target robot and the corresponding reference acceleration is higher than the acceleration threshold for a cumulative number of times and / or the duration exceeds a second preset continuity condition, it is determined that there is human position intervention behavior in the target robot.

[0082] And / or, when the speed difference between the real-time movement speed of the target robot and the corresponding expected speed is higher than the speed threshold for a cumulative number of times and / or the duration exceeds a third preset continuity condition, it is determined that there is human position intervention behavior in the target robot.

[0083] In this embodiment, the reference angle may be a standard posture angle of the robot when it is in a stationary state, or a standard posture angle of the robot in a sloped area, or an average angle of the robot in a stationary state. It is understandable that even if the robot is in a non-slope area, the ground conditions in different areas may be different, such as there are still certain undulations, such as small pits and protrusions. Therefore, obtaining the average angle of the robot in a stationary state can better balance the differences in ground conditions and reduce misjudgment.

[0084] The reference acceleration may be a standard acceleration when the robot is in a stationary state, or a preset initial acceleration of the robot, or an average acceleration of the robot in a stationary state. The expected speed may be an expected speed based on a task, such as a target speed issued by a robot navigation system.

[0085] When at least one real-time posture angle is obtained, and the cumulative number of times and / or the duration of the angle difference between the real-time posture angle of the target robot and the corresponding reference angle is higher than the angle threshold exceeds a first preset continuity condition, it is determined that there is human position intervention behavior in the target robot.

[0086] Exemplarily, it may include: when the difference between the sum of the real-time posture angles (such as the real-time pitch angle and the real-time roll angle) and the corresponding average value of the reference angles of the static state is higher than the cumulative number of times of the first angle threshold and / or the duration exceeds the first preset continuity condition, it can be determined that there is human position intervention behavior in the target robot. Here, the identified human position intervention behavior is generally: there is human moving.

[0087] It may also include, for each real-time posture angle, when the angle difference between any real-time posture angle and the corresponding reference angle (the real-time pitch angle, the real-time roll angle, and the real-time yaw angle have their own reference angles) is higher than the second angle threshold for a cumulative number of times and / or the duration exceeds a first preset continuity condition, determining that the target robot has human position intervention behavior; or for each real-time posture angle, when the angle difference between all real-time posture angles and the corresponding reference angle (the real-time pitch angle, the real-time roll angle, and the real-time yaw angle have their own reference angles) is higher than the third angle threshold for a cumulative number of times and / or the duration exceeds the first preset continuity condition, determining that the target robot has human position intervention behavior; here, the identified human position intervention behavior is generally: the presence of human pushing or moving.

[0088] When the acquired real-time acceleration is at least one, when the acceleration difference between the real-time acceleration of the target robot and the corresponding reference acceleration is higher than the acceleration threshold value for a cumulative number of times and / or a duration that exceeds a second preset persistence condition, it is determined that the target robot has an artificial position intervention behavior, including: for each real-time acceleration, when the acceleration difference between any real-time acceleration and the corresponding reference acceleration (real-time pitch acceleration, real-time roll acceleration, and real-time yaw acceleration have their own reference acceleration) is higher than the second acceleration threshold value for a cumulative number of times and / or a duration that exceeds a second preset persistence condition, it is determined that the target robot has an artificial position intervention behavior. intervention behavior; or for each real-time acceleration, when the acceleration difference between all real-time accelerations and the corresponding reference acceleration (real-time pitch acceleration, real-time roll acceleration, and real-time yaw acceleration have their own static reference acceleration) is higher than the third acceleration threshold for the cumulative number of times and / or the duration exceeds the second preset continuity condition, it is determined that the target robot has human position intervention behavior; or when the acceleration difference between the sum of all real-time accelerations and the corresponding reference acceleration (which may be the static acceleration average value) is higher than the acceleration threshold for the cumulative number of times and / or the duration exceeds the second preset continuity condition, it is determined that the target robot has human position intervention behavior.

[0089] Here, the identified human location intervention behavior is generally: there is human push.

[0090] When the real-time movement speed is the actual wheel speed of the robot, when the speed difference between the real-time movement speed of the target robot and the corresponding expected speed is higher than the speed threshold for a cumulative number of times and / or a duration that exceeds a third preset continuity condition, it is determined that the target robot has human position intervention behavior, including: when the speed difference between the actual wheel speed of the target robot and the navigation speed is higher than the speed threshold for a cumulative number of times and / or a duration that exceeds a third preset continuity condition, it is determined that the target robot has human position intervention behavior.

[0091] Here, the identified human location intervention behavior is generally: there is human push.

[0092] The first continuity condition, the second continuity condition and the third continuity condition are provided with time thresholds under corresponding judgment conditions, and the time thresholds in the three conditions are not necessarily the same.

[0093] In this way, by real-time monitoring of changes in the robot's posture angle, acceleration, and movement speed, and selecting at least one change and combining a series of preset thresholds and conditions to determine whether there is human position intervention behavior, it is possible to effectively identify and respond to human position intervention behavior, thereby enhancing the robot's adaptive ability and intelligent behavior, and helping the robot to smoothly perform tasks in complex and dynamic environments, thereby improving the accuracy and reliability of task execution.

[0094] In order to improve the accuracy of the determination result of human location intervention behavior, the present application also provides another determination method.

[0095] In another embodiment provided by the present application, the identifying whether the target robot has human position intervention behavior based on at least one of the real-time posture angle, real-time acceleration and real-time movement speed of the target robot includes: when the current driving area of ​​the target robot satisfies the flat terrain state condition, and the angle difference between the real-time posture angle of the target robot and the corresponding reference angle is higher than the cumulative number of times and / or the duration of the angle threshold exceeds a first preset continuity condition, determining that the target robot has human position intervention behavior; when the acceleration difference between the real-time acceleration of the target robot and the corresponding reference acceleration is higher than the cumulative number of times and / or the duration of the acceleration threshold exceeds a second preset continuity condition, determining that the target robot has human position intervention behavior; when the target robot is not in an emergency braking state, and the speed difference between the real-time movement speed of the target robot and the corresponding expected speed is higher than the cumulative number of times and / or the duration of the speed threshold exceeds a third preset continuity condition, determining that the target robot has human position intervention behavior.

[0096] In this embodiment, when determining that the current driving area of ​​the target robot meets the flat terrain state condition, it may specifically include: determining whether the terrain where the robot is currently located is a flat area; if yes, determining that the flat terrain state condition is met, otherwise determining that the flat terrain state condition is not met. Alternatively, determining whether the terrain where the robot is currently located is an uphill or downhill area; if no, determining that the flat terrain state condition is met, otherwise determining that the flat terrain state condition is not met.

[0097] Among them, when determining that the current driving area of ​​the target robot meets the flat terrain state condition, it can be combined with a known environment map (such as a terrain height map), or it can be determined based on the coordinate area of ​​the slope area marked on the map when the robot builds the map, combined with the robot's currently recorded position coordinates.

[0098] In this embodiment, the target robot is not in an emergency braking state, which may mean that: the target robot does not make an emergency stop, for example, does not receive an emergency stop command issued through an emergency brake button, does not generate an emergency stop command based on a collision, etc.

[0099] In this way, this embodiment takes into account uphill and downhill judgments and emergency stop judgments when determining whether the target robot has human position intervention behavior, thereby reducing the motion data fluctuations caused by the terrain being misjudged as human position intervention behavior, and reducing the emergency stop state being misjudged as human position intervention behavior, thereby further improving the accuracy of the intervention behavior judgment results.

[0100] It should also be noted that after executing step S101, if it is identified that there is human intervention in the location, step S102 is executed; if it is identified that there is no human intervention in the location, step S101 is continued to be executed.

[0101] Regarding step S102, in this step, after determining that there is human position intervention behavior, the position accuracy evaluation results of the target robot, the position evaluation results of the target robot relative to a specific area, and the preset positioning and retrieval conditions are further combined to determine whether positioning and retrieval are needed, that is, whether to execute the first positioning and retrieval strategy of this solution.

[0102] In one embodiment provided in the present application, determining whether the target robot performs a first positioning recovery based on a posture accuracy evaluation result of the target robot and a position evaluation result of the target robot relative to a specific area includes: when the posture accuracy evaluation result of the target robot indicates that the current posture does not match and the target robot is not located in the specific area, determining that the target robot performs a first positioning recovery.

[0103] Here, the pose accuracy assessment result can be determined by calculating the degree of match between the robot's current pose and the environment, and the pose accuracy assessment result can be expressed as a match score between the robot and the environment. The indication that the current pose does not match can specifically refer to: when the robot's current sensor data (determined based on sensor acquisition or algorithm estimation) has a low degree of match with the map data, it indicates that the current match cannot obtain pose data, or the obtained pose data has low credibility, and the robot may lose its position. The specific area may include an elevator area and a charging pile area, and may also include other areas.

[0104] If it is determined that the pose accuracy evaluation result of the target robot indicates that the current pose matches, then the current pose is determined to be correct, and the process returns to step S101.

[0105] If it is determined that the target robot is located in the elevator, the first positioning and retrieving operation is not performed temporarily, and step S102 is continued.

[0106] In determining whether the target robot is located in the elevator, the technical means used include but are not limited to the following methods: determination based on laser radar (LiDAR) or laser ranging, determination based on visual sensors (cameras or stereo vision), determination based on elevator door status detection, determination based on elevator control system interaction, determination based on the robot and the elevator's inertial navigation system (INS) and / or IMU monitoring results, and determination based on known maps. It can also be determined based on whether the robot is performing an elevator task. If it is in the state of performing an elevator task, the first positioning and retrieval operation will not be performed temporarily. When the robot is in the elevator, due to the relatively small environment in which the robot is located, it is difficult to locate and retrieve, the flow of people will greatly affect the positioning results, and it is not meaningful to perform positioning and retrieval at this time. Wait until the robot completes the elevator task and exits the elevator before performing the first positioning and retrieval.

[0107] If it is determined that the target robot is located in the charging pile area (specifically, it may be connected to the charging pile), a second positioning and retrieval operation may be performed, that is, the current position and posture of the target robot is updated using the charging pile position.

[0108] Among them, when determining that the target robot is located in the charging pile area, the technical means adopted include but are not limited to the following methods: determination based on physical identification of the charging pile area, determination based on communication with the charging pile system, and determination based on the monitoring results of the charging status, for example, when the robot's current automatic charging interface has an input charging current, determination is made based on sensor data.

[0109] In this way, before performing the positioning and retrieval operation, the robot's posture accuracy is comprehensively analyzed and it is determined whether the robot is in a specific area (such as an elevator or charging pile area), so that it can avoid performing an inappropriate first positioning and retrieval operation when the robot is in these specific areas and when the positioning is more accurate, thereby saving computing resources, reducing unnecessary operating costs, and improving system efficiency. And if the positioning and retrieval judgment is not performed and the positioning and retrieval operation is performed directly, frequent positioning and retrieval operations may cause the robot's task to be interrupted and affect its normal work process. Furthermore, when the identification meets certain conditions, active positioning and retrieval is performed without manual intervention at the beginning, which further improves the robot's intelligence.

[0110] Continuing with step S102, in one embodiment provided in the present application, when the sensor data collected by the robot includes laser point cloud data, the pose accuracy evaluation result of the target robot is determined by the following steps:

[0111] S1021. Acquire the current recorded position and posture of the target robot and the real-time laser point cloud data at the current recorded position and posture.

[0112] S1022: According to the currently recorded posture, the real-time laser point cloud data is matched with a known environment map to determine a posture matching score.

[0113] S1023. Determine a pose accuracy evaluation result of the target robot according to the pose matching score within a preset time period.

[0114] For step S1021, during the operation of the target robot, each recorded posture corresponds to a laser point cloud data. Here, the current recorded posture may be an accurate current posture or an erroneous current posture.

[0115] With respect to step S1022, in this step, the current position of the target robot is used to coordinately process the corresponding laser point cloud data and the known environment map, and then the unified data is matched and the error between the two is calculated; based on the calculated error, the matching score between the two is determined.

[0116] Regarding step S1023, in this step, the pose accuracy of the target robot is determined by calculating the change trend of the pose matching score within a certain period of time (eg, a few seconds, a few minutes, etc.).

[0117] If the robot's matching score remains high during this time, it means that the posture accuracy is good; if the score fluctuates greatly or is always low, it means that there may be a large error in the posture.

[0118] In this way, by matching the real-time laser point cloud data with the environment map, the current position and orientation of the robot can be determined more accurately, improving the accuracy of positioning. And by evaluating the changing trend of the pose matching score over a certain period of time, the robot's positioning stability can be fully reflected, avoiding relying solely on the matching results at a single moment, thus helping to improve the robot's autonomous navigation and operation capabilities in complex environments.

[0119] In addition, after determining that the first positioning and retrieving operation is required, the updating method further includes: determining whether the current task status of the target robot is task execution; if so, pausing the currently executed task.

[0120] Here, the task currently performed by the target robot may include, for example: a calling task, a delivery task, a cleaning task, etc.

[0121] Thus, when it is determined that the robot has lost its positioning, it may not be able to accurately perceive its current position or environment. In this case, continuing to execute the task may cause the robot to continue to move incorrectly and become more uncontrollable. Pausing the task can avoid this and ensure that the robot does not perform incorrect operations due to lost positioning. Pausing the task can also free up computing resources for the robot to recover positioning more quickly.

[0122] With respect to step S103, in this step, an example may include: when it is determined that a positioning and retrieval operation needs to be performed on the target robot, a search strategy suitable for the current state is determined (the search strategy includes an initial search step and a step adjustment strategy, and the initial search step can be determined in combination with the currently recorded position and posture of the target robot). Then, taking the recorded posture as the starting point, the search strategy is an adjustment strategy, the possible posture of the robot is adjusted according to the strategy, and real-time data of the surrounding environment is collected through its sensors (such as laser radar, visual sensor, image acquisition device, etc.) (sensor data is obtained); the real-time data is matched with the map data corresponding to the possible posture, and if the match is successful, the retrieval is successful, and the current real posture of the target robot is determined. If the match is unsuccessful, the step is adjusted, the range is expanded, and then the match is performed again.

[0123] Here, the initial search step may be set with an initial search distance and / or an initial search angle.

[0124] Step length adjustment strategy: As the search progresses, the system will adjust the search step length strategy based on the robot's feedback. For example, if the initial search range fails to successfully find the target, the system will increase the step length to expand the search area; if it gradually approaches the correct location, the step length can be reduced to improve accuracy.

[0125] For example, if the identified human intervention behavior is moving, the first initial search distance may be smaller, because the service robot is usually heavy and cannot be moved over a large range. If the identified human intervention behavior is pushing, the second initial search distance may be greater than the first initial search distance.

[0126] For example, the determined search strategy is briefly described through the following content: Assume that the current position recorded by the target robot is (x, y, z), and z is the orientation angle. At this time, the initial search step size set can be (2.5m, 10°). Then, when searching, the initial search step size is first centered on the current posture, within d = 2.5m, θ = z ± 10°, and the search is performed according to the progressive rule of increasing the distance by 0.1m or changing the angle by 1° each time. If the data obtained based on the search strategy cannot realize the positioning and retrieval of the robot, the step size adjustment strategy is executed to adjust the search step size to search again. The step size adjustment strategy can be: for the search distance in the initial search step size, increase am each time it is adjusted, and for the search angle in the initial search step size, increase b° each time it is adjusted. The values ​​of a and b can be determined adaptively. For example, when multiple searches and matches fail within the range of (2.5m, 10°), the search strategy can be adjusted to (3m, 20°) to expand the search range and improve the matching success rate.

[0127] Among them, during the search process, if only a single-frame laser point cloud data search is performed, the robot can remain still and adjust the possible posture (x, y, z) to match it with the known environment map. For example, when the possible posture is searched as (x+2.5, y+2.5, z+10°), the single-frame laser point cloud collected by the robot in real time is matched with the map data corresponding to the posture (x+2.5, y+2.5, z+10°) on the corresponding map. If the match is unsuccessful, the search strategy is adjusted.

[0128] In this way, this scheme can adjust the search accuracy and speed according to the specific situation based on the reasonable search strategy (initial search step and step adjustment strategy) determined to be suitable for the current target robot, making the search process more flexible, and collecting positioning data according to the set search strategy, and can accurately collect relevant data (sensor data) of the robot's position, so as to update the posture using sensor data, and can adapt to the robot's motion state and environmental changes in real time, further improving the accuracy and reliability of posture update.

[0129] Continuing with step S103, the sensor data may be: laser point cloud data, image data, etc. The sensor data in the embodiment of the present application adopts laser point cloud data.

[0130] In one embodiment provided in the present application, when the current posture of the target robot is updated based on the collected sensor data, the sensor data may be a frame of laser point cloud data or a point cloud subgraph composed of multiple frames of laser point cloud data.

[0131] Here, whether to use a frame of laser point cloud data to update the robot's posture or to use a point cloud sub-graph to update the robot's posture can be selected according to actual conditions.

[0132] For example, when the surrounding environment is good, such as in a hotel scene, the environment is not easy to change, and single-frame laser point cloud data is used as sensor data to update the robot's posture, so that the posture can be quickly retrieved; when the robot's environment changes, such as in a restaurant scene, the environment is easy to change, and the robot rotation is used to obtain multi-frame laser point cloud data and form a point cloud sub-graph, and then the sub-graph point cloud is used for posture retrieval, which is easier to retrieve successfully than a single-frame point cloud.

[0133] In this way, by providing two different sensor data and selecting different types of sensor data to update the robot's posture according to different situations, the present solution can be applied to a wider range of scenarios, and different types of sensor data can be used to achieve different effects. For example, using single-frame laser point cloud data can improve the retrieval efficiency, and using point cloud sub-map data can improve the retrieval success rate.

[0134] In one embodiment provided in the present application, updating the current posture of the target robot according to the current recorded posture of the target robot, sensor data, and search strategy includes:

[0135] S1041. According to the initial search step length set in the search strategy, control the target robot to perform map search and matching.

[0136] S1042: Update the current position and posture of the target robot according to the calculation result.

[0137] S1043. Adjust the initial search step length according to the step length adjustment strategy, and re-control the target robot to perform map search and matching according to the adjusted initial search step length until the current position and posture of the target robot is successfully updated or the update end condition is reached.

[0138] S1044, executing a manual posture recovery operation reminder, and determining the posture obtained based on the manual posture recovery operation as the current posture of the target robot.

[0139] For step S1041, the initial search step includes an initial moving distance and / or an initial rotation angle. This step includes: when the robot performs the search task, the moving distance and / or the moving angle of each step are determined according to the initial search step set in the search strategy, and the robot is controlled to move according to the search strategy at the issued moving speed. After the movement is completed and reaches the target sampling point, at least one positioning data (which may be current environment data) is collected at the current target sampling point according to the preset collection strategy (collecting one or more frames), and each collection obtains a set of sensor data (a frame of laser point cloud data or a point cloud sub-map composed of multiple frames of laser point cloud data). Then, based on at least one set of sensor data or all sensor data collected, and combined with the known environment map, map matching processing is performed to determine whether the current position and posture of the target robot can be calculated.

[0140] Among them, when collecting positioning data, the robot can collect it by turning in circles on the spot, or by other methods, which are not limited here.

[0141] If the current position and posture of the target robot can be calculated, it is determined that the matching is successful, otherwise it is determined that the matching fails.

[0142] If successful, execute step S1042; if failed but the update end condition is not met, execute step S1043; if failed and the update end condition is not met, execute step S1044.

[0143] The determination of calculation success and matching failure is usually determined based on certain standards and indicators, which can be determined adaptively.

[0144] Regarding step S1043, this step may be: determining the calculated position and posture as the current position of the target robot.

[0145] Regarding step S1044, this step may include: adjusting the initial search step length according to the step length adjustment strategy, and returning to execute step S1041 after the adjustment is completed.

[0146] For step S1045, when executing the manual posture recovery operation reminder, the specific moving position can be reminded, or the user can be requested to input the posture; after completion, in response to the recovery completion instruction, the posture information of the target robot is obtained, and the posture update processing is performed based on it.

[0147] For example, the reminder to perform the manual posture recovery operation may be: a pop-up window prompts the customer to push the robot to a specific area or charging station, so as to perform manual intervention to locate and retrieve the robot.

[0148] For examples, see Figure 2 , Figure 2A partial process diagram of a posture updating method provided in this application, such as Figure 2 As shown, S1: determine the search strategy; S2: control the movement of the robot according to the initial search step; S3: perform data acquisition to obtain a single-frame point cloud or a point cloud sub-graph; S4: calculate the pose; S5: determine whether the calculation is successful; if so, execute step S6, otherwise execute step S7; S7: determine whether the update end condition is met; if so, execute step S8, otherwise execute step S9; S8: manual intervention positioning retrieval; S9: adjust the initial search step according to the step adjustment strategy, and return to execute S2.

[0149] In this way, by combining search strategies, step size adjustment, sensor data, manual intervention and other means when updating the posture, an efficient, flexible and intelligent positioning data collection and update solution is provided, which can effectively improve the robot's positioning accuracy and operational reliability.

[0150] In addition, in an embodiment provided in the present application, after updating the current position and posture of the target robot, the updating method further includes: determining whether the current task status of the target robot is paused; if so, restarting the paused task.

[0151] In this way, by timely resuming the suspended task after the position and posture determination is completed and accurately determined, the smooth progress of the task can be guaranteed, and the autonomy and intelligence of the robot can be further improved.

[0152] Based on the same inventive concept, an update device corresponding to the update method is also provided in the embodiment of the present application. Since the principle of solving the problem by the device in the embodiment of the present application is similar to the above-mentioned update method in the embodiment of the present application, the implementation of the device can refer to the implementation of the method, and the repeated parts will not be repeated.

[0153] See also Figure 3 , Figure 4 , Figure 3 This is one of the structural schematic diagrams of a robot posture updating device provided in an embodiment of the present application, Figure 4 This is a second structural diagram of a robot posture updating device provided in an embodiment of the present application. Figure 3 As shown in , the updating device 300 includes:

[0154] A behavior recognition module 310 is used to identify whether the target robot has human position intervention behavior according to at least one of the real-time posture angle, real-time acceleration and real-time movement speed of the target robot;

[0155] A positioning and retrieval judgment module 320 is used to determine whether the target robot should perform a first positioning and retrieval according to a position and posture accuracy evaluation result of the target robot and a position evaluation result of the target robot relative to a specific area if human position intervention behavior is identified;

[0156] The updating module 330 is used to update the current posture of the target robot according to the current recorded posture, sensor data and search strategy of the target robot if it is determined that the target robot performs the first positioning and retrieval; wherein the search strategy is provided with an initial search step length and a step length adjustment strategy.

[0157] Optionally, when the behavior recognition module 310 is used to identify whether the target robot has human position intervention behavior according to at least one of the real-time posture angle, real-time acceleration, and real-time movement speed of the target robot, the behavior recognition module 310 is used to:

[0158] When the angle difference between the real-time posture angle of the target robot and the corresponding reference angle is higher than the angle threshold for a cumulative number of times and / or a duration that exceeds a first preset persistence condition, it is determined that the target robot has human position intervention behavior;

[0159] And / or, when the acceleration difference between the real-time acceleration of the target robot and the corresponding reference acceleration is higher than the acceleration threshold value for a cumulative number of times and / or the duration exceeds a second preset persistence condition, it is determined that the target robot has human position intervention behavior;

[0160] And / or, when the speed difference between the real-time movement speed of the target robot and the corresponding expected speed is higher than the speed threshold for a cumulative number of times and / or the duration exceeds a third preset continuity condition, it is determined that there is human position intervention behavior in the target robot.

[0161] Optionally, the sensor data includes laser point cloud data, and the positioning and retrieval judgment module 320 is further used to determine the position and posture accuracy evaluation result of the target robot through the following steps:

[0162] Acquire the current recorded position and posture of the target robot and the real-time laser point cloud data at the current recorded position and posture;

[0163] According to the currently recorded posture, matching processing is performed using the real-time laser point cloud data and a known environment map to determine a posture matching score;

[0164] The pose accuracy evaluation result of the target robot is determined according to the pose matching score within a preset time length.

[0165] Optionally, when the positioning and retrieval judgment module 320 is used to determine whether the target robot performs a first positioning and retrieval according to a posture accuracy evaluation result of the target robot and a position evaluation result of the target robot relative to a specific area, the positioning and retrieval judgment module 320 is used to:

[0166] When the pose accuracy evaluation result of the target robot indicates that the current pose does not match and the target robot is not located in the specific area, it is determined that the target robot performs a first positioning retrieval; the specific area includes an elevator area and a charging pile area.

[0167] Optionally, when the updating module 330 is used to update the current posture of the target robot according to the current recorded posture of the target robot, sensor data and search strategy, the updating module 330 is used to:

[0168] According to the initial search step length set in the search strategy, the target robot is controlled to perform map search and matching; wherein the initial search step length includes an initial moving distance and / or an initial rotation angle;

[0169] If the match is successful, the current position and posture of the target robot is updated according to the calculation result;

[0170] If the match fails but the update end condition is not met, the initial search step size is adjusted according to the step size adjustment strategy, and the target robot is re-controlled to perform map search and matching according to the adjusted initial search step size until the match is successfully updated to update the current position of the target robot or the update end condition is met;

[0171] If the matching fails and the update end condition is met, an artificial posture recovery operation reminder is executed, and the posture obtained based on the artificial posture recovery operation is determined as the current posture of the target robot.

[0172] Optional, such as Figure 4 As shown, the updating device 300 further includes a pause module 340, and the pause module 340 is used to:

[0173] After determining that the target robot needs to perform a first positioning and retrieval, determine whether the current task state of the target robot is task execution; if so, suspend the currently executed task.

[0174] The updating device 300 further includes a restart module 350, and the restart module 350 is used to:

[0175] After updating the current position and posture of the target robot, determine whether the current task status of the target robot is paused; if so, restart the paused task.

[0176] Optionally, the sensor data is a frame of laser point cloud data, or a point cloud sub-image composed of multiple frames of laser point cloud data.

[0177] Optionally, when the behavior recognition module 310 is used to identify whether the target robot has human position intervention behavior according to at least one of the real-time posture angle, real-time acceleration, and real-time movement speed of the target robot, the behavior recognition module 310 is used to:

[0178] When the current driving area of ​​the target robot meets the flat terrain state condition, and the cumulative number of times and / or the duration of the angle difference between the real-time posture angle of the target robot and the corresponding reference angle is higher than the angle threshold exceeds the first preset persistence condition, it is determined that the target robot has human position intervention behavior;

[0179] When the acceleration difference between the real-time acceleration of the target robot and the corresponding reference acceleration is higher than the acceleration threshold value for a cumulative number of times and / or the duration exceeds a second preset persistence condition, it is determined that the target robot has human position intervention behavior;

[0180] When the target robot is not in an emergency braking state, and the speed difference between the real-time movement speed of the target robot and the corresponding expected speed is higher than the speed threshold for a cumulative number of times and / or a duration that exceeds a third preset continuity condition, it is determined that the target robot has human position intervention behavior.

[0181] See also Figure 5 , Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. Figure 5 As shown in , the electronic device 500 includes a processor 510 , a memory 520 and a bus 530 .

[0182] The memory 520 stores machine-readable instructions executable by the processor 510. When the electronic device 500 is running, the processor 510 communicates with the memory 520 via the bus 530. When the machine-readable instructions are executed by the processor 510, the above-mentioned Figure 1 as well as Figure 2 The specific implementation of the steps in the method embodiment shown can be found in the method embodiment, and will not be repeated here.

[0183] The present application also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the computer program can execute the above-mentioned Figure 1 as well as Figure 2 The specific implementation of the steps in the method embodiment shown can be found in the method embodiment, and will not be repeated here.

[0184] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0185] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of devices or units can be electrical, mechanical or other forms.

[0186] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0187] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0188] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a non-volatile computer-readable storage medium that is executable by a processor. Based on this understanding, the technical solution of the present application can essentially be embodied in the form of a software product, or in other words, the part that contributes to the prior art or the part of the technical solution. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0189] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present application, which are used to illustrate the technical solutions of the present application, rather than to limit them. The protection scope of the present application is not limited thereto. Although the present application is described in detail with reference to the above-mentioned embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-mentioned embodiments within the technical scope disclosed in the present application, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A method for updating a robot posture, characterized in that: The updating method comprises: Identify whether there is human position intervention behavior on the target robot according to at least one of the real-time posture angle, real-time acceleration, and real-time motion speed of the target robot; If human position intervention behavior is identified, determining whether the target robot should perform a first positioning retrieval according to a position and posture accuracy evaluation result of the target robot and a position evaluation result of the target robot relative to a specific area; If it is determined that the target robot performs the first positioning and retrieval, the current posture of the target robot is updated according to the current recorded posture of the target robot, sensor data and search strategy; wherein the search strategy is provided with an initial search step length and a step length adjustment strategy.

2. The updating method according to claim 1, characterized in that: The step of identifying whether the target robot has human position intervention behavior according to at least one of the real-time posture angle, the real-time acceleration, and the real-time movement speed of the target robot comprises: When the angle difference between the real-time posture angle of the target robot and the corresponding reference angle is higher than the angle threshold for a cumulative number of times and / or a duration that exceeds a first preset persistence condition, it is determined that the target robot has human position intervention behavior; And / or, when the acceleration difference between the real-time acceleration of the target robot and the corresponding reference acceleration is higher than the acceleration threshold value for a cumulative number of times and / or the duration exceeds a second preset persistence condition, it is determined that the target robot has human position intervention behavior; And / or, when the speed difference between the real-time movement speed of the target robot and the corresponding expected speed is higher than the speed threshold for a cumulative number of times and / or the duration exceeds a third preset continuity condition, it is determined that there is human position intervention behavior in the target robot.

3. The updating method according to claim 1, characterized in that: The sensor data includes laser point cloud data, and the position and posture accuracy evaluation result of the target robot is determined by the following steps: Acquire the current recorded position and posture of the target robot and the real-time laser point cloud data at the current recorded position and posture; According to the currently recorded posture, matching processing is performed using the real-time laser point cloud data and a known environment map to determine a posture matching score; The pose accuracy evaluation result of the target robot is determined according to the pose matching score within a preset time length.

4. The updating method according to claim 1, characterized in that: The step of determining whether to perform a first positioning retrieval of the target robot according to a result of evaluating the accuracy of the position and posture of the target robot and a result of evaluating the position of the target robot relative to a specific area comprises: When the pose accuracy evaluation result of the target robot indicates that the current pose does not match and the target robot is not located in the specific area, it is determined that the target robot performs a first positioning retrieval; the specific area includes an elevator area and a charging pile area.

5. The updating method according to claim 3, characterized in that: The updating of the current posture of the target robot according to the current recorded posture of the target robot, sensor data and search strategy includes: According to the initial search step length set in the search strategy, the target robot is controlled to perform map search and matching; wherein the initial search step length includes an initial moving distance and / or an initial rotation angle; If the match is successful, the current position and posture of the target robot is updated according to the calculation result; If the match fails but the update end condition is not met, the initial search step size is adjusted according to the step size adjustment strategy, and the target robot is re-controlled to perform map search and matching according to the adjusted initial search step size until the match is successfully updated to update the current position of the target robot or the update end condition is met; If the matching fails and the update end condition is met, an artificial posture recovery operation reminder is executed, and the posture obtained based on the artificial posture recovery operation is determined as the current posture of the target robot.

6. The updating method according to claim 1, characterized in that: After determining that the target robot needs to perform a first positioning and retrieval, the updating method further includes: Determine whether the current task status of the target robot is task execution; If so, suspend the currently executed task; After updating the current posture of the target robot, the updating method further includes: Determine whether the current task status of the target robot is paused; If yes, restart the suspended task.

7. The updating method according to claim 3, characterized in that: The sensor data is a frame of laser point cloud data, or a point cloud sub-image composed of multiple frames of laser point cloud data.

8. The updating method according to claim 2, characterized in that: The step of identifying whether the target robot has human position intervention behavior according to at least one of the real-time posture angle, the real-time acceleration, and the real-time movement speed of the target robot comprises: When the current driving area of ​​the target robot meets the flat terrain state condition, and the cumulative number of times and / or the duration of the angle difference between the real-time posture angle of the target robot and the corresponding reference angle is higher than the angle threshold exceeds the first preset persistence condition, it is determined that the target robot has human position intervention behavior; When the acceleration difference between the real-time acceleration of the target robot and the corresponding reference acceleration is higher than the acceleration threshold value for a cumulative number of times and / or the duration exceeds a second preset persistence condition, it is determined that the target robot has human position intervention behavior; When the target robot is not in an emergency braking state, and the speed difference between the real-time movement speed of the target robot and the corresponding expected speed is higher than the speed threshold for a cumulative number of times and / or a duration that exceeds a third preset continuity condition, it is determined that the target robot has human position intervention behavior.

9. A robot posture updating device, characterized in that: The updating device comprises: A behavior recognition module, used to identify whether the target robot has human position intervention behavior based on at least one of the real-time posture angle, real-time acceleration and real-time movement speed of the target robot; A positioning and retrieval judgment module is used to determine whether the target robot should perform a first positioning and retrieval according to a position and posture accuracy evaluation result of the target robot and a position evaluation result of the target robot relative to a specific area if human position intervention behavior is identified; An updating module is used to update the current posture of the target robot according to the current recorded posture, sensor data and search strategy of the target robot if it is determined that the target robot performs the first positioning retrieval; wherein the search strategy is provided with an initial search step length and a step length adjustment strategy.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the updating method according to any one of claims 1 to 8 are executed.

Images