Robot operation method and device, computer equipment and storage medium

Abstract

The invention relates to a robot operation method and device, computer equipment and a storage medium. The method comprises the steps that environment data and operation data collected on a current contact surface in the operation process of a target robot are acquired; according to the environment data and the operation data, determining at least one effective contact point of the target robot on the future contact surface; according to a projection area defined by each effective contact point on the future contact surface, a supporting area of the target robot is determined; determining the stability margin of the target robot according to the supporting area, the operation data and the unit normal vector of each effective contact point; and according to the size relationship between the stability margin and a preset margin threshold value, adjusting the operation strategy of the target robot. By adopting the method, the target robot can be effectively controlled to autonomously run in a complex unstructured environment.

Description

Technical Field

[0001] This application relates to the field of robotics, and in particular to a robot operation method, apparatus, computer equipment, and storage medium. Background Technology

[0002] With the rapid development of robotics technology, all-terrain robots are being used more and more widely in tasks such as rubble exploration and disaster relief. These tasks typically involve extremely unstructured terrains covered with random rubble, soft soil, and loose rocks, requiring robots to have extremely high autonomy and mobility.

[0003] In traditional technologies, two approaches are typically used to address the obstacle-crossing challenges of complex terrain: one is to passively adapt to the terrain by strengthening the mechanical structure (such as increasing suspension travel or using large-diameter wheels); the other is to implement conservative active control strategies, such as slowing down or stopping the engine when excessive body tilt angle or wheel slippage is detected.

[0004] However, current traditional methods suffer from slow response and low efficiency. While reinforcing the mechanical structure can improve physical clearance limits, it cannot intelligently predict sudden risks of skidding (wheels losing support) or vehicle rollover. Furthermore, passive response control based on the current state often detects instability when the vehicle is already on the verge of danger, or even after an accident has occurred. Therefore, existing methods cannot proactively identify and intervene before risks occur, making it difficult to achieve efficient autonomous navigation of robots on complex, unstructured surfaces while ensuring safety. Summary of the Invention

[0005] Therefore, it is necessary to provide a robot operation method, device, computer equipment, and storage medium that can effectively control the target robot to drive autonomously in complex unstructured environments, addressing the aforementioned technical problems.

[0006] In a first aspect, this application provides a method for operating a robot, including:

[0007] Acquire environmental and operational data collected on the current contact surface during the operation of the target robot;

[0008] Based on environmental and operational data, determine at least one effective contact point on the future contact surface of the target robot; the future contact surface is the contact surface of the target robot within a preset forward distance; the effective contact point is the point where there is contact between the wheel of the target robot and the future contact surface and it bears the load.

[0009] The support area of ​​the target robot is determined based on the projected area enclosed by each effective contact point on the future contact surface; the support area is used to characterize the area required for the robot to maintain stability on the future contact surface.

[0010] Based on the support area, operational data, and unit normal vectors of each effective contact point, the stability margin of the target robot is determined; the stability margin is used to characterize the degree to which the robot will become airborne or overturned on the contact surface in the future.

[0011] The operating strategy of the target robot is adjusted based on the relationship between the stability margin and the preset margin threshold.

[0012] In one embodiment, the operational data includes the target robot's roll angle, pitch angle, and operational acceleration; based on the support area, operational data, and unit normal vectors of each effective contact point, the stability margin of the target robot is determined, including:

[0013] The center of gravity offset of the target robot is determined based on the shortest distance from the center of mass projection point of the target robot to the support area, and the target normal load of the target robot is determined based on the running acceleration and the unit normal vector of each effective contact point.

[0014] The stability margin of the target robot is determined based on the center of gravity offset, roll angle, pitch angle, and target normal load.

[0015] In one embodiment, the target normal load of the target robot is determined based on the unit normal vector and running acceleration of each effective contact point, including:

[0016] Obtain the quality of the target robot;

[0017] Based on the mass, running acceleration, and unit normal vector of each effective contact point, determine the wheel normal load of at least one wheel in the target robot;

[0018] Select the smallest wheel normal load from the wheel normal loads of at least one wheel as the target normal load of the target robot.

[0019] In one embodiment, the operational data includes the lowest point of at least one wheel of the target robot; based on environmental data and operational data, determining at least one effective contact point on the future contact surface of the target robot includes:

[0020] Based on environmental data, determine the obstacle height range on the future contact surface of the target robot;

[0021] The reference obstacle height is determined based on the relationship between the width of the obstacle height range and the preset width threshold.

[0022] Based on the reference obstacle height and the preset safety threshold, at least one effective contact point is selected from the lowest point of each wheel.

[0023] In one embodiment, determining a reference obstacle height based on the relationship between the width of the obstacle height range and a preset width threshold includes:

[0024] If the width of the obstacle height range is less than the preset width threshold, the midpoint of the obstacle height range will be used as the reference obstacle height.

[0025] If the width of the obstacle height range is not less than the preset width threshold, the maximum value in the obstacle height range will be used as the reference obstacle height.

[0026] In one embodiment, based on a reference obstacle height and a preset safety threshold, at least one effective contact point is selected from the lowest point of each wheel, including:

[0027] The equivalent obstacle height is determined based on the difference between the reference obstacle height and the preset safety threshold.

[0028] For each wheel's lowest point, if the height between the wheel's lowest point and the future contact surface is less than the equivalent obstacle height, the wheel's lowest point is determined as the effective contact point.

[0029] Secondly, this application also provides a robot operating device, comprising:

[0030] The acquisition module is used to acquire environmental and operational data collected on the current contact surface during the operation of the target robot;

[0031] The contact module is used to determine at least one effective contact point on the future contact surface of the target robot based on environmental data and operational data; the future contact surface is the contact surface of the target robot within a preset forward distance; the effective contact point is the point where there is contact between the wheel of the target robot and the future contact surface and the point where it bears the load.

[0032] The support module is used to determine the support area of ​​the target robot based on the projected area enclosed by each effective contact point on the future contact surface; the support area is used to characterize the area required for the robot to maintain stability on the future contact surface.

[0033] The margin module is used to determine the stability margin of the target robot based on the support area, operating data, and unit normal vectors of each effective contact point; the stability margin is used to characterize the degree to which the robot will become airborne or overturned on the contact surface in the future.

[0034] The control module is used to adjust the operating strategy of the target robot based on the relationship between the stability margin and the preset margin threshold.

[0035] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0036] Acquire environmental and operational data collected on the current contact surface during the operation of the target robot;

[0037] Based on environmental and operational data, determine at least one effective contact point on the future contact surface of the target robot; the future contact surface is the contact surface of the target robot within a preset forward distance; the effective contact point is the point where there is contact between the wheel of the target robot and the future contact surface and it bears the load.

[0038] The support area of ​​the target robot is determined based on the projected area enclosed by each effective contact point on the future contact surface; the support area is used to characterize the area required for the robot to maintain stability on the future contact surface.

[0039] Based on the support area, operational data, and unit normal vectors of each effective contact point, the stability margin of the target robot is determined; the stability margin is used to characterize the degree to which the robot will become airborne or overturned on the contact surface in the future.

[0040] The operating strategy of the target robot is adjusted based on the relationship between the stability margin and the preset margin threshold.

[0041] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0042] Acquire environmental and operational data collected on the current contact surface during the operation of the target robot;

[0043] Based on environmental and operational data, determine at least one effective contact point on the future contact surface of the target robot; the future contact surface is the contact surface of the target robot within a preset forward distance; the effective contact point is the point where there is contact between the wheel of the target robot and the future contact surface and it bears the load.

[0044] The support area of ​​the target robot is determined based on the projected area enclosed by each effective contact point on the future contact surface; the support area is used to characterize the area required for the robot to maintain stability on the future contact surface.

[0045] Based on the support area, operational data, and unit normal vectors of each effective contact point, the stability margin of the target robot is determined; the stability margin is used to characterize the degree to which the robot will become airborne or overturned on the contact surface in the future.

[0046] The operating strategy of the target robot is adjusted based on the relationship between the stability margin and the preset margin threshold.

[0047] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:

[0048] Acquire environmental and operational data collected on the current contact surface during the operation of the target robot;

[0049] Based on environmental and operational data, determine at least one effective contact point on the future contact surface of the target robot; the future contact surface is the contact surface of the target robot within a preset forward distance; the effective contact point is the point where there is contact between the wheel of the target robot and the future contact surface and it bears the load.

[0050] The support area of ​​the target robot is determined based on the projected area enclosed by each effective contact point on the future contact surface; the support area is used to characterize the area required for the robot to maintain stability on the future contact surface.

[0051] Based on the support area, operational data, and unit normal vectors of each effective contact point, the stability margin of the target robot is determined; the stability margin is used to characterize the degree to which the robot will become airborne or overturned on the contact surface in the future.

[0052] The operating strategy of the target robot is adjusted based on the relationship between the stability margin and the preset margin threshold.

[0053] The aforementioned robot operation method, device, computer equipment, and storage medium, by integrating environmental perception and its own state data, proactively predict the robot's effective contact points on future contact surfaces. Based on this, a dynamic support area and quantified stability margin are constructed, enabling advanced identification and active intervention of risks of overturning or tipping over. Compared to traditional control methods that rely on passive responses to the current state, this solution can comprehensively assess the robot's attitude, load, and support boundaries before the actual occurrence of a risk, thereby adjusting the operating strategy in advance and effectively avoiding vehicle loss of control or accidents due to delayed response. Furthermore, by combining multi-contact point constraints with dynamic stability margins, this solution can accurately identify instability trends such as wheel slippage and vehicle tilting in complex unstructured environments, and implement differentiated control strategies according to different risk levels. This significantly improves the robot's autonomous passage safety and obstacle-crossing efficiency in extreme terrains such as ruins and rubble, achieving a dynamic balance between safety and maneuverability. Attached Figure Description

[0054] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0055] Figure 1 This is an application environment diagram of a robot operation method provided in this embodiment;

[0056] Figure 2 This is a flowchart illustrating the first robot operation method provided in this embodiment;

[0057] Figure 3 A flowchart illustrating a step for determining stability margin is provided in this embodiment;

[0058] Figure 4 This is a flowchart illustrating the steps for determining a valid contact point in this embodiment.

[0059] Figure 5 This is a structural block diagram of a robot operating device provided in this embodiment;

[0060] Figure 6 This is an internal structural diagram of a computer device provided in this embodiment. Detailed Implementation

[0061] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0062] The robot operation method provided in this application embodiment can be applied to, for example... Figure 1In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be integrated onto server 104 or placed on a cloud or other network server. The computer device acquires environmental and operational data collected on the current contact surface during the target robot's operation; based on the environmental and operational data, it determines at least one effective contact point on the future contact surface of the target robot; the future contact surface is the contact surface of the target robot within a preset forward distance; an effective contact point is the point where the wheels of the target robot make contact with the future contact surface and bear the load; based on the projection area enclosed by each effective contact point on the future contact surface, it determines the support area of ​​the target robot; the support area characterizes the area required for the robot to maintain stability on the future contact surface; based on the support area, operational data, and the unit normal vector of each effective contact point, it determines the stability margin of the target robot; the stability margin characterizes the degree to which the robot may become airborne or overturned on the future contact surface; based on the relationship between the stability margin and a preset margin threshold, it adjusts the target robot's operating strategy. The computer device can be a terminal installed in the target robot or a server. Terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, smart in-vehicle devices, etc. Portable wearable devices can include smartwatches, smart bracelets, head-mounted devices, etc. Server 104 can be implemented using a standalone server or a server cluster consisting of multiple servers.

[0063] In one exemplary embodiment, such as Figure 2 As shown, a robot operation method is provided, which can be applied to... Figure 1 Taking a computer device as an example, the explanation includes the following steps S201 to S205. Wherein:

[0064] S201 acquires environmental and operational data collected on the current contact surface during the operation of the target robot.

[0065] The target robot refers to the entity executing the operational strategy adjustment method, i.e., the robotic system that needs stability assessment and control while navigating in complex unstructured environments (such as ruins or rubble). The current contact surface refers to the physical surface (such as the ground or rocks) on which the target robot's wheels are actually in contact and bearing load at the current moment. It serves as a reference surface for collecting environmental and operational data. Environmental data refers to information about the surrounding terrain and obstacles collected in real time by sensors (such as LiDAR and depth cameras) on the target robot. Specifically, this includes geometric features on the future contact surface, information used to determine the height range of obstacles, etc. Operational data refers to data describing the target robot's current motion state. Specifically, this includes: roll angle (the body's tilt angle around the forward direction), pitch angle (the body's forward and backward pitch angle around the lateral axis), operational acceleration (triaxial acceleration), and at least one wheel's lowest point (the spatial location of the wheel's circumference at its farthest point from the vehicle body or the point of contact).

[0066] In some embodiments, environmental data and operational data collected on the current contact surface are acquired during the operation of the target robot.

[0067] For example, environmental data can be collected using LiDAR and / or depth cameras to perform real-time scanning of the area in front of the all-terrain robot and the wheel tracks, laying the data foundation for subsequent obstacle information extraction; then, for each wheel of the robot, the obstacle height information within a predetermined distance in front of it is extracted independently to ensure that the obstacle data corresponding to each wheel is accurate and independent.

[0068] For example, operational data can be fused with an inertial measurement unit (IMU) and an odometry to predict the state information of an all-terrain robot at each future moment during its journey over a preset distance. This state information includes key parameters such as position, velocity, attitude angles (including roll, pitch, and yaw), and angular velocity. The IMU can frequently collect angular and linear acceleration data from the all-terrain robot, allowing for rapid calculation of dynamic information such as angular velocity and attitude angles. It offers fast response and timely capture of changes in the robot's motion state, but suffers from integral drift, which can lead to accumulated errors in position and velocity parameters over long-term use alone. The odometry, on the other hand, primarily obtains travel distance and velocity data by measuring wheel rotation. While its position measurement exhibits good short-term stability and relatively slow error growth, it is sensitive to uneven surfaces and wheel slippage, making it prone to localized data distortion and unable to directly provide attitude angle information.

[0069] It should be noted that, to achieve accurate prediction, a data fusion algorithm is needed to complement the strengths of both sensors and offset their respective weaknesses. Specifically, the raw data collected by the two sensors are first preprocessed, including zero-bias calibration and noise removal for IMU data, and slippage detection and correction for odometer data, to ensure the accuracy of the data input to the fusion system. Subsequently, based on a preset fusion model (such as Kalman filtering and its improved algorithms, extended Kalman filtering, etc.), the angular velocity and attitude angle data obtained from the IMU are fused with the speed and distance data output by the odometer: using the high-frequency dynamic data provided by the IMU as a basis, the attitude changes and instantaneous motion state of the all-terrain robot are updated in real time, compensating for the odometry's shortcomings in attitude perception and dynamic response; at the same time, using the odometry's position and velocity data as a reference, the integral drift of the IMU is corrected to suppress error accumulation and ensure the long-term stability of position and velocity parameters.

[0070] Ultimately, the system constructs and updates the all-terrain robot's state information in real time through fusion computation. This information comprehensively covers the key information needed for prediction, including the robot's position, velocity, attitude angles (roll, pitch, and yaw), and angular velocity. During the all-terrain robot's movement within a preset distance, the fusion system continuously receives real-time data from both sensors, iteratively optimizing the state information to achieve accurate predictions of the robot's state at each future moment. This provides a reliable basis for decision-making regarding the robot's path planning and motion control.

[0071] Based on environmental and operational data, S202 determines at least one effective contact point on the future contact surface of the target robot.

[0072] The future contact surface is the contact surface of the target robot within a preset forward distance; the effective contact point is the point where the wheel of the target robot makes contact with the future contact surface and bears the load.

[0073] In one alternative embodiment, an effective contact point determination model is used to determine at least one effective contact point on the future contact surface of the target robot based on environmental data and operational data.

[0074] In one alternative embodiment, at least one reference contact point on the future contact surface of the target robot is determined based on environmental data and operational data; the point among the reference contact points that bears the load is taken as the effective contact point.

[0075] S203 determines the support area of ​​the target robot based on the projection area enclosed by each effective contact point on the future contact surface.

[0076] The support region is used to characterize the area required for the robot to maintain stability on the future contact surface.

[0077] In some embodiments, the projection area enclosed by each effective contact point on the future contact surface is determined; the projection area is used as the support area of ​​the target robot.

[0078] For example, all valid contact points can be projected onto a horizontal plane, and the minimum convex hull formed by these projected points can be calculated. This convex hull is the support region used for stability analysis. For instance, its projection onto the ground forms a dynamic support polygon S: ;in, c i Let x be the lowest point of the i-th wheel. i Let y be the lateral coordinate of the lowest point of the i-th wheel. i Let z be the longitudinal coordinate of the lowest point of the i-th wheel. i Let the height coordinates of the lowest point of the i-th wheel be denoted as . Here, is the projection mapping function, and Convex Hull() is the convex hull function.

[0079] S204 determines the stability margin of the target robot based on the support area, operating data, and the unit normal vector of each effective contact point.

[0080] The stability margin characterizes the degree to which the robot may become airborne or tip over on the future contact surface. A unit normal vector is a vector perpendicular to a surface (or curve) with a magnitude (length) of 1.

[0081] In some embodiments, a stability margin determination model is used to determine the stability margin of the target robot based on the support area, operational data, and unit normal vectors of each effective contact point.

[0082] S205 adjusts the target robot's operating strategy based on the relationship between the stability margin and the preset margin threshold.

[0083] The operating strategy refers to the set of control decisions or instructions made by the computer equipment (or controller) based on the comparison between the calculated stability margin and the preset margin threshold, in order to maintain the stability of the robot and avoid it from falling or tipping over.

[0084] In some embodiments, when the stability margin is less than a preset margin threshold, the operating strategy of the target robot is adjusted. The preset margin threshold includes a preset overhead threshold and / or a preset overturning threshold, and different preset thresholds correspond to different levels of control. The preset overhead threshold is an overhead critical value, and the preset overturning threshold is an overturning critical value.

[0085] For example, when the preset overturning threshold is less than the stability margin and the preset levitation threshold is less than the preset levitation threshold, the future contact surface of the target robot corresponding to the stability margin is determined to be the levitation critical region; when the stability margin is less than the preset overturning threshold, the future contact surface of the target robot corresponding to the stability margin is determined to be the overturning critical region.

[0086] If the future contact surface is determined to be a critical area for suspension or overturning, intervention can be made in advance to proactively adjust the control operations of the target robot. Control operations based on stability margin for the target robot's state include at least one of the following: active speed limiting, wheel-end torque pulse control, and attitude angle closed-loop constraints.

[0087] It should be noted that the different control levels corresponding to different preset thresholds include: (1) different types of control operations. For example, the control operations corresponding to the preset overhead threshold may include active limiting and wheel-end torque pulse control, while the control operations corresponding to the preset overhead threshold include active speed limiting, wheel-end torque pulse control and attitude angle closed-loop constraint. (2) different operation levels of the same control operation. For example, the speed after speed limiting corresponding to the preset overhead threshold is greater than the speed after speed limiting corresponding to the preset rollover threshold.

[0088] In one possible implementation, active speed limiting includes: determining a target speed according to a preset speed adjustment formula; the preset speed adjustment formula includes:

[0089]

[0090] in, Indicates the target speed. Indicates reference speed. The saturation function is represented by DSM, and the stability margin is represented by DSM. This indicates the speed adjustment threshold.

[0091] In one possible implementation, wheel-end torque pulse control includes: determining the wheel torque according to a preset torque formula; the preset torque formula includes:

[0092]

[0093] Among them, among them, This represents the torque of the i-th wheel at time t. Indicates the nominal torque. Indicates the amount of torque adjustment. Indicates an indicator function, This represents the normal load on the i-th wheel. Indicates the normal load threshold. This represents the torque adjustment gain coefficient.

[0094] In one possible implementation, the attitude angle closed-loop constraint includes: pitch angle closed-loop constraint and / or roll angle closed-loop constraint. The pitch angle closed-loop constraint includes: first calculating the pitch attitude control variable:

[0095]

[0096] in, This indicates the pitch attitude control value. Represents the proportional gain coefficient. This represents the reference pitch angle (target value, usually given by a higher-level decision maker or stability algorithm). This represents the actual pitch angle (measured value). Represents the differential gain coefficient. This represents the reference pitch rate (target value). This represents the actual pitch angular velocity (measured value).

[0097] Then, the pitch attitude control values ​​are mapped to the differential torque of the left and right wheels: .in, This indicates the driving torque of the left wheel. This indicates the driving torque of the right wheel. This represents the torque mapping coefficient.

[0098] Roll angle closed-loop constraints include: first, calculating the roll attitude control variables:

[0099]

[0100] in, This indicates the amount of roll attitude control. Represents the proportional gain coefficient. Indicates the reference roll angle. This indicates the actual roll angle (measured value). Represents the differential gain coefficient. Indicates the reference roll rate. This indicates the actual roll rate.

[0101] Then, the roll attitude control quantity is mapped to the suspension height difference or the difference in braking force between the left and right sides: Suspension height difference or difference in braking force between the left and right sides = ;in, This represents the torque mapping coefficient.

[0102] The aforementioned robot operation method, by integrating environmental perception and its own state data, proactively predicts the robot's effective contact points on future contact surfaces and constructs a dynamic support area and quantified stability margin accordingly. This enables advanced identification and proactive intervention of risks of scrambling or tipping over. Compared to traditional control methods that rely on passive responses to the current state, this solution can comprehensively assess the robot's attitude, load, and support boundaries before the risk actually occurs, thereby adjusting the operating strategy in advance and effectively avoiding vehicle loss of control or accidents due to delayed response. Furthermore, by combining multi-contact point constraints with dynamic stability margins, this solution can accurately identify instability trends such as wheel slippage and vehicle tilting in complex unstructured environments and implement differentiated control strategies according to different risk levels. This significantly improves the robot's autonomous passage safety and obstacle-crossing efficiency in extreme terrains such as ruins and rubble, achieving a dynamic balance between safety and maneuverability.

[0103] Figure 3 This is a flowchart illustrating the steps for determining the stability margin in one embodiment. In this embodiment, the operational data includes the target robot's roll angle, pitch angle, and acceleration. Based on the support area, operational data, and the unit normal vector of each effective contact point, the stability margin of the target robot is determined, including the following steps:

[0104] S301 determines the center of gravity offset of the target robot based on the shortest distance from the center of mass projection point of the target robot to the support area; and determines the target normal load of the target robot based on the running acceleration and the unit normal vector of each effective contact point.

[0105] The center of mass projection refers to the vertical projection of the target robot's center of gravity (center of mass) onto a horizontal plane (or the plane containing the support area). It simplifies the robot's three-dimensional spatial position to its landing point on a two-dimensional plane. It is used to determine whether the line of action of gravity falls within the stable support area.

[0106] The shortest distance refers to the shortest geometric distance from the centroid projection point to the boundary of the support area (the polygon formed by the effective contact points). This is the core geometric indicator for measuring stability. The larger the distance, the closer the center of gravity is to the center of the support area, and the more stable the robot is; the smaller the distance, the closer the center of gravity is to the edge, and the more prone it is to tipping over.

[0107] The center of gravity offset quantifies the degree of deviation of the current center of gravity from the stable support foundation. If this value is 0 or negative (i.e., the projection point falls outside the support area), theoretically the robot has already begun to tip over.

[0108] Among them, motion acceleration refers to the instantaneous acceleration vector measured by sensors such as inertial measurement units (IMU) during the target robot's movement. It typically consists of three dimensions: the vector sum of gravitational acceleration and motion acceleration (such as the inertial forces generated during starting, braking, and turning).

[0109] The target normal load refers to the minimum normal load on the wheel calculated based on the running acceleration and the geometric characteristics of the contact point. This minimum value represents the critical point at which the wheel is about to leave the ground (levitate) or lose traction. It is a key indicator for determining whether the robot is at risk of levitation.

[0110] In some embodiments, the centroid projection point of the target robot is determined; the shortest distance from the centroid projection point of the target robot to the support area is determined; the shortest distance from the centroid projection point of the target robot to the support area is used as the centroid offset of the target robot; and the target normal load of the target robot is determined based on the running acceleration and the unit normal vector of each effective contact point.

[0111] For example, the method to determine the center of mass projection point of the target robot is as follows: using the angular velocity and acceleration measured by the IMU, combined with the kinematic model, to predict the movement trajectory of the center of mass projection over a short period of time (the time corresponding to the preset distance), that is, the center of mass projection trajectory.

[0112] The calculation method for each centroid projection point includes: knowing the fixed position of the all-terrain robot's centroid in the vehicle coordinate system. Transform the robot's pose to the world coordinate system: Take the projection of the above position onto the horizontal plane: ;in, This represents the position of the centroid of the all-terrain robot in the world coordinate system at time t; This represents the position vector of the all-terrain robot in the world coordinate system, that is, the coordinates of the origin of the vehicle coordinate system in the world coordinate system; This represents the rotation matrix, determined by the attitude angle, used to transform vectors in the vehicle coordinate system to the world coordinate system; Indicates the projection operator; This represents the coordinates of the centroid projection point at time t.

[0113] For example, the method for determining the shortest distance from the target robot's center of mass projection point to the support area is as follows: obtain the two-dimensional coordinates of the center of mass projection point, and use a geometric algorithm (such as calculating the perpendicular distance from the point to the straight line containing each edge in the support area, and determining whether the foot of the perpendicular falls within the edge segment) to solve for the shortest distance from the point to each edge of the support area, and take the minimum value as the shortest distance. This distance directly reflects the degree of offset of the target robot's center of mass relative to the stable support area, and is a core quantitative indicator for judging whether the entire vehicle is approaching the rollover boundary, providing a key spatial relationship input for the subsequent calculation of the comprehensive dynamic stability margin.

[0114] In one alternative embodiment, the target normal load of the target robot is determined based on the normal load determination model, according to the running acceleration and the unit normal vector of each effective contact point.

[0115] In one alternative embodiment, the mass of the target robot is obtained; based on the mass, running acceleration, and unit normal vector of each effective contact point, the wheel normal load of at least one wheel of the target robot is determined; the minimum wheel normal load is selected from the wheel normal loads of at least one wheel as the target normal load of the target robot.

[0116] Among them, the wheel normal load refers to the supporting reaction force that the ground (or contact surface) acts vertically on a single wheel through the effective contact point.

[0117] For example, the wheel normal load of at least one wheel in the target robot is determined based on mass, running acceleration, and unit normal vectors at each effective contact point. The method is as follows:

[0118]

[0119] Where m represents the mass of the target robot, and g represents the acceleration due to gravity. Indicates the number of effective contact points. This represents the unit normal vector (dimensionless) of the contact surface at the i-th effective contact point. Its direction is generally perpendicular to the contact surface and points outward from the robot; the specific definition depends on the coordinate system convention. It represents the running acceleration, which is the acceleration vector of the target robot's center of mass in the inertial coordinate system, i.e., the second derivative of the center of mass position vector.

[0120] S302 determines the stability margin of the target robot based on the center of gravity offset, roll angle, pitch angle, and target normal load.

[0121] The roll angle refers to the angle by which the target robot rotates around its direction of travel (X-axis). It describes the degree of lateral tilt of the robot. For example, when a robot travels laterally on a slope or one wheel lifts up, the robot will tilt from side to side; this tilt is the roll angle. An excessive roll angle is a direct cause of the robot tipping over to the side.

[0122] The pitch angle refers to the angle at which the target robot rotates around its lateral axis (Y-axis, i.e., the horizontal axis perpendicular to the direction of travel). It describes the degree to which the robot's nose rises or falls. For example, when the robot climbs a slope, its front rises, or when it goes down a slope or brakes suddenly, its front falls; this forward and backward pitch is the pitch angle. An excessive pitch angle may cause the robot to tip over or scrape its bottom when passing over a slope or step.

[0123] In some embodiments, the stability margin of the target robot is determined based on the center of gravity offset, roll angle, pitch angle, and target normal load as follows:

[0124]

[0125] in, Indicates stability margin, , , , All represent weights. This indicates the offset of the center of gravity. Indicates the roll angle. Indicates the pitch angle. This represents the wheel normal load of the i-th wheel.

[0126] In the above embodiments, a multi-dimensional stability margin comprehensive evaluation model is constructed by integrating geometric stability (the shortest distance from the center of mass projection point to the support area), dynamic mechanical characteristics (target normal load calculated from running acceleration and contact point normal vector), and real-time attitude information (roll angle, pitch angle). This method overcomes the limitations of traditional single criteria: on the one hand, the center of mass offset directly quantifies the spatial position risk of the robot relative to the support boundary and is the core geometric indicator for judging the overturning trend; on the other hand, by introducing the target normal load based on acceleration and contact mechanics, it can keenly detect early signs that the wheels are about to lift off the ground or that the adhesion is insufficient. The stability margin output after weighted fusion of the above multi-source information reflects both the current degree of robot attitude deviation and predicts its dynamic instability trend, thus providing the control system with a more accurate and sensitive decision-making basis, significantly improving the ability to identify overturning and lifting risks and the timeliness of early warning in complex terrain.

[0127] Figure 4This is a flowchart illustrating the steps for determining a valid contact point in one embodiment. In this embodiment, the operational data includes the lowest point of at least one wheel of the target robot; based on environmental data and operational data, determining at least one valid contact point on the future contact surface of the target robot includes the following steps:

[0128] Based on environmental data, S401 determines the obstacle height range on the future contact surface of the target robot.

[0129] The obstacle height range refers to the range of the lowest and highest possible obstacle heights detected by environmental data (such as LiDAR point clouds or depth images) on the future contact surface (within a preset forward travel distance). Because of sensor measurement errors, obstacle surface irregularities, or occlusions, the robot cannot obtain a precise single obstacle height value; therefore, a range is used to represent this uncertainty. It quantifies the degree of ambiguity in the terrain.

[0130] In some embodiments, based on an interval determination model, the obstacle height interval of the target robot on the future contact surface is determined according to environmental data.

[0131] For example, this embodiment models the obstacle height range, fully considering factors such as sensor measurement errors, local occlusion, and surface irregularities of the obstacle, to construct an uncertain obstacle height range for each wheel. This range is essentially the area comprised of the estimated "lowest height" and "highest height" of the obstacle surface within a predetermined detection distance in front of it. The lidar provides the geometric upper bound, and the depth camera provides local details and the lower bound for occlusion.

[0132] S402 determines the reference obstacle height based on the relationship between the width of the obstacle height range and the preset width threshold.

[0133] The interval width refers to the difference between the upper and lower limits of the obstacle height interval. The width directly reflects the degree of uncertainty in the perception results. A small width indicates reliable measurement results and relatively flat or well-defined terrain (such as a complete rock). A large width indicates ambiguous information, potentially including deep pits, soft grass, suspended obstacles, or severe occlusion, making it difficult for the robot to accurately determine the true terrain.

[0134] The reference obstacle height refers to a representative height value selected from the obstacle height range based on the relationship between the interval width and a preset width threshold, used for subsequent validity assessment. If the interval width is small (terrain is clear), the middle value is selected as the reference to maximize passage efficiency; if the interval width is large (terrain is uncertain), the maximum value is selected as the reference to adopt a conservative strategy to mitigate risks.

[0135] The preset width threshold can be determined based on the measurement error range of the sensors used, combined with the perception data of typical ruin terrain. The safety margin can be determined by experimentally measuring the maximum compression of the robot's suspension and tires during dynamic obstacle crossing, plus a conservative offset.

[0136] In one optional embodiment, if the width of the obstacle height range is less than a preset width threshold, the reference obstacle height is determined as a first preset height; if the width of the obstacle height range is not less than the preset width threshold, the reference obstacle height is determined as a second preset height; the second preset height is higher than the first preset height.

[0137] In one optional embodiment, if the width of the obstacle height interval is less than a preset width threshold, the midpoint of the interval is used as the reference obstacle height; if the width of the obstacle height interval is not less than the preset width threshold, the maximum value in the interval is used as the reference obstacle height. Here, the midpoint refers to the arithmetic mean of the obstacle height intervals. It represents the statistically expected value of the terrain height most likely to occur within the current uncertainty range. It is the best estimate when the confidence level is high (narrow interval).

[0138] For example, for each wheel, the midpoint and width of the obstacle height range for that wheel are calculated. If the width is less than a preset width threshold, it means the sensor's measurement of the terrain is relatively reliable, the height variation range is small, and the uncertainty is low (e.g., there is a flat, hard rock or paved road ahead). In this case, the midpoint of the range is used as the best estimate of the terrain height. If the width is greater than or equal to the preset width threshold, it means the terrain information is very vague and the uncertainty is high (e.g., there is a soft pile of rubble, dense grass, or severe obstruction that may hide cavities or protrusions). In this case, the maximum value of the range is used as the judgment criterion, and a conservative strategy is adopted.

[0139] S403 selects at least one effective contact point from the lowest point of each wheel based on the reference obstacle height and the preset safety threshold.

[0140] In one alternative embodiment, the lowest point of each wheel with a reference obstacle height greater than a preset safety threshold is selected as the effective contact point.

[0141] In one alternative embodiment, the equivalent obstacle height is determined based on the difference between the reference obstacle height and the preset safety threshold; for each wheel's lowest point, if the height between the wheel's lowest point and the future contact surface is less than the equivalent obstacle height, the wheel's lowest point is determined as the effective contact point.

[0142] The equivalent obstacle height refers to the threshold height used for actual comparison, obtained by subtracting a preset safety threshold from the reference obstacle height. This is a "discounted" height value to accommodate the effects of wheel size, suspension compression, and dynamic impacts. The height of the lowest point of the wheel is compared to the equivalent obstacle height: if the lowest point of the wheel is lower than the equivalent obstacle height, it means the wheel has enough space to press against the obstacle and make contact; if it is higher than the equivalent obstacle height, it means the wheel may be blocked by the obstacle or miss its target.

[0143] For example, if the height of the lowest point of the wheel is less than or equal to (the midpoint of the interval - the preset safety threshold), then the lowest point of the wheel is considered a valid contact point; otherwise, it is considered an invalid contact point. If the height of the lowest point of the wheel is less than or equal to (the maximum value of the interval - the preset safety threshold), then the lowest point of the wheel is considered a valid contact point; otherwise, it is considered an invalid contact point.

[0144] In the above embodiments, by introducing obstacle height ranges and their width thresholds, adaptive perception and judgment of complex terrain are achieved: when the range width is small, it indicates that the terrain is relatively clear, and using the midpoint of the range as the reference obstacle height can fully utilize terrain information and avoid misjudgment while ensuring safety, thereby improving the robot's passage efficiency; when the range width is large, it indicates that the terrain height is uncertain or there are perception blind spots, and using the maximum value of the range as the reference obstacle height can prioritize the prevention of potential risks with a conservative strategy, effectively avoiding wheel missteps or accidental contact caused by voids, soft sinkholes, or obstructions. Combining the reference obstacle height and the preset safety threshold to select effective contact points from the lowest point of each wheel significantly improves the accuracy and robustness of contact point identification in extreme unstructured environments such as ruins and rubble, laying a solid and reliable physical foundation for the subsequent construction of stable support areas and accurate assessment of stability margins.

[0145] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0146] Based on the same inventive concept, this application also provides a robot operation device for implementing the robot operation method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more robot operation device embodiments provided below can be found in the limitations of the robot operation method described above, and will not be repeated here.

[0147] In one exemplary embodiment, such as Figure 5 As shown, a robot operating device is provided, including: an acquisition module 501, a contact module 502, a support module 503, a margin module 504, and a control module 505, wherein:

[0148] The acquisition module 501 is used to acquire environmental data and operational data collected on the current contact surface during the operation of the target robot;

[0149] The contact module 502 is used to determine at least one effective contact point on the future contact surface of the target robot based on environmental data and operational data; the future contact surface is the contact surface of the target robot within a preset forward distance; the effective contact point is the point where there is contact between the wheel of the target robot and the future contact surface and the wheel bears the load.

[0150] The support module 503 is used to determine the support area of ​​the target robot based on the projection area enclosed by each effective contact point on the future contact surface; the support area is used to characterize the area required for the robot to maintain stability on the future contact surface.

[0151] The margin module 504 is used to determine the stability margin of the target robot based on the support area, operating data and the unit normal vector of each effective contact point; the stability margin is used to characterize the degree to which the robot will be suspended or overturned on the contact surface in the future.

[0152] The control module 505 is used to adjust the operating strategy of the target robot according to the relationship between the stability margin and the preset margin threshold.

[0153] In some embodiments, the margin module 504 is further configured to determine the center of gravity offset of the target robot based on the shortest distance from the center of gravity projection point of the target robot to the support area; and to determine the target normal load of the target robot based on the running acceleration and the unit normal vector of each effective contact point; and to determine the stability margin of the target robot based on the center of gravity offset, roll angle, pitch angle and target normal load.

[0154] In some embodiments, the margin module 504 is further configured to obtain the mass of the target robot; determine the wheel normal load of at least one wheel of the target robot based on the mass, running acceleration and the unit normal vector of each effective contact point; and select the smallest wheel normal load from the wheel normal loads of at least one wheel as the target normal load of the target robot.

[0155] In some embodiments, the contact module 502 is further configured to determine the obstacle height range on the future contact surface of the target robot based on environmental data; determine a reference obstacle height based on the relationship between the width of the obstacle height range and a preset width threshold; and select at least one effective contact point from the lowest point of each wheel based on the reference obstacle height and a preset safety threshold.

[0156] In some embodiments, the contact module 502 is further configured to use the midpoint of the obstacle height range as a reference obstacle height when the width of the obstacle height range is less than a preset width threshold; and to use the maximum value of the obstacle height range as a reference obstacle height when the width of the obstacle height range is not less than the preset width threshold.

[0157] In some embodiments, the contact module 502 is further configured to determine the equivalent obstacle height based on the difference between the reference obstacle height and the preset safety threshold; and for each wheel lowest point, if the height between the wheel lowest point and the future contact surface is less than the equivalent obstacle height, determine the wheel lowest point as an effective contact point.

[0158] Each module in the aforementioned robot operating device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.

[0159] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 6As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores data. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When the computer program is executed by the processor, it implements a robot operation method.

[0160] Those skilled in the art will understand that Figure 6 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0161] In one embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above method embodiments.

[0162] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the steps in the above method embodiments.

[0163] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0164] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0165] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0166] 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.

[0167] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for operating a robot, characterized in that, The method includes: Acquire environmental and operational data collected on the current contact surface during the operation of the target robot; Based on the environmental data and the operational data, at least one effective contact point on the future contact surface of the target robot is determined; the future contact surface is the contact surface of the target robot within a preset forward distance; the effective contact point is the point where there is contact between the wheel of the target robot and the future contact surface and it bears the load. The support region of the target robot is determined based on the projected area enclosed by each of the effective contact points on the future contact surface; the support region is used to characterize the area required for the robot to maintain stability on the future contact surface. Based on the support area, the operational data, and the unit normal vector of each effective contact point, the stability margin of the target robot is determined; the stability margin is used to characterize the degree to which the robot may become airborne or overturn on the future contact surface. The operating strategy of the target robot is adjusted according to the relationship between the stability margin and the preset margin threshold.

2. The method according to claim 1, characterized in that, The operational data includes the roll angle, pitch angle, and acceleration of the target robot; determining the stability margin of the target robot based on the support area, the operational data, and the unit normal vector of each effective contact point includes: The center of gravity offset of the target robot is determined based on the shortest distance from the projection point of the target robot's center of mass to the support area; and, The target normal load of the target robot is determined based on the running acceleration and the unit normal vector of each effective contact point; The stability margin of the target robot is determined based on the center of gravity offset, the roll angle, the pitch angle, and the target normal load.

3. The method according to claim 2, characterized in that, Determining the target normal load of the target robot based on the unit normal vector of each effective contact point and the running acceleration includes: Obtain the mass of the target robot; Based on the mass, the running acceleration, and the unit normal vector of each effective contact point, determine the wheel normal load of at least one wheel in the target robot; The minimum wheel normal load is selected from the wheel normal loads of at least one wheel as the target normal load of the target robot.

4. The method according to claim 1, characterized in that, The operational data includes the lowest point of at least one wheel of the target robot; determining at least one effective contact point of the target robot on the future contact surface based on the environmental data and the operational data includes: Based on the environmental data, the obstacle height range of the target robot on the future contact surface is determined; The reference obstacle height is determined based on the relationship between the width of the obstacle height range and a preset width threshold. Based on the reference obstacle height and the preset safety threshold, at least one effective contact point is selected from the lowest point of each wheel.

5. The method according to claim 4, characterized in that, Determining the reference obstacle height based on the relationship between the width of the obstacle height range and a preset width threshold includes: If the width of the obstacle height range is less than a preset width threshold, the midpoint of the obstacle height range shall be used as the reference obstacle height. If the width of the obstacle height range is not less than a preset width threshold, the maximum value in the obstacle height range shall be used as the reference obstacle height.

6. The method according to claim 4, characterized in that, The step of selecting at least one effective contact point from the lowest point of each wheel based on the reference obstacle height and a preset safety threshold includes: The equivalent obstacle height is determined based on the difference between the reference obstacle height and the preset safety threshold. For each of the wheel's lowest points, if the height between the wheel's lowest point and the future contact surface is less than the equivalent obstacle height, the wheel's lowest point is determined as the effective contact point.

7. A robot operating device, characterized in that, The device includes: The acquisition module is used to acquire environmental and operational data collected on the current contact surface during the operation of the target robot; A contact module is used to determine at least one effective contact point on the future contact surface of the target robot based on the environmental data and the operational data; the future contact surface is the contact surface of the target robot within a preset forward distance; the effective contact point is the point where there is contact between the wheel of the target robot and the future contact surface and the point that bears the load; A support module is used to determine the support area of ​​the target robot based on the projection area enclosed by each of the effective contact points on the future contact surface; the support area is used to characterize the area required for the robot to maintain stability on the future contact surface; The margin module is used to determine the stability margin of the target robot based on the support area, the operating data, and the unit normal vector of each effective contact point; the stability margin is used to characterize the degree to which the robot will become airborne or overturned on the future contact surface; The control module is used to adjust the operating strategy of the target robot according to the relationship between the stability margin and the preset margin threshold.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

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

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images