A robot collision detection calibration, collision detection method, device and system

Abstract

The embodiment of the application belongs to the technical field of robot collision detection, and relates to a collision detection calibration method of a robot. The method comprises the following steps: acquiring a calibration grid corresponding to the robot; determining position information of a contact point at which an actuator presses the surface of the robot based on the calibration grid; generating a control instruction based on the position information of the contact point and in combination with a preset action strategy, and instructing the actuator to press each contact point on the surface of the robot with a corresponding force according to the preset action strategy through the control instruction; acquiring corresponding calibration haptic information of each contact point each time the contact point is pressed; and constructing collision detection calibration data of the robot by using the position information of each contact point, the corresponding force and the corresponding calibration haptic information. The embodiment of the application also relates to related robot collision detection methods, devices and systems, etc. The technical solution of the application can improve the calibration efficiency of robot collision detection.

Description

Technical Field

[0001] This application relates to the field of robot collision detection technology, and in particular to a robot collision detection calibration, collision detection method, device and system. Background Technology

[0002] Collision detection is widely used in robots (such as cleaning robots, AGVs, and humanoid robots with mobility functions). Taking cleaning robots as an example, collision detection can prevent them from getting stuck or damaging furniture.

[0003] Currently, various similar robots on the market, including robotic vacuum cleaners, typically use collision detection with a collision plate equipped with an optical coupler sensor. As an improvement, robots based on force / tactile sensors for collision detection have just been introduced. Due to factors such as installation position errors of the force / tactile sensors and inconsistent remanence of permanent magnets, factory calibration is necessary to ensure consistency in collision position detection across a large number of products.

[0004] In existing technologies, calibration is often performed manually, which is time-consuming and inefficient. Summary of the Invention

[0005] The purpose of this application is to provide a robot collision detection calibration, collision detection method, apparatus and system to improve the calibration efficiency of robot collision detection.

[0006] In a first aspect, embodiments of this application provide a robot collision detection calibration method, which adopts the following technical solution:

[0007] A collision detection calibration method for a robot is applied to a robot calibration system. The system includes an actuator, a fixing device, and a first controller. The fixing device is used to fix the robot. The force / tactile sensor includes multiple sensing units arranged in an array. The multiple sensing units arranged in an array are disposed on the surface of the robot at preset positions. The method includes the following steps:

[0008] Obtain the calibration grid corresponding to the robot;

[0009] Based on the calibration grid, determine the position information of the contact point where the actuator presses against the surface of the robot;

[0010] Based on the location information of the contact point and combined with the preset action strategy, a control command is generated. The control command instructs the actuator to press the contact point on the surface of the robot with a preset calibration force according to the preset action strategy.

[0011] Acquire the calibrated tactile information corresponding to each press;

[0012] The collision detection calibration data of the robot is constructed based on the position information of the corresponding contact point, the calibration force, and the calibration tactile information.

[0013] Furthermore, determining the position information of the contact point where the actuator presses against the robot's surface based on the calibration grid specifically includes the following steps:

[0014] Construct the robot's coordinate system;

[0015] At least a portion of the nodes in the calibration mesh are extracted as the contact points;

[0016] Extract the position information of the contact point in the robot coordinate system.

[0017] Furthermore, based on the location information of the contact point and in conjunction with a preset action strategy, a control command is generated, specifically including the following steps:

[0018] Obtain the position information of the current contact point; the position information of the current contact point is obtained from the position information of the plurality of contact points arranged in sequence;

[0019] Based on the current contact point position information, a motion command is generated to instruct the actuator to move to the position corresponding to the current contact point;

[0020] When the actuator moves to the position corresponding to the current contact point, a pressing command is generated to instruct the actuator to press the position corresponding to the current contact point with the current force.

[0021] Furthermore, the step of constructing collision detection calibration data for the robot based on the position information of the corresponding contact point, the calibration force, and the calibration tactile information specifically includes the following steps:

[0022] Construct a list relating the location information of the contact point, the calibration force, and the calibration tactile information;

[0023] Use the list as the calibration data; or,

[0024] Based on the position information of the corresponding contact point, the calibration force, and the calibration tactile information, a mapping function is constructed to establish the correspondence between the position information, force, and tactile information of the collision detection point.

[0025] The mapping function is used as calibration data.

[0026] Furthermore, the step of constructing collision detection calibration data for the robot based on the position information of the corresponding contact point, the calibration force, and the calibration tactile information specifically includes the following steps:

[0027] The calibrated tactile information is used as sample data, and the corresponding position information of the contact point and the corresponding force are used as labels for the sample data.

[0028] The collision detection model is iteratively trained based on the sample data and the corresponding labels.

[0029] The trained collision detection model is used as the calibration data.

[0030] Furthermore, before obtaining the calibration mesh corresponding to the robot, the following steps are also included:

[0031] Obtain the robot model corresponding to the robot;

[0032] The calibration mesh is constructed based on the robot model.

[0033] Secondly, embodiments of this application provide a collision detection method for a robot based on calibration data calibrated using any of the robot collision detection calibration methods described above, the method comprising the following steps:

[0034] Obtain the currently detected tactile information;

[0035] Based on the current detected tactile information and combined with the collision detection calibration data, the current collision force information is predicted; the current collision force information includes: current force information and current force location information.

[0036] Thirdly, embodiments of this application provide a collision detection calibration device for a robot, the device comprising:

[0037] The mesh acquisition module is used to acquire the calibration mesh corresponding to the robot;

[0038] The contact determination module is used to determine the position information of the contact point where the actuator presses on the surface of the robot based on the calibration grid;

[0039] The instruction generation module is used to generate control instructions based on the position information of the contact point and in combination with a preset action strategy. The control instructions instruct the actuator to press the contact point on the surface of the robot with a preset calibration force according to the preset action strategy.

[0040] The tactile acquisition module is used to acquire the calibrated tactile information corresponding to each press.

[0041] The calibration construction module is used to construct collision detection calibration data for the robot based on the position information of the corresponding contact point, the calibration force, and the calibration tactile information.

[0042] Fourthly, embodiments of this application provide a collision detection device for a robot, the device comprising:

[0043] The information acquisition module is used to acquire the currently detected tactile information;

[0044] The collision prediction module is used to predict the current collision force information based on the currently detected tactile information and the collision detection calibration data; the current collision force information includes: current force information and current force location information.

[0045] Fifthly, embodiments of this application provide a robot collision detection calibration system, the system comprising: an actuator, a fixing device, and a first controller; a force / tactile sensor comprising multiple sensing units arranged in an array; the multiple sensing units arranged in an array are positioned at preset locations on the collision detection surface of the robot;

[0046] The fixing device is used to fix the robot;

[0047] The force / tactile sensor is used to collect the calibrated tactile information;

[0048] The first controller is communicatively connected to both the force / tactile sensor and the actuator.

[0049] The first controller is used to implement the steps of the collision detection and calibration method for the robot described in any of the above embodiments.

[0050] Sixthly, embodiments of this application provide a controller, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the robot collision detection calibration method and / or robot collision detection method described above.

[0051] In a seventh aspect, embodiments of this application provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the robot collision detection calibration method and / or the robot collision detection method described above.

[0052] Compared with the prior art, the embodiments of this application have the following main advantages:

[0053] This application embodiment determines the position information of the contact points where the actuator and the robot surface make pressing contact based on the calibration grid corresponding to the robot. Based on the position information of the contact points and combined with the preset action strategy, control commands are generated. The control commands instruct the actuator to press each contact point on the robot surface with a preset calibration force according to the preset action strategy. The corresponding calibration tactile information of each contact point is obtained each time it is pressed. The collision detection calibration data of the robot is constructed with the position information of each contact point, the preset calibration force, and the corresponding calibration tactile information to realize the automatic calibration of robot collision detection. Attached Figure Description

[0054] To more clearly illustrate the solutions in this application, the accompanying drawings used in the description of the embodiments of this application will be briefly introduced below. Obviously, the accompanying drawings described below are 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 architectural diagram of one embodiment of the robot system to which this application can be applied.

[0056] Figure 2 This is a schematic diagram of the calibration grid corresponding to the robot in this application.

[0057] Figure 3 Based on Figure 2 A schematic diagram of the structure used to determine the contact points using the calibration grid.

[0058] Figure 4 This is a flowchart illustrating one embodiment of the robot collision detection and calibration method of this application.

[0059] Figure 5 This is a flowchart illustrating one embodiment of the robot collision detection method of this application.

[0060] Figure 6 This is a structural block diagram of one embodiment of the robot collision detection and calibration device of this application.

[0061] Figure 7 This is a structural block diagram of one embodiment of the robot collision detection device of this application.

[0062] Figure 8 This is a schematic diagram of the structure of one embodiment of the computer device of this application. Detailed Implementation

[0063] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein in the specification of the application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application; the terms "comprising" and "having," and any variations thereof, in the specification, claims, and foregoing drawings of this application, are intended to cover non-exclusive inclusion. The terms "first," "second," etc., in the specification, claims, or foregoing drawings of this application are used to distinguish different objects, not to describe a particular order.

[0064] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0065] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0066] like Figure 1 As shown, Figure 1 This is an architecture diagram of one embodiment of the calibration system to which this application can be applied.

[0067] This application provides a robot calibration system 100 for calibrating a robot 200. The system includes an actuator 110, a fixing device 130, and a first controller 140. The fixing device 130 is used to fix the robot 200; the force / tactile sensor 300 includes multiple sensing units arranged in an array; the multiple sensing units arranged in an array are positioned at preset locations on the collision detection surface of the robot.

[0068] Specifically, the robot described in this application embodiment can be any existing or future robot that needs to perform collision detection during movement, such as: cleaning robot (e.g., sweeping robot, mopping robot, window cleaning robot), AGV vehicle, humanoid robot or humanoid robot chassis.

[0069] The robot 200 includes a collision detection surface. The force / tactile sensor 300 includes multiple sensing units arranged in an array; these multiple sensing units are positioned at preset locations on the robot's collision detection surface. For example, taking a cubic robotic vacuum cleaner as an example, the four sides of the cubic robotic vacuum cleaner can typically be used as the collision detection surface. The corresponding sensing units can be set according to factors such as the detection range of the cubic robotic vacuum cleaner, its actual size, and the detection range of each sensing unit. For example, such as... Figure 1 As shown, two sensor units can be installed on each of the four sides of the cube-shaped robot vacuum cleaner.

[0070] The first controller 130 is connected to the actuator 110 and the force / tactile sensor 300 via wired or wireless means.

[0071] It should be noted that the aforementioned wireless connection methods may include, but are not limited to, 3G / 4G / 5G connections, WiFi connections, Bluetooth connections, WiMAX connections, Zigbee connections, UWB (ultra-wideband) connections, and other currently known or future wireless connection methods.

[0072] Force / tactile sensor

[0073] Force / tactile sensors (hereinafter referred to as "sensors") refer to array-type force sensors and / or tactile sensors. Array-type force / tactile sensors comprise multiple sensing units arranged in an array, enabling the perception of force distribution information at multiple points. Typically, each sensing unit includes a data acquisition unit and a sensing unit arranged correspondingly to each other. The data acquisition unit is used to acquire force / tactile related signals; the sensing unit is used to generate force / tactile information (hereinafter referred to as "tactile information") based on the acquired signals.

[0074] The force / tactile sensor described in the embodiments of this application may be, but is not limited to, capacitive, inductive, photoelectric, Hall effect, piezoresistive, piezoelectric, or visual.

[0075] Specifically, force sensors can be, but are not limited to, one-dimensional or multi-dimensional force sensors used to measure pressure or three-dimensional force data, etc.

[0076] Tactile sensors measure contact force information with objects. This contact force information includes, but is not limited to, array-based multidimensional contact force information, surface deformation information, temperature information, and texture information. The contact surface between the tactile sensor and the object is typically flexible and has good resilience. The implementation of a tactile sensor includes a flexible contact surface, sensing circuitry, computing devices, and contact force information parsing algorithms. Compared to traditional force sensors, tactile sensors can more sensitively perceive various forces from multiple dimensions; for example, they can detect dense tangential frictional forces. Tactile sensors can be applied in various fields as needed. For example, tactile sensors can be placed in dexterous hands or other grasping actuators to measure contact force information while working with the dexterous hand to achieve grasping functions, thereby enabling the grasping of objects of different shapes and softness.

[0077] Actuator

[0078] The actuator 110 includes an actuator body and a pressing actuator (e.g., a pressure head, to apply a calibration force to the surface of the robot) disposed at the actuation end of the actuator body, and the pressing actuator applies a calibration force to the surface of the robot.

[0079] Specifically, the actuator body may be, but is not limited to: a robot (e.g., a robotic arm or humanoid robot); an XYZ platform or an actuator that includes an XYZ platform.

[0080] For ease of understanding, such as Figure 1 As shown, this application embodiment mainly takes the robotic arm 111 as the actuator body as an example for detailed description, and the execution end of the robotic arm 111 is provided with a pressing actuator 112.

[0081] like Figure 2 As shown, Figure 2 This is a three-dimensional structural schematic diagram of an embodiment of the pressing actuator of the calibration system of this application.

[0082] In one embodiment, the pressing actuator 112 described in this application may include: a mounting plate 1121 and a pressing head 1122 extending from the mounting plate toward the actuating end away from the actuator body.

[0083] In one embodiment, the free end of the pressure head 1122 is arc-shaped, thereby reducing the occurrence of damage to the robot's surface caused by the free end of the pressure head during the application of calibration force to the robot.

[0084] Mounting plate 1121 is used to fix pressure head 1122 to the execution end of actuator body.

[0085] In one embodiment, the mounting plate may have mounting holes to be fixedly connected to the actuator body's actuator end by means of fixing bolts, fixing pins, etc. passing through the mounting holes; otherwise, any other existing or future-developed structure or method may be used to achieve the fixed connection between the pressing actuator and the actuator body's actuator end, such as: fixed connection by mutually cooperating bayonet and slot; by adhesive bonding; or prefabrication into one piece.

[0086] In one embodiment, the diameter of the pressure head can be 4mm to 6mm.

[0087] In this embodiment, by using the aforementioned diameter of the pressure head, the contact area between the pressure head and the sensor surface during calibration experiments can be reduced due to an excessively large diameter, which would result in a smaller strain on the sensor surface under the same pressure, thus weakening the detected displacement signal. In addition, it can also reduce the excessive pressure on the sensor caused by an excessively small contact area between the pressure head and the sensor surface during calibration experiments, which could lead to the sensor being punctured.

[0088] It should be noted that, in addition to the point contact pressure head described in the above embodiments, the pressure head can also adopt any other shape as needed, such as: straight, square, etc., and the pressure head can apply line load or surface load to the surface of the robot.

[0089] It should be noted that the aforementioned pressure head is usually made of rigid (and insulating or non-magnetic) materials such as aluminum alloy or ABS engineering plastic to prevent interference with the sensor to be calibrated, thereby reducing the calibration accuracy of the sensor.

[0090] Fixture

[0091] Specifically, the fixing device 130 can be any device that can fix the robot. For example, the fixing device 130 may include a worktable 131 and a fixing clamp 132. The fixing clamp 132 is provided on the worktable 131, and the robot 200 to be calibrated is fixed in position by the fixing clamp 132.

[0092] Specifically, the aforementioned fixture can be any existing or future structure that can serve to fix the sensor.

[0093] First controller

[0094] The first controller 130 is used to execute the steps of the collision detection calibration method for the robot described in the embodiments of this application.

[0095] The collision detection and calibration method for robots provided in this invention can be applied to computer terminals (PCs); industrial personal computers (IPCs); mobile terminals; servers; systems including terminals and servers, implemented through interaction between the terminals and servers; programmable logic controllers (PLCs); field-programmable gate arrays (FPGAs); digital signal processors (DSPs) or microcontroller units (MCUs) and similar first controllers. The first controller generates program instructions based on a pre-defined program and / or in conjunction with data signals collected from external actuators, etc. Specifically, it can be applied to, for example... Figure 8 The computer equipment shown.

[0096] Based on the robot calibration system described in the above embodiments, this application provides a collision detection calibration method for a robot. This method is generally executed by a first controller 130. Accordingly, the collision detection calibration device for the robot described in the following embodiments is generally disposed in the first controller 130.

[0097] like Figure 4 As shown, Figure 4 This is a flowchart of an embodiment of the robot collision detection calibration method of this application; the robot collision detection calibration method may include the following method steps:

[0098] Step 210: Obtain the calibration grid corresponding to the robot.

[0099] Step 220 determines the location information of the contact point where the actuator and the robot's surface make pressing contact based on the calibration mesh.

[0100] Step 230 generates control commands based on the position information of the contact points and in conjunction with the preset action strategy. The control commands instruct the actuator to press the contact points on the robot's surface with a preset calibrated force according to the preset action strategy.

[0101] Step 240: Obtain the calibrated tactile information corresponding to each press.

[0102] Step 250 constructs collision detection calibration data for the robot based on the position information of the corresponding contact point, the calibration force, and the calibration tactile information.

[0103] This application embodiment determines the position information of the contact points where the actuator and the robot surface make pressing contact based on the calibration grid corresponding to the robot. Based on the position information of the contact points and combined with the preset action strategy, control commands are generated. The control commands instruct the actuator to press each contact point on the robot surface with a preset calibration force according to the preset action strategy. The calibration tactile information corresponding to each contact point being pressed is obtained. The collision detection calibration data of the robot is constructed with the position information of each contact point, the preset calibration force, and the corresponding calibration tactile information to realize the automatic calibration of robot collision detection, thereby improving the calibration efficiency.

[0104] To facilitate understanding, the above methods and steps will be explained in further detail below.

[0105] Step 210: Obtain the calibration grid corresponding to the robot.

[0106] In one embodiment, the first controller retrieves a pre-built and stored calibration mesh corresponding to the robot model or the real-time generated calibration mesh from a memory or server according to a preset address.

[0107] In one embodiment, before step 210 obtains the calibration mesh corresponding to the robot, the following method steps may also be included:

[0108] Obtain the robot model corresponding to the robot.

[0109] Step 260: Obtain the robot model corresponding to the robot.

[0110] In one embodiment, the first controller can retrieve a pre-generated CAD model of the robot from a memory or server according to a preset address, or retrieve a 3D model of the robot generated in real time based on the acquired images of the robot.

[0111] For example, for ease of understanding, this application uses a cube-shaped sweeping robot as an example for detailed description, and the robot model is a cube.

[0112] Step 270: Construct a calibration mesh based on the robot model.

[0113] It should be noted that the specifications of the calibration grid can be set based on the application scenario requirements of the robot, the recognition accuracy, and the robot's own self-stored specifications.

[0114] Continuing with the example above, such as Figure 2As shown, a recognition accuracy of less than 3cm is generally acceptable. Therefore, given that the side length of the cleaning robot in this embodiment is 24cm, the four side detection surfaces (front, back, left, and right) of the cube robot can be divided into 8 equal parts along the X / Y direction and 4 equal parts along the Z direction; the top detection surface can be divided into 8 equal parts along the X / Y direction to form a calibration grid.

[0115] This application embodiment obtains a robot model corresponding to the robot and constructs a virtual calibration mesh based on the robot model. This facilitates the subsequent determination of contact points based on the nodes of the calibration mesh, and the virtual contact points determined based on the calibration mesh can then be transformed into the real robot coordinate system.

[0116] Step 220 determines the location information of the contact point where the actuator and the robot's surface make pressing contact based on the calibration mesh.

[0117] In one embodiment, before step 220 determines the location information of the contact point where the actuator and the robot's surface make pressing contact based on the calibration mesh, the following method steps may also be included:

[0118] Step 280: Construct the robot coordinate system.

[0119] In this embodiment, the robot is pre-positioned using a clamp.

[0120] For example, taking a cube-shaped robotic vacuum cleaner as an example, a three-dimensional rectangular coordinate system can be established with the center of the lower end of the back of the robot's main direction of travel as the origin, the perpendicular direction of travel as the X-axis, the direction in which the main direction of travel points as the y-axis, and the height direction as the Z-axis.

[0121] Step 290: Extract at least some nodes from the calibration mesh as contact points.

[0122] For example, such as Figure 3 As shown, the contact points in the calibration grid can be determined, such as the four corner points of each grid. In this example, a total of 7*7=49 contact points were generated on the top surface, a total of 4*3*7=84 contact points were generated on the sides, and a total of 4*3=12 contact points were generated at the corners.

[0123] Step 300: Determine the coordinates of the contact point in the robot coordinate system.

[0124] In this embodiment of the application, since a robot coordinate system is constructed and the robot model corresponds to the robot, the coordinate system can also be applied to the robot model. Based on the robot model coordinate system and the robot model, the coordinates of each mesh contact point in the robot coordinate system can be determined.

[0125] For example, taking the robot model as a CAD model, the coordinates of each determined mesh contact point in the robot model coordinate system can be obtained, that is, the coordinates of each point in the robot coordinate system.

[0126] The embodiments of this application obtain the coordinates of the contact points in the robot coordinate system by constructing a robot coordinate system, extracting at least some nodes in the calibration mesh as contact points, and then determining the coordinates of the contact points in the robot coordinate system.

[0127] Step 230 generates control commands based on the position information of the contact points and in conjunction with the preset action strategy. The control commands instruct the actuator to press the contact points on the robot's surface with a preset calibrated force according to the preset action strategy.

[0128] Specifically, the aforementioned preset action strategy can be set arbitrarily as needed. For example, taking a point-type pressure head as an example, the same force can be applied sequentially to each contact point, and then the force can be gradually changed (e.g., increased or decreased). After each change in force, the same force is applied to each contact point repeatedly until a preset termination condition is met. For example, the force change range can be set to: 5N, 15N, 25N, 40N, 60N, 90N, 110N, 130N. Then, under each force, the actuator is controlled by a pressing command to sequentially apply the same force to each contact point.

[0129] Furthermore, based on the previous examples, when the pressure head is in the shape of a straight line or a square, it can apply line loads or surface loads to the surface of the force / tactile sensor. This requires pressing a group of contact points (i.e., multiple contact points) arranged linearly or planarly each time. Depending on the form of the applied load, collisions between the robot and obstacles in different scenarios can be simulated, allowing for the development of different action strategies.

[0130] In one embodiment, step 230 generates a control command based on the location information of the contact point and a preset action strategy, which may specifically include the following method steps:

[0131] Step 231: Obtain the position information of the current contact point; the position information of the current contact point is obtained from the position information of multiple contact points arranged in sequence.

[0132] Based on the previous embodiments, depending on the shape of the pressure head, the current contact point location information can be the location information of one contact point or it can correspond to multiple contact points.

[0133] Step 232 generates motion commands based on the current contact point position information to instruct the actuator to move to the position on the robot surface corresponding to the current contact point.

[0134] Step 233 generates a pressing command when the actuator moves to the position corresponding to the current contact point, instructing the actuator to press the position corresponding to the current contact point with the current force.

[0135] Specifically, the position information of the current contact point that the actuator needs to reach can be used as the target position for trajectory planning, and then the robot is instructed to move to the current contact point each time.

[0136] Alternatively, based on the above embodiments, since the robot moves in a straight line, the trajectory control of the actuator can be easily achieved simply by setting the actuator's direction of travel and distance parameters as needed.

[0137] This application embodiment obtains the position information of sequentially arranged contact points, generates motion commands based on the position information of sequentially arranged contact points, instructs the actuator to move sequentially to each contact point on the robot surface, and then generates a pressing command to instruct the actuator to press the position corresponding to the current contact point according to the preset current pressing force.

[0138] Step 240: Obtain the calibrated tactile information corresponding to each press.

[0139] Specifically, each time a contact point on the surface of the robot is pressed with the current preset calibration pressing force, the first controller can obtain tactile information collected by the force / tactile sensors arranged in a preset array in the robot or tactile information after some preprocessing from the memory or server according to the preset address, and use the tactile information as the calibration tactile information.

[0140] Taking the array-arranged force / tactile sensor described in the above embodiment, which includes eight sensing units, as an example, each press will correspondingly acquire eight sets of tactile information based on the eight sensing units. For example, using a magnetic tactile sensor, the aforementioned calibration tactile information may include: displacement change information (this displacement change information can be obtained based on magnetic field change information). In some cases, magnetic field change information can also be used as calibration tactile information. For ease of understanding, this application embodiment will continue to describe in detail using displacement change information as an example of calibration tactile information.

[0141] Step 250 constructs collision detection calibration data for the robot based on the position information of the corresponding contact point, the calibration force, and the calibration tactile information.

[0142] In this embodiment, each or every contact point location and corresponding calibration force corresponds to a set of calibration tactile information collected by an array of force / tactile sensing units. Therefore, collision detection calibration data for the robot can be constructed using various existing or future methods, based on the corresponding contact point location information, calibration force, and calibration tactile information. A detailed explanation is provided below with examples.

[0143] In an optional embodiment of this application, step 250 constructs collision detection calibration data for the robot based on the position information of the corresponding contact point, the calibration force, and the calibration tactile information. Specifically, this may include the following method steps:

[0144] Construct a list of the correspondence between the location information of the contact point, the calibrated force, and the calibrated tactile information;

[0145] Use a list as the label data.

[0146] This application embodiment constructs a list of correspondences between the location information of the contact point, the calibration force, and the calibration tactile information (e.g., displacement information). The list is used as calibration data, and subsequent processing such as table lookup and matching can be used to predict the current collision force information.

[0147] If the list is used as calibration data, due to nonlinear factors such as structure, material, and contact, it may not be possible to directly match the current collision force information during the actual collision detection process based on the list. In this case, a polynomial interpolation method, such as the Lagrange method or the Newton method, can be used to obtain the current collision force information through the difference.

[0148] In an optional embodiment of this application, step 250 constructs collision detection calibration data for the robot based on the position information of the corresponding contact point, the calibration force, and the calibration tactile information. Specifically, this may include the following method steps:

[0149] Step 251: Based on the position information of the corresponding contact point, the calibration force, and the calibration tactile information, construct a mapping function for the correspondence between the position information, force, and tactile information of the collision detection point.

[0150] The purpose of this application is to establish a continuous function to describe the mapping relationship between position information, force, and tactile information by recording the position information of the contact point, the calibration force, and the calibration tactile information (e.g., displacement information).

[0151] Step 253 uses the mapping function as calibration data.

[0152] In some cases, it is desirable to construct a mapping function that establishes the correspondence between the location information, force, and tactile information of the collision detection point. However, in other cases, it may be impossible to construct such a mapping function.

[0153] This application embodiment constructs a list of location information and corresponding force for each contact point, as well as a list of corresponding calibrated tactile information. The list is used as the calibration data, and subsequent processing such as table lookup and matching can be used to predict the current collision force information.

[0154] In an optional embodiment of this application, step 250 constructs collision detection calibration data for the robot based on the position information of the corresponding contact point, the calibration force, and the calibration tactile information. Specifically, this may include the following method steps:

[0155] Step 252 uses calibrated tactile information as sample data, and the corresponding contact point location information and corresponding force as labels for the sample data.

[0156] Step 254 Iteratively trains the collision detection model based on the sample data and corresponding labels.

[0157] Step 256 uses the trained collision detection model as calibration data.

[0158] This application embodiment uses calibrated tactile information as sample data, and the corresponding contact point position information and corresponding force as labels for the sample data. Based on the sample data and corresponding labels, a collision detection model is iteratively trained using supervised and semi-supervised methods. Subsequently, based on the collision detection calibration model, the current collision force information can be predicted for the current array tactile information, thereby improving the applicability and robustness of subsequent collision detection prediction.

[0159] This application embodiment determines the position information of the contact points where the actuator and the robot surface make pressing contact based on the calibration grid corresponding to the robot. Based on the position information of the contact points and combined with the preset action strategy, control commands are generated. The control commands instruct the actuator to press each contact point on the robot surface with a preset calibration force according to the preset action strategy. The corresponding calibration tactile information of each contact point is obtained each time it is pressed. The collision detection calibration data of the robot is constructed with the position information of each contact point, the preset calibration force, and the corresponding calibration tactile information to realize the automatic calibration of robot collision detection.

[0160] In one embodiment, this application also provides a robot collision detection system (figures omitted). This robot collision detection system includes the force / tactile sensor and the second controller found in the calibration system of the preceding embodiments.

[0161] The force / tactile sensor comprises multiple sensing units arranged in an array. For details regarding the force / tactile sensor, please refer to the description in the preceding embodiments; it will not be repeated here.

[0162] The second controller communicates with the force / tactile sensor via wired or wireless means.

[0163] It should be noted that the second controller and the first controller in the calibration system may be the same or different controllers, both of which are within the scope of protection of this application.

[0164] The collision detection method for robots provided in this invention can be applied to computer terminals (PCs); industrial personal computers (IPCs); mobile terminals; servers; systems including terminals and servers, implemented through interaction between the terminals and servers; programmable logic controllers (PLCs); field-programmable gate arrays (FPGAs); digital signal processors (DSPs) or microcontroller units (MCUs) and similar second controllers. The second controller generates program instructions based on a pre-set program and / or in conjunction with data signals collected from external actuators, etc. Specifically, it can be applied to, for example... Figure 8 The computer equipment shown.

[0165] Based on the robot system and robot collision detection calibration method described in the above embodiments, this application embodiment can also provide a robot collision detection method, which is generally executed by a second controller. Accordingly, the robot collision detection device described in the following embodiments is generally set in the second controller.

[0166] Step 410: Obtain the currently detected tactile information.

[0167] Step 420 predicts the current collision force information based on the current detected tactile information and the calibration data of the collision detection; the current collision force information includes: the magnitude of the current force and the position of the current force.

[0168] This application embodiment performs robot collision detection calibration in advance based on the robot collision detection calibration method described in the above embodiment. Subsequently, based on the acquired current detection data information and the calibration data, the current collision force information can be quickly predicted.

[0169] In an optional embodiment of this application, based on the above embodiments, if a list is used as calibration data, due to nonlinear factors such as structure, material, and contact, the corresponding current collision force information may not be directly matched during the actual collision detection process based on the list. Therefore, step 420 may specifically include the following method steps:

[0170] Step 421: Based on the current detection data, match the current collision force information corresponding to the current detection data in the list;

[0171] Step 422: If the current collision force information cannot be directly matched based on the list, obtain the collision force information of the nearest collision force in the list, and calculate the current force information based on the difference method.

[0172] Specifically, interpolation methods can be polynomial interpolation, such as the Lagrange method and Newton's method.

[0173] In this embodiment of the application, when the current collision force information corresponding to the current detection data information cannot be directly matched based on the list, the collision force information of the neighboring objects in the list is obtained, and the current force information can also be obtained based on the difference method.

[0174] 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. This computer program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The aforementioned storage medium can be a non-volatile storage medium such as a magnetic disk, optical disk, or read-only memory (ROM), or random access memory (RAM).

[0175] It should be understood that although the steps in the flowcharts of the accompanying figures are shown sequentially as indicated by 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 accompanying figures may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0176] Further reference Figure 6 As a response to the above Figure 4The implementation of the method shown in this application provides an embodiment of a robot collision detection calibration device, which is similar to... Figure 4 The method embodiments shown correspond to those described.

[0177] like Figure 6 As shown, the robot collision detection calibration device 500 described in this embodiment may include:

[0178] The mesh acquisition module 510 is used to acquire the calibration mesh corresponding to the robot.

[0179] The contact determination module 520 is used to determine the position information of the contact point where the actuator presses on the surface of the robot based on the calibration grid;

[0180] The instruction generation module 530 is used to generate control instructions based on the position information of the contact point and in combination with the preset action strategy. The control instructions instruct the actuator to press the contact point on the surface of the robot with a preset calibration force according to the preset action strategy.

[0181] The tactile acquisition module 540 is used to acquire the calibrated tactile information corresponding to each press.

[0182] The calibration construction module 550 is used to construct collision detection calibration data for the robot based on the position information of the corresponding contact point, the calibration force, and the calibration tactile information.

[0183] In an optional embodiment, the contact determination module 520 may include:

[0184] The coordinate construction submodule is used to construct the robot's coordinate system;

[0185] The contact extraction submodule is used to extract at least a portion of the nodes in the calibration mesh as contact points.

[0186] The position extraction submodule is used to extract the position information of the contact point in the robot coordinate system.

[0187] In an optional embodiment, the instruction generation module 530 may include:

[0188] The location acquisition submodule acquires the location information of the current contact point; the location information of the current contact point is obtained from the location information of multiple contact points arranged in sequence.

[0189] The motion generation submodule is used to generate motion commands based on the current contact point position information, so as to instruct the actuator to move to the position corresponding to the current contact point;

[0190] The press generation submodule is used to generate a press command when the actuator moves to the position corresponding to the current contact point, so as to instruct the actuator to press the current contact point with the current force.

[0191] In an optional embodiment, the calibration building module 550 may include:

[0192] The list building submodule is used to build a list of the correspondence between the location information of the contact point, the calibration force, and the calibration tactile information;

[0193] The first determination submodule is used to use a list as calibration data.

[0194] In an optional embodiment, the calibration building module 550 may include:

[0195] The function construction submodule is used to construct a mapping function for the correspondence between the position information, force and tactile information of the collision detection point based on the position information, calibration force and calibration tactile information of the corresponding contact point;

[0196] The second determination submodule is used to use the mapping function as calibration data.

[0197] In an optional embodiment, the calibration building module 550 may include:

[0198] The sample determination submodule is used to use calibrated tactile information as sample data, and the corresponding contact point location information and corresponding force as labels for the sample data.

[0199] The model training submodule is used to iteratively train the collision detection model based on sample data and corresponding labels.

[0200] The calibration determination submodule is used to use the trained collision detection model as calibration data.

[0201] In an optional embodiment, the robot collision detection calibration device 500 may further include:

[0202] The model acquisition submodule is used to acquire the robot model corresponding to the robot.

[0203] The mesh construction submodule is used to build calibration meshes based on robot models.

[0204] Further reference Figure 7 As a response to the above Figure 5 The implementation of the method shown in this application provides an embodiment of a collision detection device for a robot, which is similar to... Figure 5 The method embodiments shown correspond to those described.

[0205] like Figure 6As shown, the robot collision detection device 600 in this embodiment may include:

[0206] Information acquisition module 610 is used to acquire the currently detected tactile information;

[0207] The collision prediction module 620 is used to predict the current collision force information based on the currently detected tactile information and the collision detection calibration data; the current collision force information includes: the current force information and the position information of the current force.

[0208] To address the aforementioned technical problems, embodiments of this application also provide a computer device. Please refer to [link / reference needed]. Figure 8 , Figure 8 This embodiment presents a basic structural block diagram of a computer device.

[0209] The computer device can be a terminal or a server.

[0210] The computer device 7 includes a memory 71, a processor 72, and a network interface 73 that are interconnected via a system bus. It should be noted that only the computer device 7 with components 71-73 is shown in the figure; however, it should be understood that it is not required to implement all the shown components, and more or fewer components can be implemented alternatively. Those skilled in the art will understand that the computer device described here is a device capable of automatically performing numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes, but is not limited to, microprocessors, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0211] The computer device can be a desktop computer, laptop, handheld computer, or cloud server, etc. The computer device can interact with the user via a keyboard, mouse, remote control, touchpad, or voice control.

[0212] The memory 71 includes at least one type of readable storage medium, including flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 71 may be an internal storage unit of the computer device 7, such as the hard disk or memory of the computer device 7. In other embodiments, the memory 71 may also be an external storage device of the computer device 7, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the computer device 7. Of course, the memory 71 may include both the internal storage unit and its external storage device of the computer device 7. In this embodiment, the memory 71 is typically used to store the operating system and various application software installed on the computer device 7, such as the program code of robot collision detection calibration methods and / or robot collision detection methods. In addition, the memory 71 can also be used to temporarily store various types of data that have been output or will be output.

[0213] In some embodiments, the processor 72 may be a central processing unit (CPU), a first controller, a microcontroller, a microprocessor, or other data processing chip. The processor 72 is typically used to control the overall operation of the computer device 7. In this embodiment, the processor 72 is used to run program code stored in the memory 71 or process data, for example, to run program code for a robot collision detection calibration method and / or a robot collision detection method.

[0214] The network interface 73 may include a wireless network interface or a wired network interface, which is typically used to establish communication connections between the computer device 7 and other electronic devices.

[0215] This application also provides another embodiment, namely, providing a computer-readable storage medium storing a robot collision detection calibration and / or robot collision detection program, which can be executed by at least one processor to cause the at least one processor to perform the steps of the robot collision detection calibration method and / or robot collision detection method as described above.

[0216] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0217] Obviously, the embodiments described above are only some embodiments of this application, not all embodiments. The accompanying drawings show preferred embodiments of this application, but do not limit the patent scope of this application. This application can be implemented in many different forms; rather, the purpose of providing these embodiments is to provide a more thorough and comprehensive understanding of the disclosure of this application. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing specific embodiments, or make equivalent substitutions for some of the technical features. Any equivalent structures made using the content of this application's specification and drawings, directly or indirectly applied to other related technical fields, are similarly within the scope of patent protection of this application.

Claims

1. A collision detection and calibration method for a robot, applied to a robot calibration system, the system comprising an actuator, a fixing device, and a first controller; the fixing device for fixing the robot; a force / tactile sensor comprising multiple sensing units arranged in an array; the multiple sensing units arranged in an array are disposed on the surface of the robot at preset positions, characterized in that, The method includes the following steps: Obtain the calibration grid corresponding to the robot; Based on the calibration grid, determine the position information of the contact point where the actuator presses against the surface of the robot; Based on the location information of the contact point and combined with the preset action strategy, a control command is generated. The control command instructs the actuator to press the contact point on the surface of the robot with a preset calibration force according to the preset action strategy. Acquire the calibrated tactile information corresponding to each press; The collision detection calibration data of the robot is constructed based on the position information of the corresponding contact point, the calibration force, and the calibration tactile information.

2. The collision detection calibration method for a robot according to claim 1, characterized in that, The step of determining the position information of the contact point where the actuator presses against the surface of the robot based on the calibration grid specifically includes the following steps: Construct the robot's coordinate system; At least a portion of the nodes in the calibration mesh are extracted as the contact points; Extract the position information of the contact point in the robot coordinate system.

3. The collision detection calibration method for a robot according to claim 1 or 2, characterized in that, Based on the location information of the contact point and in conjunction with a preset action strategy, a control command is generated, specifically including the following steps: Obtain the position information of the current contact point; the position information of the current contact point is obtained from the position information of the plurality of contact points arranged in sequence; Based on the current contact point position information, a motion command is generated to instruct the actuator to move to the position corresponding to the current contact point; When the actuator moves to the position corresponding to the current contact point, a pressing command is generated to instruct the actuator to press the position corresponding to the current contact point with the current force.

4. The collision detection calibration method for a robot according to claim 1 or 2, characterized in that, The process of constructing collision detection calibration data for the robot based on the position information of the corresponding contact point, the calibration force, and the calibration tactile information specifically includes the following steps: Construct a list relating the location information of the contact point, the calibration force, and the calibration tactile information; Use the list as the calibration data; or, Based on the position information of the corresponding contact point, the calibration force, and the calibration tactile information, a mapping function is constructed to establish the correspondence between the position information, force, and tactile information of the collision detection point. The mapping function is used as calibration data.

5. The collision detection calibration method for a robot according to claim 1 or 2, characterized in that, The process of constructing collision detection calibration data for the robot based on the position information of the corresponding contact point, the calibration force, and the calibration tactile information specifically includes the following steps: Construct a list relating the location information of the contact point, the calibration force, and the calibration tactile information; The calibrated tactile information in the list is used as sample data, and the corresponding position information of the contact point and the corresponding force are used as labels for the sample data. The collision detection model is iteratively trained based on the sample data and the corresponding labels. The trained collision detection model is used as the calibration data.

6. The collision detection calibration method for a robot according to claim 1 or 2, characterized in that, Before obtaining the calibration grid corresponding to the robot, the following steps are also included: Obtain the robot model corresponding to the robot; The calibration mesh is constructed based on the robot model.

7. A collision detection method for a robot based on calibration data calibrated using the collision detection calibration method for a robot according to any one of claims 1 to 6, characterized in that, The method includes the following steps: Obtain the currently detected tactile information; Based on the current detected tactile information and combined with the collision detection calibration data, the current collision force information is predicted; the current collision force information includes: current force information and current force location information.

8. A collision detection and calibration device for a robot, characterized in that, The device includes: The mesh acquisition module is used to acquire the calibration mesh corresponding to the robot; The contact determination module is used to determine the position information of the contact point where the actuator presses on the surface of the robot based on the calibration grid; The instruction generation module is used to generate control instructions based on the position information of the contact point and in combination with a preset action strategy. The control instructions instruct the actuator to press the contact point on the surface of the robot with a preset calibration force according to the preset action strategy. The tactile acquisition module is used to acquire the calibrated tactile information corresponding to each press. The calibration construction module is used to construct collision detection calibration data for the robot based on the position information of the corresponding contact point, the calibration force, and the calibration tactile information.

9. A collision detection device for a robot, characterized in that, The device includes: The information acquisition module is used to acquire the currently detected tactile information; The collision prediction module is used to predict the current collision force information based on the currently detected tactile information and the collision detection calibration data; the current collision force information includes: current force information and current force location information.

10. A robot collision detection and calibration system, characterized in that, The system includes: an actuator, a fixing device, and a first controller; the force / tactile sensor includes multiple sensing units arranged in an array; the multiple sensing units arranged in an array are positioned at preset locations on the collision detection surface of the robot; The fixing device is used to fix the robot; The force / tactile sensor is used to collect the calibrated tactile information; The first controller is communicatively connected to both the force / tactile sensor and the actuator. The first controller is used to implement the steps of the collision detection calibration method for the robot according to any one of claims 1 to 6.

Images